1 acpi= [HW,ACPI,X86,ARM64,RISCV64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64 and RISCV64, ONLY "acpi=off", "acpi=on" or 14 "acpi=force" are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 { vendor | video | native | none } 26 If set to vendor, prefer vendor-specific driver 27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 28 of the ACPI video.ko driver. 29 If set to video, use the ACPI video.ko driver. 30 If set to native, use the device's native backlight mode. 31 If set to none, disable the ACPI backlight interface. 32 33 acpi_force_32bit_fadt_addr 34 force FADT to use 32 bit addresses rather than the 35 64 bit X_* addresses. Some firmware have broken 64 36 bit addresses for force ACPI ignore these and use 37 the older legacy 32 bit addresses. 38 39 acpica_no_return_repair [HW, ACPI] 40 Disable AML predefined validation mechanism 41 This mechanism can repair the evaluation result to make 42 the return objects more ACPI specification compliant. 43 This option is useful for developers to identify the 44 root cause of an AML interpreter issue when the issue 45 has something to do with the repair mechanism. 46 47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 49 Format: <int> 50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 51 debug output. Bits in debug_layer correspond to a 52 _COMPONENT in an ACPI source file, e.g., 53 #define _COMPONENT ACPI_EVENTS 54 Bits in debug_level correspond to a level in 55 ACPI_DEBUG_PRINT statements, e.g., 56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 57 The debug_level mask defaults to "info". See 58 Documentation/firmware-guide/acpi/debug.rst for more information about 59 debug layers and levels. 60 61 Enable processor driver info messages: 62 acpi.debug_layer=0x20000000 63 Enable AML "Debug" output, i.e., stores to the Debug 64 object while interpreting AML: 65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 66 Enable all messages related to ACPI hardware: 67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 68 69 Some values produce so much output that the system is 70 unusable. The "log_buf_len" parameter may be useful 71 if you need to capture more output. 72 73 acpi_enforce_resources= [ACPI] 74 { strict | lax | no } 75 Check for resource conflicts between native drivers 76 and ACPI OperationRegions (SystemIO and SystemMemory 77 only). IO ports and memory declared in ACPI might be 78 used by the ACPI subsystem in arbitrary AML code and 79 can interfere with legacy drivers. 80 strict (default): access to resources claimed by ACPI 81 is denied; legacy drivers trying to access reserved 82 resources will fail to bind to device using them. 83 lax: access to resources claimed by ACPI is allowed; 84 legacy drivers trying to access reserved resources 85 will bind successfully but a warning message is logged. 86 no: ACPI OperationRegions are not marked as reserved, 87 no further checks are performed. 88 89 acpi_force_table_verification [HW,ACPI] 90 Enable table checksum verification during early stage. 91 By default, this is disabled due to x86 early mapping 92 size limitation. 93 94 acpi_irq_balance [HW,ACPI] 95 ACPI will balance active IRQs 96 default in APIC mode 97 98 acpi_irq_nobalance [HW,ACPI] 99 ACPI will not move active IRQs (default) 100 default in PIC mode 101 102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 103 Format: <irq>,<irq>... 104 105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 106 use by PCI 107 Format: <irq>,<irq>... 108 109 acpi_mask_gpe= [HW,ACPI] 110 Due to the existence of _Lxx/_Exx, some GPEs triggered 111 by unsupported hardware/firmware features can result in 112 GPE floodings that cannot be automatically disabled by 113 the GPE dispatcher. 114 This facility can be used to prevent such uncontrolled 115 GPE floodings. 116 Format: <byte> or <bitmap-list> 117 118 acpi_no_auto_serialize [HW,ACPI] 119 Disable auto-serialization of AML methods 120 AML control methods that contain the opcodes to create 121 named objects will be marked as "Serialized" by the 122 auto-serialization feature. 123 This feature is enabled by default. 124 This option allows to turn off the feature. 125 126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 127 kernels. 128 129 acpi_no_static_ssdt [HW,ACPI] 130 Disable installation of static SSDTs at early boot time 131 By default, SSDTs contained in the RSDT/XSDT will be 132 installed automatically and they will appear under 133 /sys/firmware/acpi/tables. 134 This option turns off this feature. 135 Note that specifying this option does not affect 136 dynamic table installation which will install SSDT 137 tables to /sys/firmware/acpi/tables/dynamic. 138 139 acpi_no_watchdog [HW,ACPI,WDT] 140 Ignore the ACPI-based watchdog interface (WDAT) and let 141 a native driver control the watchdog device instead. 142 143 acpi_rsdp= [ACPI,EFI,KEXEC] 144 Pass the RSDP address to the kernel, mostly used 145 on machines running EFI runtime service to boot the 146 second kernel for kdump. 147 148 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 149 Format: To spoof as Windows 98: ="Microsoft Windows" 150 151 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 152 of 2 which is mandated by ACPI 6) as the supported ACPI 153 specification revision (when using this switch, it may 154 be necessary to carry out a cold reboot _twice_ in a 155 row to make it take effect on the platform firmware). 156 157 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 158 acpi_osi="string1" # add string1 159 acpi_osi="!string2" # remove string2 160 acpi_osi=!* # remove all strings 161 acpi_osi=! # disable all built-in OS vendor 162 strings 163 acpi_osi=!! # enable all built-in OS vendor 164 strings 165 acpi_osi= # disable all strings 166 167 'acpi_osi=!' can be used in combination with single or 168 multiple 'acpi_osi="string1"' to support specific OS 169 vendor string(s). Note that such command can only 170 affect the default state of the OS vendor strings, thus 171 it cannot affect the default state of the feature group 172 strings and the current state of the OS vendor strings, 173 specifying it multiple times through kernel command line 174 is meaningless. This command is useful when one do not 175 care about the state of the feature group strings which 176 should be controlled by the OSPM. 177 Examples: 178 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 179 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 180 can make '_OSI("Windows 2000")' TRUE. 181 182 'acpi_osi=' cannot be used in combination with other 183 'acpi_osi=' command lines, the _OSI method will not 184 exist in the ACPI namespace. NOTE that such command can 185 only affect the _OSI support state, thus specifying it 186 multiple times through kernel command line is also 187 meaningless. 188 Examples: 189 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 190 FALSE. 191 192 'acpi_osi=!*' can be used in combination with single or 193 multiple 'acpi_osi="string1"' to support specific 194 string(s). Note that such command can affect the 195 current state of both the OS vendor strings and the 196 feature group strings, thus specifying it multiple times 197 through kernel command line is meaningful. But it may 198 still not able to affect the final state of a string if 199 there are quirks related to this string. This command 200 is useful when one want to control the state of the 201 feature group strings to debug BIOS issues related to 202 the OSPM features. 203 Examples: 204 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 205 '_OSI("Module Device")' FALSE. 206 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 207 '_OSI("Module Device")' TRUE. 208 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 209 equivalent to 210 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 211 and 212 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 213 they all will make '_OSI("Windows 2000")' TRUE. 214 215 acpi_pm_good [X86] 216 Override the pmtimer bug detection: force the kernel 217 to assume that this machine's pmtimer latches its value 218 and always returns good values. 219 220 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 221 Format: { level | edge | high | low } 222 223 acpi_skip_timer_override [HW,ACPI] 224 Recognize and ignore IRQ0/pin2 Interrupt Override. 225 For broken nForce2 BIOS resulting in XT-PIC timer. 226 227 acpi_sleep= [HW,ACPI] Sleep options 228 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 229 s4_nohwsig, old_ordering, nonvs, 230 sci_force_enable, nobl } 231 See Documentation/power/video.rst for information on 232 s3_bios and s3_mode. 233 s3_beep is for debugging; it makes the PC's speaker beep 234 as soon as the kernel's real-mode entry point is called. 235 s4_hwsig causes the kernel to check the ACPI hardware 236 signature during resume from hibernation, and gracefully 237 refuse to resume if it has changed. This complies with 238 the ACPI specification but not with reality, since 239 Windows does not do this and many laptops do change it 240 on docking. So the default behaviour is to allow resume 241 and simply warn when the signature changes, unless the 242 s4_hwsig option is enabled. 243 s4_nohwsig prevents ACPI hardware signature from being 244 used (or even warned about) during resume. 245 old_ordering causes the ACPI 1.0 ordering of the _PTS 246 control method, with respect to putting devices into 247 low power states, to be enforced (the ACPI 2.0 ordering 248 of _PTS is used by default). 249 nonvs prevents the kernel from saving/restoring the 250 ACPI NVS memory during suspend/hibernation and resume. 251 sci_force_enable causes the kernel to set SCI_EN directly 252 on resume from S1/S3 (which is against the ACPI spec, 253 but some broken systems don't work without it). 254 nobl causes the internal blacklist of systems known to 255 behave incorrectly in some ways with respect to system 256 suspend and resume to be ignored (use wisely). 257 258 acpi_use_timer_override [HW,ACPI] 259 Use timer override. For some broken Nvidia NF5 boards 260 that require a timer override, but don't have HPET 261 262 add_efi_memmap [EFI; X86] Include EFI memory map in 263 kernel's map of available physical RAM. 264 265 agp= [AGP] 266 { off | try_unsupported } 267 off: disable AGP support 268 try_unsupported: try to drive unsupported chipsets 269 (may crash computer or cause data corruption) 270 271 ALSA [HW,ALSA] 272 See Documentation/sound/alsa-configuration.rst 273 274 alignment= [KNL,ARM] 275 Allow the default userspace alignment fault handler 276 behaviour to be specified. Bit 0 enables warnings, 277 bit 1 enables fixups, and bit 2 sends a segfault. 278 279 align_va_addr= [X86-64] 280 Align virtual addresses by clearing slice [14:12] when 281 allocating a VMA at process creation time. This option 282 gives you up to 3% performance improvement on AMD F15h 283 machines (where it is enabled by default) for a 284 CPU-intensive style benchmark, and it can vary highly in 285 a microbenchmark depending on workload and compiler. 286 287 32: only for 32-bit processes 288 64: only for 64-bit processes 289 on: enable for both 32- and 64-bit processes 290 off: disable for both 32- and 64-bit processes 291 292 alloc_snapshot [FTRACE] 293 Allocate the ftrace snapshot buffer on boot up when the 294 main buffer is allocated. This is handy if debugging 295 and you need to use tracing_snapshot() on boot up, and 296 do not want to use tracing_snapshot_alloc() as it needs 297 to be done where GFP_KERNEL allocations are allowed. 298 299 allow_mismatched_32bit_el0 [ARM64] 300 Allow execve() of 32-bit applications and setting of the 301 PER_LINUX32 personality on systems where only a strict 302 subset of the CPUs support 32-bit EL0. When this 303 parameter is present, the set of CPUs supporting 32-bit 304 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 305 and hot-unplug operations may be restricted. 306 307 See Documentation/arch/arm64/asymmetric-32bit.rst for more 308 information. 309 310 amd_iommu= [HW,X86-64] 311 Pass parameters to the AMD IOMMU driver in the system. 312 Possible values are: 313 fullflush - Deprecated, equivalent to iommu.strict=1 314 off - do not initialize any AMD IOMMU found in 315 the system 316 force_isolation - Force device isolation for all 317 devices. The IOMMU driver is not 318 allowed anymore to lift isolation 319 requirements as needed. This option 320 does not override iommu=pt 321 force_enable - Force enable the IOMMU on platforms known 322 to be buggy with IOMMU enabled. Use this 323 option with care. 324 pgtbl_v1 - Use v1 page table for DMA-API (Default). 325 pgtbl_v2 - Use v2 page table for DMA-API. 326 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 327 328 amd_iommu_dump= [HW,X86-64] 329 Enable AMD IOMMU driver option to dump the ACPI table 330 for AMD IOMMU. With this option enabled, AMD IOMMU 331 driver will print ACPI tables for AMD IOMMU during 332 IOMMU initialization. 333 334 amd_iommu_intr= [HW,X86-64] 335 Specifies one of the following AMD IOMMU interrupt 336 remapping modes: 337 legacy - Use legacy interrupt remapping mode. 338 vapic - Use virtual APIC mode, which allows IOMMU 339 to inject interrupts directly into guest. 340 This mode requires kvm-amd.avic=1. 341 (Default when IOMMU HW support is present.) 342 343 amd_pstate= [X86] 344 disable 345 Do not enable amd_pstate as the default 346 scaling driver for the supported processors 347 passive 348 Use amd_pstate with passive mode as a scaling driver. 349 In this mode autonomous selection is disabled. 350 Driver requests a desired performance level and platform 351 tries to match the same performance level if it is 352 satisfied by guaranteed performance level. 353 active 354 Use amd_pstate_epp driver instance as the scaling driver, 355 driver provides a hint to the hardware if software wants 356 to bias toward performance (0x0) or energy efficiency (0xff) 357 to the CPPC firmware. then CPPC power algorithm will 358 calculate the runtime workload and adjust the realtime cores 359 frequency. 360 guided 361 Activate guided autonomous mode. Driver requests minimum and 362 maximum performance level and the platform autonomously 363 selects a performance level in this range and appropriate 364 to the current workload. 365 366 amijoy.map= [HW,JOY] Amiga joystick support 367 Map of devices attached to JOY0DAT and JOY1DAT 368 Format: <a>,<b> 369 See also Documentation/input/joydev/joystick.rst 370 371 analog.map= [HW,JOY] Analog joystick and gamepad support 372 Specifies type or capabilities of an analog joystick 373 connected to one of 16 gameports 374 Format: <type1>,<type2>,..<type16> 375 376 apc= [HW,SPARC] 377 Power management functions (SPARCstation-4/5 + deriv.) 378 Format: noidle 379 Disable APC CPU standby support. SPARCstation-Fox does 380 not play well with APC CPU idle - disable it if you have 381 APC and your system crashes randomly. 382 383 apic= [APIC,X86] Advanced Programmable Interrupt Controller 384 Change the output verbosity while booting 385 Format: { quiet (default) | verbose | debug } 386 Change the amount of debugging information output 387 when initialising the APIC and IO-APIC components. 388 For X86-32, this can also be used to specify an APIC 389 driver name. 390 Format: apic=driver_name 391 Examples: apic=bigsmp 392 393 apic_extnmi= [APIC,X86] External NMI delivery setting 394 Format: { bsp (default) | all | none } 395 bsp: External NMI is delivered only to CPU 0 396 all: External NMIs are broadcast to all CPUs as a 397 backup of CPU 0 398 none: External NMI is masked for all CPUs. This is 399 useful so that a dump capture kernel won't be 400 shot down by NMI 401 402 autoconf= [IPV6] 403 See Documentation/networking/ipv6.rst. 404 405 apm= [APM] Advanced Power Management 406 See header of arch/x86/kernel/apm_32.c. 407 408 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 409 Format: { "0" | "1" } 410 See security/apparmor/Kconfig help text 411 0 -- disable. 412 1 -- enable. 413 Default value is set via kernel config option. 414 415 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 416 Format: <io>,<irq>,<nodeID> 417 418 arm64.nobti [ARM64] Unconditionally disable Branch Target 419 Identification support 420 421 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 422 Set instructions support 423 424 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 425 support 426 427 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 428 support 429 430 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 431 Extension support 432 433 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 434 Extension support 435 436 ataflop= [HW,M68k] 437 438 atarimouse= [HW,MOUSE] Atari Mouse 439 440 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 441 EzKey and similar keyboards 442 443 atkbd.reset= [HW] Reset keyboard during initialization 444 445 atkbd.set= [HW] Select keyboard code set 446 Format: <int> (2 = AT (default), 3 = PS/2) 447 448 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 449 keyboards 450 451 atkbd.softraw= [HW] Choose between synthetic and real raw mode 452 Format: <bool> (0 = real, 1 = synthetic (default)) 453 454 atkbd.softrepeat= [HW] 455 Use software keyboard repeat 456 457 audit= [KNL] Enable the audit sub-system 458 Format: { "0" | "1" | "off" | "on" } 459 0 | off - kernel audit is disabled and can not be 460 enabled until the next reboot 461 unset - kernel audit is initialized but disabled and 462 will be fully enabled by the userspace auditd. 463 1 | on - kernel audit is initialized and partially 464 enabled, storing at most audit_backlog_limit 465 messages in RAM until it is fully enabled by the 466 userspace auditd. 467 Default: unset 468 469 audit_backlog_limit= [KNL] Set the audit queue size limit. 470 Format: <int> (must be >=0) 471 Default: 64 472 473 bau= [X86_UV] Enable the BAU on SGI UV. The default 474 behavior is to disable the BAU (i.e. bau=0). 475 Format: { "0" | "1" } 476 0 - Disable the BAU. 477 1 - Enable the BAU. 478 unset - Disable the BAU. 479 480 baycom_epp= [HW,AX25] 481 Format: <io>,<mode> 482 483 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 484 Format: <io>,<mode> 485 See header of drivers/net/hamradio/baycom_par.c. 486 487 baycom_ser_fdx= [HW,AX25] 488 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 489 Format: <io>,<irq>,<mode>[,<baud>] 490 See header of drivers/net/hamradio/baycom_ser_fdx.c. 491 492 baycom_ser_hdx= [HW,AX25] 493 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 494 Format: <io>,<irq>,<mode> 495 See header of drivers/net/hamradio/baycom_ser_hdx.c. 496 497 bert_disable [ACPI] 498 Disable BERT OS support on buggy BIOSes. 499 500 bgrt_disable [ACPI][X86] 501 Disable BGRT to avoid flickering OEM logo. 502 503 blkdevparts= Manual partition parsing of block device(s) for 504 embedded devices based on command line input. 505 See Documentation/block/cmdline-partition.rst 506 507 boot_delay= Milliseconds to delay each printk during boot. 508 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 509 and you may also have to specify "lpj=". Boot_delay 510 values larger than 10 seconds (10000) are assumed 511 erroneous and ignored. 512 Format: integer 513 514 bootconfig [KNL] 515 Extended command line options can be added to an initrd 516 and this will cause the kernel to look for it. 517 518 See Documentation/admin-guide/bootconfig.rst 519 520 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 521 bttv.radio= Most important insmod options are available as 522 kernel args too. 523 bttv.pll= See Documentation/admin-guide/media/bttv.rst 524 bttv.tuner= 525 526 bulk_remove=off [PPC] This parameter disables the use of the pSeries 527 firmware feature for flushing multiple hpte entries 528 at a time. 529 530 c101= [NET] Moxa C101 synchronous serial card 531 532 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 533 Sometimes CPU hardware bugs make them report the cache 534 size incorrectly. The kernel will attempt work arounds 535 to fix known problems, but for some CPUs it is not 536 possible to determine what the correct size should be. 537 This option provides an override for these situations. 538 539 carrier_timeout= 540 [NET] Specifies amount of time (in seconds) that 541 the kernel should wait for a network carrier. By default 542 it waits 120 seconds. 543 544 ca_keys= [KEYS] This parameter identifies a specific key(s) on 545 the system trusted keyring to be used for certificate 546 trust validation. 547 format: { id:<keyid> | builtin } 548 549 cca= [MIPS] Override the kernel pages' cache coherency 550 algorithm. Accepted values range from 0 to 7 551 inclusive. See arch/mips/include/asm/pgtable-bits.h 552 for platform specific values (SB1, Loongson3 and 553 others). 554 555 ccw_timeout_log [S390] 556 See Documentation/arch/s390/common_io.rst for details. 557 558 cgroup_disable= [KNL] Disable a particular controller or optional feature 559 Format: {name of the controller(s) or feature(s) to disable} 560 The effects of cgroup_disable=foo are: 561 - foo isn't auto-mounted if you mount all cgroups in 562 a single hierarchy 563 - foo isn't visible as an individually mountable 564 subsystem 565 - if foo is an optional feature then the feature is 566 disabled and corresponding cgroup files are not 567 created 568 {Currently only "memory" controller deal with this and 569 cut the overhead, others just disable the usage. So 570 only cgroup_disable=memory is actually worthy} 571 Specifying "pressure" disables per-cgroup pressure 572 stall information accounting feature 573 574 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 575 Format: { { controller | "all" | "named" } 576 [,{ controller | "all" | "named" }...] } 577 Like cgroup_disable, but only applies to cgroup v1; 578 the blacklisted controllers remain available in cgroup2. 579 "all" blacklists all controllers and "named" disables 580 named mounts. Specifying both "all" and "named" disables 581 all v1 hierarchies. 582 583 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 584 Format: <string> 585 nosocket -- Disable socket memory accounting. 586 nokmem -- Disable kernel memory accounting. 587 nobpf -- Disable BPF memory accounting. 588 589 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 590 Format: { "0" | "1" } 591 See security/selinux/Kconfig help text. 592 0 -- check protection applied by kernel (includes 593 any implied execute protection). 594 1 -- check protection requested by application. 595 Default value is set via a kernel config option. 596 Value can be changed at runtime via 597 /sys/fs/selinux/checkreqprot. 598 Setting checkreqprot to 1 is deprecated. 599 600 cio_ignore= [S390] 601 See Documentation/arch/s390/common_io.rst for details. 602 603 clearcpuid=X[,X...] [X86] 604 Disable CPUID feature X for the kernel. See 605 arch/x86/include/asm/cpufeatures.h for the valid bit 606 numbers X. Note the Linux-specific bits are not necessarily 607 stable over kernel options, but the vendor-specific 608 ones should be. 609 X can also be a string as appearing in the flags: line 610 in /proc/cpuinfo which does not have the above 611 instability issue. However, not all features have names 612 in /proc/cpuinfo. 613 Note that using this option will taint your kernel. 614 Also note that user programs calling CPUID directly 615 or using the feature without checking anything 616 will still see it. This just prevents it from 617 being used by the kernel or shown in /proc/cpuinfo. 618 Also note the kernel might malfunction if you disable 619 some critical bits. 620 621 clk_ignore_unused 622 [CLK] 623 Prevents the clock framework from automatically gating 624 clocks that have not been explicitly enabled by a Linux 625 device driver but are enabled in hardware at reset or 626 by the bootloader/firmware. Note that this does not 627 force such clocks to be always-on nor does it reserve 628 those clocks in any way. This parameter is useful for 629 debug and development, but should not be needed on a 630 platform with proper driver support. For more 631 information, see Documentation/driver-api/clk.rst. 632 633 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 634 [Deprecated] 635 Forces specified clocksource (if available) to be used 636 when calculating gettimeofday(). If specified 637 clocksource is not available, it defaults to PIT. 638 Format: { pit | tsc | cyclone | pmtmr } 639 640 clocksource= Override the default clocksource 641 Format: <string> 642 Override the default clocksource and use the clocksource 643 with the name specified. 644 Some clocksource names to choose from, depending on 645 the platform: 646 [all] jiffies (this is the base, fallback clocksource) 647 [ACPI] acpi_pm 648 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 649 pxa_timer,timer3,32k_counter,timer0_1 650 [X86-32] pit,hpet,tsc; 651 scx200_hrt on Geode; cyclone on IBM x440 652 [MIPS] MIPS 653 [PARISC] cr16 654 [S390] tod 655 [SH] SuperH 656 [SPARC64] tick 657 [X86-64] hpet,tsc 658 659 clocksource.arm_arch_timer.evtstrm= 660 [ARM,ARM64] 661 Format: <bool> 662 Enable/disable the eventstream feature of the ARM 663 architected timer so that code using WFE-based polling 664 loops can be debugged more effectively on production 665 systems. 666 667 clocksource.max_cswd_read_retries= [KNL] 668 Number of clocksource_watchdog() retries due to 669 external delays before the clock will be marked 670 unstable. Defaults to two retries, that is, 671 three attempts to read the clock under test. 672 673 clocksource.verify_n_cpus= [KNL] 674 Limit the number of CPUs checked for clocksources 675 marked with CLOCK_SOURCE_VERIFY_PERCPU that 676 are marked unstable due to excessive skew. 677 A negative value says to check all CPUs, while 678 zero says not to check any. Values larger than 679 nr_cpu_ids are silently truncated to nr_cpu_ids. 680 The actual CPUs are chosen randomly, with 681 no replacement if the same CPU is chosen twice. 682 683 clocksource-wdtest.holdoff= [KNL] 684 Set the time in seconds that the clocksource 685 watchdog test waits before commencing its tests. 686 Defaults to zero when built as a module and to 687 10 seconds when built into the kernel. 688 689 cma=nn[MG]@[start[MG][-end[MG]]] 690 [KNL,CMA] 691 Sets the size of kernel global memory area for 692 contiguous memory allocations and optionally the 693 placement constraint by the physical address range of 694 memory allocations. A value of 0 disables CMA 695 altogether. For more information, see 696 kernel/dma/contiguous.c 697 698 cma_pernuma=nn[MG] 699 [KNL,CMA] 700 Sets the size of kernel per-numa memory area for 701 contiguous memory allocations. A value of 0 disables 702 per-numa CMA altogether. And If this option is not 703 specified, the default value is 0. 704 With per-numa CMA enabled, DMA users on node nid will 705 first try to allocate buffer from the pernuma area 706 which is located in node nid, if the allocation fails, 707 they will fallback to the global default memory area. 708 709 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 710 [KNL,CMA] 711 Sets the size of kernel numa memory area for 712 contiguous memory allocations. It will reserve CMA 713 area for the specified node. 714 715 With numa CMA enabled, DMA users on node nid will 716 first try to allocate buffer from the numa area 717 which is located in node nid, if the allocation fails, 718 they will fallback to the global default memory area. 719 720 cmo_free_hint= [PPC] Format: { yes | no } 721 Specify whether pages are marked as being inactive 722 when they are freed. This is used in CMO environments 723 to determine OS memory pressure for page stealing by 724 a hypervisor. 725 Default: yes 726 727 coherent_pool=nn[KMG] [ARM,KNL] 728 Sets the size of memory pool for coherent, atomic dma 729 allocations, by default set to 256K. 730 731 com20020= [HW,NET] ARCnet - COM20020 chipset 732 Format: 733 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 734 735 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 736 Format: <io>[,<irq>] 737 738 com90xx= [HW,NET] 739 ARCnet - COM90xx chipset (memory-mapped buffers) 740 Format: <io>[,<irq>[,<memstart>]] 741 742 condev= [HW,S390] console device 743 conmode= 744 745 con3215_drop= [S390] 3215 console drop mode. 746 Format: y|n|Y|N|1|0 747 When set to true, drop data on the 3215 console when 748 the console buffer is full. In this case the 749 operator using a 3270 terminal emulator (for example 750 x3270) does not have to enter the clear key for the 751 console output to advance and the kernel to continue. 752 This leads to a much faster boot time when a 3270 753 terminal emulator is active. If no 3270 terminal 754 emulator is used, this parameter has no effect. 755 756 console= [KNL] Output console device and options. 757 758 tty<n> Use the virtual console device <n>. 759 760 ttyS<n>[,options] 761 ttyUSB0[,options] 762 Use the specified serial port. The options are of 763 the form "bbbbpnf", where "bbbb" is the baud rate, 764 "p" is parity ("n", "o", or "e"), "n" is number of 765 bits, and "f" is flow control ("r" for RTS or 766 omit it). Default is "9600n8". 767 768 See Documentation/admin-guide/serial-console.rst for more 769 information. See 770 Documentation/networking/netconsole.rst for an 771 alternative. 772 773 uart[8250],io,<addr>[,options] 774 uart[8250],mmio,<addr>[,options] 775 uart[8250],mmio16,<addr>[,options] 776 uart[8250],mmio32,<addr>[,options] 777 uart[8250],0x<addr>[,options] 778 Start an early, polled-mode console on the 8250/16550 779 UART at the specified I/O port or MMIO address, 780 switching to the matching ttyS device later. 781 MMIO inter-register address stride is either 8-bit 782 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 783 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 784 to be equivalent to 'mmio'. 'options' are specified in 785 the same format described for ttyS above; if unspecified, 786 the h/w is not re-initialized. 787 788 hvc<n> Use the hypervisor console device <n>. This is for 789 both Xen and PowerPC hypervisors. 790 791 { null | "" } 792 Use to disable console output, i.e., to have kernel 793 console messages discarded. 794 This must be the only console= parameter used on the 795 kernel command line. 796 797 If the device connected to the port is not a TTY but a braille 798 device, prepend "brl," before the device type, for instance 799 console=brl,ttyS0 800 For now, only VisioBraille is supported. 801 802 console_msg_format= 803 [KNL] Change console messages format 804 default 805 By default we print messages on consoles in 806 "[time stamp] text\n" format (time stamp may not be 807 printed, depending on CONFIG_PRINTK_TIME or 808 `printk_time' param). 809 syslog 810 Switch to syslog format: "<%u>[time stamp] text\n" 811 IOW, each message will have a facility and loglevel 812 prefix. The format is similar to one used by syslog() 813 syscall, or to executing "dmesg -S --raw" or to reading 814 from /proc/kmsg. 815 816 consoleblank= [KNL] The console blank (screen saver) timeout in 817 seconds. A value of 0 disables the blank timer. 818 Defaults to 0. 819 820 coredump_filter= 821 [KNL] Change the default value for 822 /proc/<pid>/coredump_filter. 823 See also Documentation/filesystems/proc.rst. 824 825 coresight_cpu_debug.enable 826 [ARM,ARM64] 827 Format: <bool> 828 Enable/disable the CPU sampling based debugging. 829 0: default value, disable debugging 830 1: enable debugging at boot time 831 832 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 833 Format: 834 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 835 836 cpuidle.off=1 [CPU_IDLE] 837 disable the cpuidle sub-system 838 839 cpuidle.governor= 840 [CPU_IDLE] Name of the cpuidle governor to use. 841 842 cpufreq.off=1 [CPU_FREQ] 843 disable the cpufreq sub-system 844 845 cpufreq.default_governor= 846 [CPU_FREQ] Name of the default cpufreq governor or 847 policy to use. This governor must be registered in the 848 kernel before the cpufreq driver probes. 849 850 cpu_init_udelay=N 851 [X86] Delay for N microsec between assert and de-assert 852 of APIC INIT to start processors. This delay occurs 853 on every CPU online, such as boot, and resume from suspend. 854 Default: 10000 855 856 cpuhp.parallel= 857 [SMP] Enable/disable parallel bringup of secondary CPUs 858 Format: <bool> 859 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 860 the parameter has no effect. 861 862 crash_kexec_post_notifiers 863 Run kdump after running panic-notifiers and dumping 864 kmsg. This only for the users who doubt kdump always 865 succeeds in any situation. 866 Note that this also increases risks of kdump failure, 867 because some panic notifiers can make the crashed 868 kernel more unstable. 869 870 crashkernel=size[KMG][@offset[KMG]] 871 [KNL] Using kexec, Linux can switch to a 'crash kernel' 872 upon panic. This parameter reserves the physical 873 memory region [offset, offset + size] for that kernel 874 image. If '@offset' is omitted, then a suitable offset 875 is selected automatically. 876 [KNL, X86-64, ARM64, RISCV] Select a region under 4G first, and 877 fall back to reserve region above 4G when '@offset' 878 hasn't been specified. 879 See Documentation/admin-guide/kdump/kdump.rst for further details. 880 881 crashkernel=range1:size1[,range2:size2,...][@offset] 882 [KNL] Same as above, but depends on the memory 883 in the running system. The syntax of range is 884 start-[end] where start and end are both 885 a memory unit (amount[KMG]). See also 886 Documentation/admin-guide/kdump/kdump.rst for an example. 887 888 crashkernel=size[KMG],high 889 [KNL, X86-64, ARM64, RISCV] range could be above 4G. 890 Allow kernel to allocate physical memory region from top, 891 so could be above 4G if system have more than 4G ram 892 installed. Otherwise memory region will be allocated 893 below 4G, if available. 894 It will be ignored if crashkernel=X is specified. 895 crashkernel=size[KMG],low 896 [KNL, X86-64, ARM64, RISCV] range under 4G. When crashkernel=X,high 897 is passed, kernel could allocate physical memory region 898 above 4G, that cause second kernel crash on system 899 that require some amount of low memory, e.g. swiotlb 900 requires at least 64M+32K low memory, also enough extra 901 low memory is needed to make sure DMA buffers for 32-bit 902 devices won't run out. Kernel would try to allocate 903 default size of memory below 4G automatically. The default 904 size is platform dependent. 905 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 906 --> arm64: 128MiB 907 --> riscv: 128MiB 908 This one lets the user specify own low range under 4G 909 for second kernel instead. 910 0: to disable low allocation. 911 It will be ignored when crashkernel=X,high is not used 912 or memory reserved is below 4G. 913 914 cryptomgr.notests 915 [KNL] Disable crypto self-tests 916 917 cs89x0_dma= [HW,NET] 918 Format: <dma> 919 920 cs89x0_media= [HW,NET] 921 Format: { rj45 | aui | bnc } 922 923 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 924 function call handling. When switched on, 925 additional debug data is printed to the console 926 in case a hanging CPU is detected, and that 927 CPU is pinged again in order to try to resolve 928 the hang situation. The default value of this 929 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 930 Kconfig option. 931 932 dasd= [HW,NET] 933 See header of drivers/s390/block/dasd_devmap.c. 934 935 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 936 (one device per port) 937 Format: <port#>,<type> 938 See also Documentation/input/devices/joystick-parport.rst 939 940 debug [KNL] Enable kernel debugging (events log level). 941 942 debug_boot_weak_hash 943 [KNL] Enable printing [hashed] pointers early in the 944 boot sequence. If enabled, we use a weak hash instead 945 of siphash to hash pointers. Use this option if you are 946 seeing instances of '(___ptrval___)') and need to see a 947 value (hashed pointer) instead. Cryptographically 948 insecure, please do not use on production kernels. 949 950 debug_locks_verbose= 951 [KNL] verbose locking self-tests 952 Format: <int> 953 Print debugging info while doing the locking API 954 self-tests. 955 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 956 (no extra messages), setting it to -1 (all bits set) 957 will print _a_lot_ more information - normally only 958 useful to lockdep developers. 959 960 debug_objects [KNL] Enable object debugging 961 962 debug_guardpage_minorder= 963 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 964 parameter allows control of the order of pages that will 965 be intentionally kept free (and hence protected) by the 966 buddy allocator. Bigger value increase the probability 967 of catching random memory corruption, but reduce the 968 amount of memory for normal system use. The maximum 969 possible value is MAX_ORDER/2. Setting this parameter 970 to 1 or 2 should be enough to identify most random 971 memory corruption problems caused by bugs in kernel or 972 driver code when a CPU writes to (or reads from) a 973 random memory location. Note that there exists a class 974 of memory corruptions problems caused by buggy H/W or 975 F/W or by drivers badly programming DMA (basically when 976 memory is written at bus level and the CPU MMU is 977 bypassed) which are not detectable by 978 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 979 tracking down these problems. 980 981 debug_pagealloc= 982 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 983 enables the feature at boot time. By default, it is 984 disabled and the system will work mostly the same as a 985 kernel built without CONFIG_DEBUG_PAGEALLOC. 986 Note: to get most of debug_pagealloc error reports, it's 987 useful to also enable the page_owner functionality. 988 on: enable the feature 989 990 debugfs= [KNL] This parameter enables what is exposed to userspace 991 and debugfs internal clients. 992 Format: { on, no-mount, off } 993 on: All functions are enabled. 994 no-mount: 995 Filesystem is not registered but kernel clients can 996 access APIs and a crashkernel can be used to read 997 its content. There is nothing to mount. 998 off: Filesystem is not registered and clients 999 get a -EPERM as result when trying to register files 1000 or directories within debugfs. 1001 This is equivalent of the runtime functionality if 1002 debugfs was not enabled in the kernel at all. 1003 Default value is set in build-time with a kernel configuration. 1004 1005 debugpat [X86] Enable PAT debugging 1006 1007 default_hugepagesz= 1008 [HW] The size of the default HugeTLB page. This is 1009 the size represented by the legacy /proc/ hugepages 1010 APIs. In addition, this is the default hugetlb size 1011 used for shmget(), mmap() and mounting hugetlbfs 1012 filesystems. If not specified, defaults to the 1013 architecture's default huge page size. Huge page 1014 sizes are architecture dependent. See also 1015 Documentation/admin-guide/mm/hugetlbpage.rst. 1016 Format: size[KMG] 1017 1018 deferred_probe_timeout= 1019 [KNL] Debugging option to set a timeout in seconds for 1020 deferred probe to give up waiting on dependencies to 1021 probe. Only specific dependencies (subsystems or 1022 drivers) that have opted in will be ignored. A timeout 1023 of 0 will timeout at the end of initcalls. If the time 1024 out hasn't expired, it'll be restarted by each 1025 successful driver registration. This option will also 1026 dump out devices still on the deferred probe list after 1027 retrying. 1028 1029 delayacct [KNL] Enable per-task delay accounting 1030 1031 dell_smm_hwmon.ignore_dmi= 1032 [HW] Continue probing hardware even if DMI data 1033 indicates that the driver is running on unsupported 1034 hardware. 1035 1036 dell_smm_hwmon.force= 1037 [HW] Activate driver even if SMM BIOS signature does 1038 not match list of supported models and enable otherwise 1039 blacklisted features. 1040 1041 dell_smm_hwmon.power_status= 1042 [HW] Report power status in /proc/i8k 1043 (disabled by default). 1044 1045 dell_smm_hwmon.restricted= 1046 [HW] Allow controlling fans only if SYS_ADMIN 1047 capability is set. 1048 1049 dell_smm_hwmon.fan_mult= 1050 [HW] Factor to multiply fan speed with. 1051 1052 dell_smm_hwmon.fan_max= 1053 [HW] Maximum configurable fan speed. 1054 1055 dfltcc= [HW,S390] 1056 Format: { on | off | def_only | inf_only | always } 1057 on: s390 zlib hardware support for compression on 1058 level 1 and decompression (default) 1059 off: No s390 zlib hardware support 1060 def_only: s390 zlib hardware support for deflate 1061 only (compression on level 1) 1062 inf_only: s390 zlib hardware support for inflate 1063 only (decompression) 1064 always: Same as 'on' but ignores the selected compression 1065 level always using hardware support (used for debugging) 1066 1067 dhash_entries= [KNL] 1068 Set number of hash buckets for dentry cache. 1069 1070 disable_1tb_segments [PPC] 1071 Disables the use of 1TB hash page table segments. This 1072 causes the kernel to fall back to 256MB segments which 1073 can be useful when debugging issues that require an SLB 1074 miss to occur. 1075 1076 disable= [IPV6] 1077 See Documentation/networking/ipv6.rst. 1078 1079 disable_radix [PPC] 1080 Disable RADIX MMU mode on POWER9 1081 1082 disable_tlbie [PPC] 1083 Disable TLBIE instruction. Currently does not work 1084 with KVM, with HASH MMU, or with coherent accelerators. 1085 1086 disable_cpu_apicid= [X86,APIC,SMP] 1087 Format: <int> 1088 The number of initial APIC ID for the 1089 corresponding CPU to be disabled at boot, 1090 mostly used for the kdump 2nd kernel to 1091 disable BSP to wake up multiple CPUs without 1092 causing system reset or hang due to sending 1093 INIT from AP to BSP. 1094 1095 disable_ddw [PPC/PSERIES] 1096 Disable Dynamic DMA Window support. Use this 1097 to workaround buggy firmware. 1098 1099 disable_ipv6= [IPV6] 1100 See Documentation/networking/ipv6.rst. 1101 1102 disable_mtrr_cleanup [X86] 1103 The kernel tries to adjust MTRR layout from continuous 1104 to discrete, to make X server driver able to add WB 1105 entry later. This parameter disables that. 1106 1107 disable_mtrr_trim [X86, Intel and AMD only] 1108 By default the kernel will trim any uncacheable 1109 memory out of your available memory pool based on 1110 MTRR settings. This parameter disables that behavior, 1111 possibly causing your machine to run very slowly. 1112 1113 disable_timer_pin_1 [X86] 1114 Disable PIN 1 of APIC timer 1115 Can be useful to work around chipset bugs. 1116 1117 dis_ucode_ldr [X86] Disable the microcode loader. 1118 1119 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1120 this option disables the debugging code at boot. 1121 1122 dma_debug_entries=<number> 1123 This option allows to tune the number of preallocated 1124 entries for DMA-API debugging code. One entry is 1125 required per DMA-API allocation. Use this if the 1126 DMA-API debugging code disables itself because the 1127 architectural default is too low. 1128 1129 dma_debug_driver=<driver_name> 1130 With this option the DMA-API debugging driver 1131 filter feature can be enabled at boot time. Just 1132 pass the driver to filter for as the parameter. 1133 The filter can be disabled or changed to another 1134 driver later using sysfs. 1135 1136 driver_async_probe= [KNL] 1137 List of driver names to be probed asynchronously. * 1138 matches with all driver names. If * is specified, the 1139 rest of the listed driver names are those that will NOT 1140 match the *. 1141 Format: <driver_name1>,<driver_name2>... 1142 1143 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1144 Broken monitors, graphic adapters, KVMs and EDIDless 1145 panels may send no or incorrect EDID data sets. 1146 This parameter allows to specify an EDID data sets 1147 in the /lib/firmware directory that are used instead. 1148 Generic built-in EDID data sets are used, if one of 1149 edid/1024x768.bin, edid/1280x1024.bin, 1150 edid/1680x1050.bin, or edid/1920x1080.bin is given 1151 and no file with the same name exists. Details and 1152 instructions how to build your own EDID data are 1153 available in Documentation/admin-guide/edid.rst. An EDID 1154 data set will only be used for a particular connector, 1155 if its name and a colon are prepended to the EDID 1156 name. Each connector may use a unique EDID data 1157 set by separating the files with a comma. An EDID 1158 data set with no connector name will be used for 1159 any connectors not explicitly specified. 1160 1161 dscc4.setup= [NET] 1162 1163 dt_cpu_ftrs= [PPC] 1164 Format: {"off" | "known"} 1165 Control how the dt_cpu_ftrs device-tree binding is 1166 used for CPU feature discovery and setup (if it 1167 exists). 1168 off: Do not use it, fall back to legacy cpu table. 1169 known: Do not pass through unknown features to guests 1170 or userspace, only those that the kernel is aware of. 1171 1172 dump_apple_properties [X86] 1173 Dump name and content of EFI device properties on 1174 x86 Macs. Useful for driver authors to determine 1175 what data is available or for reverse-engineering. 1176 1177 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1178 <module>.dyndbg[="val"] 1179 Enable debug messages at boot time. See 1180 Documentation/admin-guide/dynamic-debug-howto.rst 1181 for details. 1182 1183 early_ioremap_debug [KNL] 1184 Enable debug messages in early_ioremap support. This 1185 is useful for tracking down temporary early mappings 1186 which are not unmapped. 1187 1188 earlycon= [KNL] Output early console device and options. 1189 1190 When used with no options, the early console is 1191 determined by stdout-path property in device tree's 1192 chosen node or the ACPI SPCR table if supported by 1193 the platform. 1194 1195 cdns,<addr>[,options] 1196 Start an early, polled-mode console on a Cadence 1197 (xuartps) serial port at the specified address. Only 1198 supported option is baud rate. If baud rate is not 1199 specified, the serial port must already be setup and 1200 configured. 1201 1202 uart[8250],io,<addr>[,options[,uartclk]] 1203 uart[8250],mmio,<addr>[,options[,uartclk]] 1204 uart[8250],mmio32,<addr>[,options[,uartclk]] 1205 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1206 uart[8250],0x<addr>[,options] 1207 Start an early, polled-mode console on the 8250/16550 1208 UART at the specified I/O port or MMIO address. 1209 MMIO inter-register address stride is either 8-bit 1210 (mmio) or 32-bit (mmio32 or mmio32be). 1211 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1212 to be equivalent to 'mmio'. 'options' are specified 1213 in the same format described for "console=ttyS<n>"; if 1214 unspecified, the h/w is not initialized. 'uartclk' is 1215 the uart clock frequency; if unspecified, it is set 1216 to 'BASE_BAUD' * 16. 1217 1218 pl011,<addr> 1219 pl011,mmio32,<addr> 1220 Start an early, polled-mode console on a pl011 serial 1221 port at the specified address. The pl011 serial port 1222 must already be setup and configured. Options are not 1223 yet supported. If 'mmio32' is specified, then only 1224 the driver will use only 32-bit accessors to read/write 1225 the device registers. 1226 1227 liteuart,<addr> 1228 Start an early console on a litex serial port at the 1229 specified address. The serial port must already be 1230 setup and configured. Options are not yet supported. 1231 1232 meson,<addr> 1233 Start an early, polled-mode console on a meson serial 1234 port at the specified address. The serial port must 1235 already be setup and configured. Options are not yet 1236 supported. 1237 1238 msm_serial,<addr> 1239 Start an early, polled-mode console on an msm serial 1240 port at the specified address. The serial port 1241 must already be setup and configured. Options are not 1242 yet supported. 1243 1244 msm_serial_dm,<addr> 1245 Start an early, polled-mode console on an msm serial 1246 dm port at the specified address. The serial port 1247 must already be setup and configured. Options are not 1248 yet supported. 1249 1250 owl,<addr> 1251 Start an early, polled-mode console on a serial port 1252 of an Actions Semi SoC, such as S500 or S900, at the 1253 specified address. The serial port must already be 1254 setup and configured. Options are not yet supported. 1255 1256 rda,<addr> 1257 Start an early, polled-mode console on a serial port 1258 of an RDA Micro SoC, such as RDA8810PL, at the 1259 specified address. The serial port must already be 1260 setup and configured. Options are not yet supported. 1261 1262 sbi 1263 Use RISC-V SBI (Supervisor Binary Interface) for early 1264 console. 1265 1266 smh Use ARM semihosting calls for early console. 1267 1268 s3c2410,<addr> 1269 s3c2412,<addr> 1270 s3c2440,<addr> 1271 s3c6400,<addr> 1272 s5pv210,<addr> 1273 exynos4210,<addr> 1274 Use early console provided by serial driver available 1275 on Samsung SoCs, requires selecting proper type and 1276 a correct base address of the selected UART port. The 1277 serial port must already be setup and configured. 1278 Options are not yet supported. 1279 1280 lantiq,<addr> 1281 Start an early, polled-mode console on a lantiq serial 1282 (lqasc) port at the specified address. The serial port 1283 must already be setup and configured. Options are not 1284 yet supported. 1285 1286 lpuart,<addr> 1287 lpuart32,<addr> 1288 Use early console provided by Freescale LP UART driver 1289 found on Freescale Vybrid and QorIQ LS1021A processors. 1290 A valid base address must be provided, and the serial 1291 port must already be setup and configured. 1292 1293 ec_imx21,<addr> 1294 ec_imx6q,<addr> 1295 Start an early, polled-mode, output-only console on the 1296 Freescale i.MX UART at the specified address. The UART 1297 must already be setup and configured. 1298 1299 ar3700_uart,<addr> 1300 Start an early, polled-mode console on the 1301 Armada 3700 serial port at the specified 1302 address. The serial port must already be setup 1303 and configured. Options are not yet supported. 1304 1305 qcom_geni,<addr> 1306 Start an early, polled-mode console on a Qualcomm 1307 Generic Interface (GENI) based serial port at the 1308 specified address. The serial port must already be 1309 setup and configured. Options are not yet supported. 1310 1311 efifb,[options] 1312 Start an early, unaccelerated console on the EFI 1313 memory mapped framebuffer (if available). On cache 1314 coherent non-x86 systems that use system memory for 1315 the framebuffer, pass the 'ram' option so that it is 1316 mapped with the correct attributes. 1317 1318 linflex,<addr> 1319 Use early console provided by Freescale LINFlexD UART 1320 serial driver for NXP S32V234 SoCs. A valid base 1321 address must be provided, and the serial port must 1322 already be setup and configured. 1323 1324 earlyprintk= [X86,SH,ARM,M68k,S390] 1325 earlyprintk=vga 1326 earlyprintk=sclp 1327 earlyprintk=xen 1328 earlyprintk=serial[,ttySn[,baudrate]] 1329 earlyprintk=serial[,0x...[,baudrate]] 1330 earlyprintk=ttySn[,baudrate] 1331 earlyprintk=dbgp[debugController#] 1332 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1333 earlyprintk=xdbc[xhciController#] 1334 1335 earlyprintk is useful when the kernel crashes before 1336 the normal console is initialized. It is not enabled by 1337 default because it has some cosmetic problems. 1338 1339 Append ",keep" to not disable it when the real console 1340 takes over. 1341 1342 Only one of vga, serial, or usb debug port can 1343 be used at a time. 1344 1345 Currently only ttyS0 and ttyS1 may be specified by 1346 name. Other I/O ports may be explicitly specified 1347 on some architectures (x86 and arm at least) by 1348 replacing ttySn with an I/O port address, like this: 1349 earlyprintk=serial,0x1008,115200 1350 You can find the port for a given device in 1351 /proc/tty/driver/serial: 1352 2: uart:ST16650V2 port:00001008 irq:18 ... 1353 1354 Interaction with the standard serial driver is not 1355 very good. 1356 1357 The VGA output is eventually overwritten by 1358 the real console. 1359 1360 The xen option can only be used in Xen domains. 1361 1362 The sclp output can only be used on s390. 1363 1364 The optional "force" to "pciserial" enables use of a 1365 PCI device even when its classcode is not of the 1366 UART class. 1367 1368 edac_report= [HW,EDAC] Control how to report EDAC event 1369 Format: {"on" | "off" | "force"} 1370 on: enable EDAC to report H/W event. May be overridden 1371 by other higher priority error reporting module. 1372 off: disable H/W event reporting through EDAC. 1373 force: enforce the use of EDAC to report H/W event. 1374 default: on. 1375 1376 edd= [EDD] 1377 Format: {"off" | "on" | "skip[mbr]"} 1378 1379 efi= [EFI] 1380 Format: { "debug", "disable_early_pci_dma", 1381 "nochunk", "noruntime", "nosoftreserve", 1382 "novamap", "no_disable_early_pci_dma" } 1383 debug: enable misc debug output. 1384 disable_early_pci_dma: disable the busmaster bit on all 1385 PCI bridges while in the EFI boot stub. 1386 nochunk: disable reading files in "chunks" in the EFI 1387 boot stub, as chunking can cause problems with some 1388 firmware implementations. 1389 noruntime : disable EFI runtime services support 1390 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1391 attribute may cause the kernel to reserve the 1392 memory range for a memory mapping driver to 1393 claim. Specify efi=nosoftreserve to disable this 1394 reservation and treat the memory by its base type 1395 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1396 novamap: do not call SetVirtualAddressMap(). 1397 no_disable_early_pci_dma: Leave the busmaster bit set 1398 on all PCI bridges while in the EFI boot stub 1399 1400 efi_no_storage_paranoia [EFI; X86] 1401 Using this parameter you can use more than 50% of 1402 your efi variable storage. Use this parameter only if 1403 you are really sure that your UEFI does sane gc and 1404 fulfills the spec otherwise your board may brick. 1405 1406 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1407 Add arbitrary attribute to specific memory range by 1408 updating original EFI memory map. 1409 Region of memory which aa attribute is added to is 1410 from ss to ss+nn. 1411 1412 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1413 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1414 attribute is added to range 0x100000000-0x180000000 and 1415 0x10a0000000-0x1120000000. 1416 1417 If efi_fake_mem=8G@9G:0x40000 is specified, the 1418 EFI_MEMORY_SP(0x40000) attribute is added to 1419 range 0x240000000-0x43fffffff. 1420 1421 Using this parameter you can do debugging of EFI memmap 1422 related features. For example, you can do debugging of 1423 Address Range Mirroring feature even if your box 1424 doesn't support it, or mark specific memory as 1425 "soft reserved". 1426 1427 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1428 that is to be dynamically loaded by Linux. If there are 1429 multiple variables with the same name but with different 1430 vendor GUIDs, all of them will be loaded. See 1431 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1432 1433 1434 eisa_irq_edge= [PARISC,HW] 1435 See header of drivers/parisc/eisa.c. 1436 1437 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1438 Format: ekgdboc=kbd 1439 1440 This is designed to be used in conjunction with 1441 the boot argument: earlyprintk=vga 1442 1443 This parameter works in place of the kgdboc parameter 1444 but can only be used if the backing tty is available 1445 very early in the boot process. For early debugging 1446 via a serial port see kgdboc_earlycon instead. 1447 1448 elanfreq= [X86-32] 1449 See comment before function elanfreq_setup() in 1450 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1451 1452 elfcorehdr=[size[KMG]@]offset[KMG] [PPC,SH,X86,S390] 1453 Specifies physical address of start of kernel core 1454 image elf header and optionally the size. Generally 1455 kexec loader will pass this option to capture kernel. 1456 See Documentation/admin-guide/kdump/kdump.rst for details. 1457 1458 enable_mtrr_cleanup [X86] 1459 The kernel tries to adjust MTRR layout from continuous 1460 to discrete, to make X server driver able to add WB 1461 entry later. This parameter enables that. 1462 1463 enable_timer_pin_1 [X86] 1464 Enable PIN 1 of APIC timer 1465 Can be useful to work around chipset bugs 1466 (in particular on some ATI chipsets). 1467 The kernel tries to set a reasonable default. 1468 1469 enforcing= [SELINUX] Set initial enforcing status. 1470 Format: {"0" | "1"} 1471 See security/selinux/Kconfig help text. 1472 0 -- permissive (log only, no denials). 1473 1 -- enforcing (deny and log). 1474 Default value is 0. 1475 Value can be changed at runtime via 1476 /sys/fs/selinux/enforce. 1477 1478 erst_disable [ACPI] 1479 Disable Error Record Serialization Table (ERST) 1480 support. 1481 1482 ether= [HW,NET] Ethernet cards parameters 1483 This option is obsoleted by the "netdev=" option, which 1484 has equivalent usage. See its documentation for details. 1485 1486 evm= [EVM] 1487 Format: { "fix" } 1488 Permit 'security.evm' to be updated regardless of 1489 current integrity status. 1490 1491 early_page_ext [KNL] Enforces page_ext initialization to earlier 1492 stages so cover more early boot allocations. 1493 Please note that as side effect some optimizations 1494 might be disabled to achieve that (e.g. parallelized 1495 memory initialization is disabled) so the boot process 1496 might take longer, especially on systems with a lot of 1497 memory. Available with CONFIG_PAGE_EXTENSION=y. 1498 1499 failslab= 1500 fail_usercopy= 1501 fail_page_alloc= 1502 fail_make_request=[KNL] 1503 General fault injection mechanism. 1504 Format: <interval>,<probability>,<space>,<times> 1505 See also Documentation/fault-injection/. 1506 1507 fb_tunnels= [NET] 1508 Format: { initns | none } 1509 See Documentation/admin-guide/sysctl/net.rst for 1510 fb_tunnels_only_for_init_ns 1511 1512 floppy= [HW] 1513 See Documentation/admin-guide/blockdev/floppy.rst. 1514 1515 forcepae [X86-32] 1516 Forcefully enable Physical Address Extension (PAE). 1517 Many Pentium M systems disable PAE but may have a 1518 functionally usable PAE implementation. 1519 Warning: use of this parameter will taint the kernel 1520 and may cause unknown problems. 1521 1522 ftrace=[tracer] 1523 [FTRACE] will set and start the specified tracer 1524 as early as possible in order to facilitate early 1525 boot debugging. 1526 1527 ftrace_boot_snapshot 1528 [FTRACE] On boot up, a snapshot will be taken of the 1529 ftrace ring buffer that can be read at: 1530 /sys/kernel/tracing/snapshot. 1531 This is useful if you need tracing information from kernel 1532 boot up that is likely to be overridden by user space 1533 start up functionality. 1534 1535 Optionally, the snapshot can also be defined for a tracing 1536 instance that was created by the trace_instance= command 1537 line parameter. 1538 1539 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1540 1541 The above will cause the "foo" tracing instance to trigger 1542 a snapshot at the end of boot up. 1543 1544 ftrace_dump_on_oops[=orig_cpu] 1545 [FTRACE] will dump the trace buffers on oops. 1546 If no parameter is passed, ftrace will dump 1547 buffers of all CPUs, but if you pass orig_cpu, it will 1548 dump only the buffer of the CPU that triggered the 1549 oops. 1550 1551 ftrace_filter=[function-list] 1552 [FTRACE] Limit the functions traced by the function 1553 tracer at boot up. function-list is a comma-separated 1554 list of functions. This list can be changed at run 1555 time by the set_ftrace_filter file in the debugfs 1556 tracing directory. 1557 1558 ftrace_notrace=[function-list] 1559 [FTRACE] Do not trace the functions specified in 1560 function-list. This list can be changed at run time 1561 by the set_ftrace_notrace file in the debugfs 1562 tracing directory. 1563 1564 ftrace_graph_filter=[function-list] 1565 [FTRACE] Limit the top level callers functions traced 1566 by the function graph tracer at boot up. 1567 function-list is a comma-separated list of functions 1568 that can be changed at run time by the 1569 set_graph_function file in the debugfs tracing directory. 1570 1571 ftrace_graph_notrace=[function-list] 1572 [FTRACE] Do not trace from the functions specified in 1573 function-list. This list is a comma-separated list of 1574 functions that can be changed at run time by the 1575 set_graph_notrace file in the debugfs tracing directory. 1576 1577 ftrace_graph_max_depth=<uint> 1578 [FTRACE] Used with the function graph tracer. This is 1579 the max depth it will trace into a function. This value 1580 can be changed at run time by the max_graph_depth file 1581 in the tracefs tracing directory. default: 0 (no limit) 1582 1583 fw_devlink= [KNL] Create device links between consumer and supplier 1584 devices by scanning the firmware to infer the 1585 consumer/supplier relationships. This feature is 1586 especially useful when drivers are loaded as modules as 1587 it ensures proper ordering of tasks like device probing 1588 (suppliers first, then consumers), supplier boot state 1589 clean up (only after all consumers have probed), 1590 suspend/resume & runtime PM (consumers first, then 1591 suppliers). 1592 Format: { off | permissive | on | rpm } 1593 off -- Don't create device links from firmware info. 1594 permissive -- Create device links from firmware info 1595 but use it only for ordering boot state clean 1596 up (sync_state() calls). 1597 on -- Create device links from firmware info and use it 1598 to enforce probe and suspend/resume ordering. 1599 rpm -- Like "on", but also use to order runtime PM. 1600 1601 fw_devlink.strict=<bool> 1602 [KNL] Treat all inferred dependencies as mandatory 1603 dependencies. This only applies for fw_devlink=on|rpm. 1604 Format: <bool> 1605 1606 fw_devlink.sync_state = 1607 [KNL] When all devices that could probe have finished 1608 probing, this parameter controls what to do with 1609 devices that haven't yet received their sync_state() 1610 calls. 1611 Format: { strict | timeout } 1612 strict -- Default. Continue waiting on consumers to 1613 probe successfully. 1614 timeout -- Give up waiting on consumers and call 1615 sync_state() on any devices that haven't yet 1616 received their sync_state() calls after 1617 deferred_probe_timeout has expired or by 1618 late_initcall() if !CONFIG_MODULES. 1619 1620 gamecon.map[2|3]= 1621 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1622 support via parallel port (up to 5 devices per port) 1623 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1624 See also Documentation/input/devices/joystick-parport.rst 1625 1626 gamma= [HW,DRM] 1627 1628 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART 1629 Format: off | on 1630 default: on 1631 1632 gather_data_sampling= 1633 [X86,INTEL] Control the Gather Data Sampling (GDS) 1634 mitigation. 1635 1636 Gather Data Sampling is a hardware vulnerability which 1637 allows unprivileged speculative access to data which was 1638 previously stored in vector registers. 1639 1640 This issue is mitigated by default in updated microcode. 1641 The mitigation may have a performance impact but can be 1642 disabled. On systems without the microcode mitigation 1643 disabling AVX serves as a mitigation. 1644 1645 force: Disable AVX to mitigate systems without 1646 microcode mitigation. No effect if the microcode 1647 mitigation is present. Known to cause crashes in 1648 userspace with buggy AVX enumeration. 1649 1650 off: Disable GDS mitigation. 1651 1652 gcov_persist= [GCOV] When non-zero (default), profiling data for 1653 kernel modules is saved and remains accessible via 1654 debugfs, even when the module is unloaded/reloaded. 1655 When zero, profiling data is discarded and associated 1656 debugfs files are removed at module unload time. 1657 1658 goldfish [X86] Enable the goldfish android emulator platform. 1659 Don't use this when you are not running on the 1660 android emulator 1661 1662 gpio-mockup.gpio_mockup_ranges 1663 [HW] Sets the ranges of gpiochip of for this device. 1664 Format: <start1>,<end1>,<start2>,<end2>... 1665 gpio-mockup.gpio_mockup_named_lines 1666 [HW] Let the driver know GPIO lines should be named. 1667 1668 gpt [EFI] Forces disk with valid GPT signature but 1669 invalid Protective MBR to be treated as GPT. If the 1670 primary GPT is corrupted, it enables the backup/alternate 1671 GPT to be used instead. 1672 1673 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1674 the "Enable 0" bit of the configuration register. 1675 Format: 0 | 1 1676 Default: 0 1677 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1678 the "Enable 0" bit of the configuration register. 1679 Format: 0 | 1 1680 Default: 0 1681 grcan.select= [HW] Select which physical interface to use. 1682 Format: 0 | 1 1683 Default: 0 1684 grcan.txsize= [HW] Sets the size of the tx buffer. 1685 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1686 Default: 1024 1687 grcan.rxsize= [HW] Sets the size of the rx buffer. 1688 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1689 Default: 1024 1690 1691 hardened_usercopy= 1692 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1693 hardening is enabled for this boot. Hardened 1694 usercopy checking is used to protect the kernel 1695 from reading or writing beyond known memory 1696 allocation boundaries as a proactive defense 1697 against bounds-checking flaws in the kernel's 1698 copy_to_user()/copy_from_user() interface. 1699 on Perform hardened usercopy checks (default). 1700 off Disable hardened usercopy checks. 1701 1702 hardlockup_all_cpu_backtrace= 1703 [KNL] Should the hard-lockup detector generate 1704 backtraces on all cpus. 1705 Format: 0 | 1 1706 1707 hashdist= [KNL,NUMA] Large hashes allocated during boot 1708 are distributed across NUMA nodes. Defaults on 1709 for 64-bit NUMA, off otherwise. 1710 Format: 0 | 1 (for off | on) 1711 1712 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1713 1714 hd= [EIDE] (E)IDE hard drive subsystem geometry 1715 Format: <cyl>,<head>,<sect> 1716 1717 hest_disable [ACPI] 1718 Disable Hardware Error Source Table (HEST) support; 1719 corresponding firmware-first mode error processing 1720 logic will be disabled. 1721 1722 hibernate= [HIBERNATION] 1723 noresume Don't check if there's a hibernation image 1724 present during boot. 1725 nocompress Don't compress/decompress hibernation images. 1726 no Disable hibernation and resume. 1727 protect_image Turn on image protection during restoration 1728 (that will set all pages holding image data 1729 during restoration read-only). 1730 1731 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1732 size of <nn>. This works even on boxes that have no 1733 highmem otherwise. This also works to reduce highmem 1734 size on bigger boxes. 1735 1736 highres= [KNL] Enable/disable high resolution timer mode. 1737 Valid parameters: "on", "off" 1738 Default: "on" 1739 1740 hlt [BUGS=ARM,SH] 1741 1742 hostname= [KNL] Set the hostname (aka UTS nodename). 1743 Format: <string> 1744 This allows setting the system's hostname during early 1745 startup. This sets the name returned by gethostname. 1746 Using this parameter to set the hostname makes it 1747 possible to ensure the hostname is correctly set before 1748 any userspace processes run, avoiding the possibility 1749 that a process may call gethostname before the hostname 1750 has been explicitly set, resulting in the calling 1751 process getting an incorrect result. The string must 1752 not exceed the maximum allowed hostname length (usually 1753 64 characters) and will be truncated otherwise. 1754 1755 hpet= [X86-32,HPET] option to control HPET usage 1756 Format: { enable (default) | disable | force | 1757 verbose } 1758 disable: disable HPET and use PIT instead 1759 force: allow force enabled of undocumented chips (ICH4, 1760 VIA, nVidia) 1761 verbose: show contents of HPET registers during setup 1762 1763 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1764 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1765 1766 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1767 If this follows hugepagesz (below), it specifies 1768 the number of pages of hugepagesz to be allocated. 1769 If this is the first HugeTLB parameter on the command 1770 line, it specifies the number of pages to allocate for 1771 the default huge page size. If using node format, the 1772 number of pages to allocate per-node can be specified. 1773 See also Documentation/admin-guide/mm/hugetlbpage.rst. 1774 Format: <integer> or (node format) 1775 <node>:<integer>[,<node>:<integer>] 1776 1777 hugepagesz= 1778 [HW] The size of the HugeTLB pages. This is used in 1779 conjunction with hugepages (above) to allocate huge 1780 pages of a specific size at boot. The pair 1781 hugepagesz=X hugepages=Y can be specified once for 1782 each supported huge page size. Huge page sizes are 1783 architecture dependent. See also 1784 Documentation/admin-guide/mm/hugetlbpage.rst. 1785 Format: size[KMG] 1786 1787 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation 1788 of gigantic hugepages. Or using node format, the size 1789 of a CMA area per node can be specified. 1790 Format: nn[KMGTPE] or (node format) 1791 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 1792 1793 Reserve a CMA area of given size and allocate gigantic 1794 hugepages using the CMA allocator. If enabled, the 1795 boot-time allocation of gigantic hugepages is skipped. 1796 1797 hugetlb_free_vmemmap= 1798 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 1799 enabled. 1800 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 1801 Allows heavy hugetlb users to free up some more 1802 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 1803 Format: { on | off (default) } 1804 1805 on: enable HVO 1806 off: disable HVO 1807 1808 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 1809 the default is on. 1810 1811 Note that the vmemmap pages may be allocated from the added 1812 memory block itself when memory_hotplug.memmap_on_memory is 1813 enabled, those vmemmap pages cannot be optimized even if this 1814 feature is enabled. Other vmemmap pages not allocated from 1815 the added memory block itself do not be affected. 1816 1817 hung_task_panic= 1818 [KNL] Should the hung task detector generate panics. 1819 Format: 0 | 1 1820 1821 A value of 1 instructs the kernel to panic when a 1822 hung task is detected. The default value is controlled 1823 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1824 option. The value selected by this boot parameter can 1825 be changed later by the kernel.hung_task_panic sysctl. 1826 1827 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1828 terminal devices. Valid values: 0..8 1829 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1830 If specified, z/VM IUCV HVC accepts connections 1831 from listed z/VM user IDs only. 1832 1833 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1834 which allow the hypervisor to 'idle' the 1835 guest on lock contention. 1836 1837 i2c_bus= [HW] Override the default board specific I2C bus speed 1838 or register an additional I2C bus that is not 1839 registered from board initialization code. 1840 Format: 1841 <bus_id>,<clkrate> 1842 1843 i8042.debug [HW] Toggle i8042 debug mode 1844 i8042.unmask_kbd_data 1845 [HW] Enable printing of interrupt data from the KBD port 1846 (disabled by default, and as a pre-condition 1847 requires that i8042.debug=1 be enabled) 1848 i8042.direct [HW] Put keyboard port into non-translated mode 1849 i8042.dumbkbd [HW] Pretend that controller can only read data from 1850 keyboard and cannot control its state 1851 (Don't attempt to blink the leds) 1852 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1853 i8042.nokbd [HW] Don't check/create keyboard port 1854 i8042.noloop [HW] Disable the AUX Loopback command while probing 1855 for the AUX port 1856 i8042.nomux [HW] Don't check presence of an active multiplexing 1857 controller 1858 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1859 controllers 1860 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1861 i8042.reset [HW] Reset the controller during init, cleanup and 1862 suspend-to-ram transitions, only during s2r 1863 transitions, or never reset 1864 Format: { 1 | Y | y | 0 | N | n } 1865 1, Y, y: always reset controller 1866 0, N, n: don't ever reset controller 1867 Default: only on s2r transitions on x86; most other 1868 architectures force reset to be always executed 1869 i8042.unlock [HW] Unlock (ignore) the keylock 1870 i8042.kbdreset [HW] Reset device connected to KBD port 1871 i8042.probe_defer 1872 [HW] Allow deferred probing upon i8042 probe errors 1873 1874 i810= [HW,DRM] 1875 1876 i915.invert_brightness= 1877 [DRM] Invert the sense of the variable that is used to 1878 set the brightness of the panel backlight. Normally a 1879 brightness value of 0 indicates backlight switched off, 1880 and the maximum of the brightness value sets the backlight 1881 to maximum brightness. If this parameter is set to 0 1882 (default) and the machine requires it, or this parameter 1883 is set to 1, a brightness value of 0 sets the backlight 1884 to maximum brightness, and the maximum of the brightness 1885 value switches the backlight off. 1886 -1 -- never invert brightness 1887 0 -- machine default 1888 1 -- force brightness inversion 1889 1890 icn= [HW,ISDN] 1891 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1892 1893 1894 idle= [X86] 1895 Format: idle=poll, idle=halt, idle=nomwait 1896 Poll forces a polling idle loop that can slightly 1897 improve the performance of waking up a idle CPU, but 1898 will use a lot of power and make the system run hot. 1899 Not recommended. 1900 idle=halt: Halt is forced to be used for CPU idle. 1901 In such case C2/C3 won't be used again. 1902 idle=nomwait: Disable mwait for CPU C-states 1903 1904 idxd.sva= [HW] 1905 Format: <bool> 1906 Allow force disabling of Shared Virtual Memory (SVA) 1907 support for the idxd driver. By default it is set to 1908 true (1). 1909 1910 idxd.tc_override= [HW] 1911 Format: <bool> 1912 Allow override of default traffic class configuration 1913 for the device. By default it is set to false (0). 1914 1915 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1916 Format: { strict | legacy | 2008 | relaxed } 1917 Default: strict 1918 1919 Choose which programs will be accepted for execution 1920 based on the IEEE 754 NaN encoding(s) supported by 1921 the FPU and the NaN encoding requested with the value 1922 of an ELF file header flag individually set by each 1923 binary. Hardware implementations are permitted to 1924 support either or both of the legacy and the 2008 NaN 1925 encoding mode. 1926 1927 Available settings are as follows: 1928 strict accept binaries that request a NaN encoding 1929 supported by the FPU 1930 legacy only accept legacy-NaN binaries, if supported 1931 by the FPU 1932 2008 only accept 2008-NaN binaries, if supported 1933 by the FPU 1934 relaxed accept any binaries regardless of whether 1935 supported by the FPU 1936 1937 The FPU emulator is always able to support both NaN 1938 encodings, so if no FPU hardware is present or it has 1939 been disabled with 'nofpu', then the settings of 1940 'legacy' and '2008' strap the emulator accordingly, 1941 'relaxed' straps the emulator for both legacy-NaN and 1942 2008-NaN, whereas 'strict' enables legacy-NaN only on 1943 legacy processors and both NaN encodings on MIPS32 or 1944 MIPS64 CPUs. 1945 1946 The setting for ABS.fmt/NEG.fmt instruction execution 1947 mode generally follows that for the NaN encoding, 1948 except where unsupported by hardware. 1949 1950 ignore_loglevel [KNL] 1951 Ignore loglevel setting - this will print /all/ 1952 kernel messages to the console. Useful for debugging. 1953 We also add it as printk module parameter, so users 1954 could change it dynamically, usually by 1955 /sys/module/printk/parameters/ignore_loglevel. 1956 1957 ignore_rlimit_data 1958 Ignore RLIMIT_DATA setting for data mappings, 1959 print warning at first misuse. Can be changed via 1960 /sys/module/kernel/parameters/ignore_rlimit_data. 1961 1962 ihash_entries= [KNL] 1963 Set number of hash buckets for inode cache. 1964 1965 ima_appraise= [IMA] appraise integrity measurements 1966 Format: { "off" | "enforce" | "fix" | "log" } 1967 default: "enforce" 1968 1969 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1970 The builtin appraise policy appraises all files 1971 owned by uid=0. 1972 1973 ima_canonical_fmt [IMA] 1974 Use the canonical format for the binary runtime 1975 measurements, instead of host native format. 1976 1977 ima_hash= [IMA] 1978 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1979 | sha512 | ... } 1980 default: "sha1" 1981 1982 The list of supported hash algorithms is defined 1983 in crypto/hash_info.h. 1984 1985 ima_policy= [IMA] 1986 The builtin policies to load during IMA setup. 1987 Format: "tcb | appraise_tcb | secure_boot | 1988 fail_securely | critical_data" 1989 1990 The "tcb" policy measures all programs exec'd, files 1991 mmap'd for exec, and all files opened with the read 1992 mode bit set by either the effective uid (euid=0) or 1993 uid=0. 1994 1995 The "appraise_tcb" policy appraises the integrity of 1996 all files owned by root. 1997 1998 The "secure_boot" policy appraises the integrity 1999 of files (eg. kexec kernel image, kernel modules, 2000 firmware, policy, etc) based on file signatures. 2001 2002 The "fail_securely" policy forces file signature 2003 verification failure also on privileged mounted 2004 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2005 flag. 2006 2007 The "critical_data" policy measures kernel integrity 2008 critical data. 2009 2010 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2011 Load a policy which meets the needs of the Trusted 2012 Computing Base. This means IMA will measure all 2013 programs exec'd, files mmap'd for exec, and all files 2014 opened for read by uid=0. 2015 2016 ima_template= [IMA] 2017 Select one of defined IMA measurements template formats. 2018 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2019 "ima-sigv2" } 2020 Default: "ima-ng" 2021 2022 ima_template_fmt= 2023 [IMA] Define a custom template format. 2024 Format: { "field1|...|fieldN" } 2025 2026 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2027 Format: <min_file_size> 2028 Set the minimal file size for using asynchronous hash. 2029 If left unspecified, ahash usage is disabled. 2030 2031 ahash performance varies for different data sizes on 2032 different crypto accelerators. This option can be used 2033 to achieve the best performance for a particular HW. 2034 2035 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2036 Format: <bufsize> 2037 Set hashing buffer size. Default: 4k. 2038 2039 ahash performance varies for different chunk sizes on 2040 different crypto accelerators. This option can be used 2041 to achieve best performance for particular HW. 2042 2043 init= [KNL] 2044 Format: <full_path> 2045 Run specified binary instead of /sbin/init as init 2046 process. 2047 2048 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2049 for working out where the kernel is dying during 2050 startup. 2051 2052 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2053 initcall functions. Useful for debugging built-in 2054 modules and initcalls. 2055 2056 initramfs_async= [KNL] 2057 Format: <bool> 2058 Default: 1 2059 This parameter controls whether the initramfs 2060 image is unpacked asynchronously, concurrently 2061 with devices being probed and 2062 initialized. This should normally just work, 2063 but as a debugging aid, one can get the 2064 historical behaviour of the initramfs 2065 unpacking being completed before device_ and 2066 late_ initcalls. 2067 2068 initrd= [BOOT] Specify the location of the initial ramdisk 2069 2070 initrdmem= [KNL] Specify a physical address and size from which to 2071 load the initrd. If an initrd is compiled in or 2072 specified in the bootparams, it takes priority over this 2073 setting. 2074 Format: ss[KMG],nn[KMG] 2075 Default is 0, 0 2076 2077 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 2078 zeroes. 2079 Format: 0 | 1 2080 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2081 2082 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 2083 Format: 0 | 1 2084 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2085 2086 init_pkru= [X86] Specify the default memory protection keys rights 2087 register contents for all processes. 0x55555554 by 2088 default (disallow access to all but pkey 0). Can 2089 override in debugfs after boot. 2090 2091 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2092 Format: <irq> 2093 2094 int_pln_enable [X86] Enable power limit notification interrupt 2095 2096 integrity_audit=[IMA] 2097 Format: { "0" | "1" } 2098 0 -- basic integrity auditing messages. (Default) 2099 1 -- additional integrity auditing messages. 2100 2101 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2102 on 2103 Enable intel iommu driver. 2104 off 2105 Disable intel iommu driver. 2106 igfx_off [Default Off] 2107 By default, gfx is mapped as normal device. If a gfx 2108 device has a dedicated DMAR unit, the DMAR unit is 2109 bypassed by not enabling DMAR with this option. In 2110 this case, gfx device will use physical address for 2111 DMA. 2112 strict [Default Off] 2113 Deprecated, equivalent to iommu.strict=1. 2114 sp_off [Default Off] 2115 By default, super page will be supported if Intel IOMMU 2116 has the capability. With this option, super page will 2117 not be supported. 2118 sm_on 2119 Enable the Intel IOMMU scalable mode if the hardware 2120 advertises that it has support for the scalable mode 2121 translation. 2122 sm_off 2123 Disallow use of the Intel IOMMU scalable mode. 2124 tboot_noforce [Default Off] 2125 Do not force the Intel IOMMU enabled under tboot. 2126 By default, tboot will force Intel IOMMU on, which 2127 could harm performance of some high-throughput 2128 devices like 40GBit network cards, even if identity 2129 mapping is enabled. 2130 Note that using this option lowers the security 2131 provided by tboot because it makes the system 2132 vulnerable to DMA attacks. 2133 2134 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2135 0 disables intel_idle and fall back on acpi_idle. 2136 1 to 9 specify maximum depth of C-state. 2137 2138 intel_pstate= [X86] 2139 disable 2140 Do not enable intel_pstate as the default 2141 scaling driver for the supported processors 2142 active 2143 Use intel_pstate driver to bypass the scaling 2144 governors layer of cpufreq and provides it own 2145 algorithms for p-state selection. There are two 2146 P-state selection algorithms provided by 2147 intel_pstate in the active mode: powersave and 2148 performance. The way they both operate depends 2149 on whether or not the hardware managed P-states 2150 (HWP) feature has been enabled in the processor 2151 and possibly on the processor model. 2152 passive 2153 Use intel_pstate as a scaling driver, but configure it 2154 to work with generic cpufreq governors (instead of 2155 enabling its internal governor). This mode cannot be 2156 used along with the hardware-managed P-states (HWP) 2157 feature. 2158 force 2159 Enable intel_pstate on systems that prohibit it by default 2160 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2161 instead of acpi-cpufreq may disable platform features, such 2162 as thermal controls and power capping, that rely on ACPI 2163 P-States information being indicated to OSPM and therefore 2164 should be used with caution. This option does not work with 2165 processors that aren't supported by the intel_pstate driver 2166 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2167 no_hwp 2168 Do not enable hardware P state control (HWP) 2169 if available. 2170 hwp_only 2171 Only load intel_pstate on systems which support 2172 hardware P state control (HWP) if available. 2173 support_acpi_ppc 2174 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2175 Description Table, specifies preferred power management 2176 profile as "Enterprise Server" or "Performance Server", 2177 then this feature is turned on by default. 2178 per_cpu_perf_limits 2179 Allow per-logical-CPU P-State performance control limits using 2180 cpufreq sysfs interface 2181 2182 intremap= [X86-64, Intel-IOMMU] 2183 on enable Interrupt Remapping (default) 2184 off disable Interrupt Remapping 2185 nosid disable Source ID checking 2186 no_x2apic_optout 2187 BIOS x2APIC opt-out request will be ignored 2188 nopost disable Interrupt Posting 2189 2190 iomem= Disable strict checking of access to MMIO memory 2191 strict regions from userspace. 2192 relaxed 2193 2194 iommu= [X86] 2195 off 2196 force 2197 noforce 2198 biomerge 2199 panic 2200 nopanic 2201 merge 2202 nomerge 2203 soft 2204 pt [X86] 2205 nopt [X86] 2206 nobypass [PPC/POWERNV] 2207 Disable IOMMU bypass, using IOMMU for PCI devices. 2208 2209 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices. 2210 Format: { "0" | "1" } 2211 0 - Try to allocate a 32-bit DMA address first, before 2212 falling back to the full range if needed. 2213 1 - Allocate directly from the full usable range, 2214 forcing Dual Address Cycle for PCI cards supporting 2215 greater than 32-bit addressing. 2216 2217 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour 2218 Format: { "0" | "1" } 2219 0 - Lazy mode. 2220 Request that DMA unmap operations use deferred 2221 invalidation of hardware TLBs, for increased 2222 throughput at the cost of reduced device isolation. 2223 Will fall back to strict mode if not supported by 2224 the relevant IOMMU driver. 2225 1 - Strict mode. 2226 DMA unmap operations invalidate IOMMU hardware TLBs 2227 synchronously. 2228 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2229 Note: on x86, strict mode specified via one of the 2230 legacy driver-specific options takes precedence. 2231 2232 iommu.passthrough= 2233 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 2234 Format: { "0" | "1" } 2235 0 - Use IOMMU translation for DMA. 2236 1 - Bypass the IOMMU for DMA. 2237 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2238 2239 io7= [HW] IO7 for Marvel-based Alpha systems 2240 See comment before marvel_specify_io7 in 2241 arch/alpha/kernel/core_marvel.c. 2242 2243 io_delay= [X86] I/O delay method 2244 0x80 2245 Standard port 0x80 based delay 2246 0xed 2247 Alternate port 0xed based delay (needed on some systems) 2248 udelay 2249 Simple two microseconds delay 2250 none 2251 No delay 2252 2253 ip= [IP_PNP] 2254 See Documentation/admin-guide/nfs/nfsroot.rst. 2255 2256 ipcmni_extend [KNL] Extend the maximum number of unique System V 2257 IPC identifiers from 32,768 to 16,777,216. 2258 2259 irqaffinity= [SMP] Set the default irq affinity mask 2260 The argument is a cpu list, as described above. 2261 2262 irqchip.gicv2_force_probe= 2263 [ARM, ARM64] 2264 Format: <bool> 2265 Force the kernel to look for the second 4kB page 2266 of a GICv2 controller even if the memory range 2267 exposed by the device tree is too small. 2268 2269 irqchip.gicv3_nolpi= 2270 [ARM, ARM64] 2271 Force the kernel to ignore the availability of 2272 LPIs (and by consequence ITSs). Intended for system 2273 that use the kernel as a bootloader, and thus want 2274 to let secondary kernels in charge of setting up 2275 LPIs. 2276 2277 irqchip.gicv3_pseudo_nmi= [ARM64] 2278 Enables support for pseudo-NMIs in the kernel. This 2279 requires the kernel to be built with 2280 CONFIG_ARM64_PSEUDO_NMI. 2281 2282 irqfixup [HW] 2283 When an interrupt is not handled search all handlers 2284 for it. Intended to get systems with badly broken 2285 firmware running. 2286 2287 irqpoll [HW] 2288 When an interrupt is not handled search all handlers 2289 for it. Also check all handlers each timer 2290 interrupt. Intended to get systems with badly broken 2291 firmware running. 2292 2293 isapnp= [ISAPNP] 2294 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2295 2296 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2297 [Deprecated - use cpusets instead] 2298 Format: [flag-list,]<cpu-list> 2299 2300 Specify one or more CPUs to isolate from disturbances 2301 specified in the flag list (default: domain): 2302 2303 nohz 2304 Disable the tick when a single task runs. 2305 2306 A residual 1Hz tick is offloaded to workqueues, which you 2307 need to affine to housekeeping through the global 2308 workqueue's affinity configured via the 2309 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2310 by using the 'domain' flag described below. 2311 2312 NOTE: by default the global workqueue runs on all CPUs, 2313 so to protect individual CPUs the 'cpumask' file has to 2314 be configured manually after bootup. 2315 2316 domain 2317 Isolate from the general SMP balancing and scheduling 2318 algorithms. Note that performing domain isolation this way 2319 is irreversible: it's not possible to bring back a CPU to 2320 the domains once isolated through isolcpus. It's strongly 2321 advised to use cpusets instead to disable scheduler load 2322 balancing through the "cpuset.sched_load_balance" file. 2323 It offers a much more flexible interface where CPUs can 2324 move in and out of an isolated set anytime. 2325 2326 You can move a process onto or off an "isolated" CPU via 2327 the CPU affinity syscalls or cpuset. 2328 <cpu number> begins at 0 and the maximum value is 2329 "number of CPUs in system - 1". 2330 2331 managed_irq 2332 2333 Isolate from being targeted by managed interrupts 2334 which have an interrupt mask containing isolated 2335 CPUs. The affinity of managed interrupts is 2336 handled by the kernel and cannot be changed via 2337 the /proc/irq/* interfaces. 2338 2339 This isolation is best effort and only effective 2340 if the automatically assigned interrupt mask of a 2341 device queue contains isolated and housekeeping 2342 CPUs. If housekeeping CPUs are online then such 2343 interrupts are directed to the housekeeping CPU 2344 so that IO submitted on the housekeeping CPU 2345 cannot disturb the isolated CPU. 2346 2347 If a queue's affinity mask contains only isolated 2348 CPUs then this parameter has no effect on the 2349 interrupt routing decision, though interrupts are 2350 only delivered when tasks running on those 2351 isolated CPUs submit IO. IO submitted on 2352 housekeeping CPUs has no influence on those 2353 queues. 2354 2355 The format of <cpu-list> is described above. 2356 2357 iucv= [HW,NET] 2358 2359 ivrs_ioapic [HW,X86-64] 2360 Provide an override to the IOAPIC-ID<->DEVICE-ID 2361 mapping provided in the IVRS ACPI table. 2362 By default, PCI segment is 0, and can be omitted. 2363 2364 For example, to map IOAPIC-ID decimal 10 to 2365 PCI segment 0x1 and PCI device 00:14.0, 2366 write the parameter as: 2367 ivrs_ioapic=10@0001:00:14.0 2368 2369 Deprecated formats: 2370 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2371 write the parameter as: 2372 ivrs_ioapic[10]=00:14.0 2373 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2374 PCI device 00:14.0 write the parameter as: 2375 ivrs_ioapic[10]=0001:00:14.0 2376 2377 ivrs_hpet [HW,X86-64] 2378 Provide an override to the HPET-ID<->DEVICE-ID 2379 mapping provided in the IVRS ACPI table. 2380 By default, PCI segment is 0, and can be omitted. 2381 2382 For example, to map HPET-ID decimal 10 to 2383 PCI segment 0x1 and PCI device 00:14.0, 2384 write the parameter as: 2385 ivrs_hpet=10@0001:00:14.0 2386 2387 Deprecated formats: 2388 * To map HPET-ID decimal 0 to PCI device 00:14.0 2389 write the parameter as: 2390 ivrs_hpet[0]=00:14.0 2391 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2392 PCI device 00:14.0 write the parameter as: 2393 ivrs_ioapic[10]=0001:00:14.0 2394 2395 ivrs_acpihid [HW,X86-64] 2396 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2397 mapping provided in the IVRS ACPI table. 2398 By default, PCI segment is 0, and can be omitted. 2399 2400 For example, to map UART-HID:UID AMD0020:0 to 2401 PCI segment 0x1 and PCI device ID 00:14.5, 2402 write the parameter as: 2403 ivrs_acpihid=AMD0020:0@0001:00:14.5 2404 2405 Deprecated formats: 2406 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2407 PCI device ID 00:14.5, write the parameter as: 2408 ivrs_acpihid[00:14.5]=AMD0020:0 2409 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2410 PCI device ID 00:14.5, write the parameter as: 2411 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2412 2413 js= [HW,JOY] Analog joystick 2414 See Documentation/input/joydev/joystick.rst. 2415 2416 kasan_multi_shot 2417 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2418 report on every invalid memory access. Without this 2419 parameter KASAN will print report only for the first 2420 invalid access. 2421 2422 keep_bootcon [KNL] 2423 Do not unregister boot console at start. This is only 2424 useful for debugging when something happens in the window 2425 between unregistering the boot console and initializing 2426 the real console. 2427 2428 keepinitrd [HW,ARM] 2429 2430 kernelcore= [KNL,X86,IA-64,PPC] 2431 Format: nn[KMGTPE] | nn% | "mirror" 2432 This parameter specifies the amount of memory usable by 2433 the kernel for non-movable allocations. The requested 2434 amount is spread evenly throughout all nodes in the 2435 system as ZONE_NORMAL. The remaining memory is used for 2436 movable memory in its own zone, ZONE_MOVABLE. In the 2437 event, a node is too small to have both ZONE_NORMAL and 2438 ZONE_MOVABLE, kernelcore memory will take priority and 2439 other nodes will have a larger ZONE_MOVABLE. 2440 2441 ZONE_MOVABLE is used for the allocation of pages that 2442 may be reclaimed or moved by the page migration 2443 subsystem. Note that allocations like PTEs-from-HighMem 2444 still use the HighMem zone if it exists, and the Normal 2445 zone if it does not. 2446 2447 It is possible to specify the exact amount of memory in 2448 the form of "nn[KMGTPE]", a percentage of total system 2449 memory in the form of "nn%", or "mirror". If "mirror" 2450 option is specified, mirrored (reliable) memory is used 2451 for non-movable allocations and remaining memory is used 2452 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2453 are exclusive, so you cannot specify multiple forms. 2454 2455 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2456 Format: <Controller#>[,poll interval] 2457 The controller # is the number of the ehci usb debug 2458 port as it is probed via PCI. The poll interval is 2459 optional and is the number seconds in between 2460 each poll cycle to the debug port in case you need 2461 the functionality for interrupting the kernel with 2462 gdb or control-c on the dbgp connection. When 2463 not using this parameter you use sysrq-g to break into 2464 the kernel debugger. 2465 2466 kgdboc= [KGDB,HW] kgdb over consoles. 2467 Requires a tty driver that supports console polling, 2468 or a supported polling keyboard driver (non-usb). 2469 Serial only format: <serial_device>[,baud] 2470 keyboard only format: kbd 2471 keyboard and serial format: kbd,<serial_device>[,baud] 2472 Optional Kernel mode setting: 2473 kms, kbd format: kms,kbd 2474 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2475 2476 kgdboc_earlycon= [KGDB,HW] 2477 If the boot console provides the ability to read 2478 characters and can work in polling mode, you can use 2479 this parameter to tell kgdb to use it as a backend 2480 until the normal console is registered. Intended to 2481 be used together with the kgdboc parameter which 2482 specifies the normal console to transition to. 2483 2484 The name of the early console should be specified 2485 as the value of this parameter. Note that the name of 2486 the early console might be different than the tty 2487 name passed to kgdboc. It's OK to leave the value 2488 blank and the first boot console that implements 2489 read() will be picked. 2490 2491 kgdbwait [KGDB] Stop kernel execution and enter the 2492 kernel debugger at the earliest opportunity. 2493 2494 kmac= [MIPS] Korina ethernet MAC address. 2495 Configure the RouterBoard 532 series on-chip 2496 Ethernet adapter MAC address. 2497 2498 kmemleak= [KNL] Boot-time kmemleak enable/disable 2499 Valid arguments: on, off 2500 Default: on 2501 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2502 the default is off. 2503 2504 kprobe_event=[probe-list] 2505 [FTRACE] Add kprobe events and enable at boot time. 2506 The probe-list is a semicolon delimited list of probe 2507 definitions. Each definition is same as kprobe_events 2508 interface, but the parameters are comma delimited. 2509 For example, to add a kprobe event on vfs_read with 2510 arg1 and arg2, add to the command line; 2511 2512 kprobe_event=p,vfs_read,$arg1,$arg2 2513 2514 See also Documentation/trace/kprobetrace.rst "Kernel 2515 Boot Parameter" section. 2516 2517 kpti= [ARM64] Control page table isolation of user 2518 and kernel address spaces. 2519 Default: enabled on cores which need mitigation. 2520 0: force disabled 2521 1: force enabled 2522 2523 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2524 CONFIG_KUNIT to be set to be fully enabled. The 2525 default value can be overridden via 2526 KUNIT_DEFAULT_ENABLED. 2527 Default is 1 (enabled) 2528 2529 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2530 Default is 0 (don't ignore, but inject #GP) 2531 2532 kvm.eager_page_split= 2533 [KVM,X86] Controls whether or not KVM will try to 2534 proactively split all huge pages during dirty logging. 2535 Eager page splitting reduces interruptions to vCPU 2536 execution by eliminating the write-protection faults 2537 and MMU lock contention that would otherwise be 2538 required to split huge pages lazily. 2539 2540 VM workloads that rarely perform writes or that write 2541 only to a small region of VM memory may benefit from 2542 disabling eager page splitting to allow huge pages to 2543 still be used for reads. 2544 2545 The behavior of eager page splitting depends on whether 2546 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 2547 disabled, all huge pages in a memslot will be eagerly 2548 split when dirty logging is enabled on that memslot. If 2549 enabled, eager page splitting will be performed during 2550 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 2551 cleared. 2552 2553 Eager page splitting is only supported when kvm.tdp_mmu=Y. 2554 2555 Default is Y (on). 2556 2557 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2558 Default is false (don't support). 2559 2560 kvm.nx_huge_pages= 2561 [KVM] Controls the software workaround for the 2562 X86_BUG_ITLB_MULTIHIT bug. 2563 force : Always deploy workaround. 2564 off : Never deploy workaround. 2565 auto : Deploy workaround based on the presence of 2566 X86_BUG_ITLB_MULTIHIT. 2567 2568 Default is 'auto'. 2569 2570 If the software workaround is enabled for the host, 2571 guests do need not to enable it for nested guests. 2572 2573 kvm.nx_huge_pages_recovery_ratio= 2574 [KVM] Controls how many 4KiB pages are periodically zapped 2575 back to huge pages. 0 disables the recovery, otherwise if 2576 the value is N KVM will zap 1/Nth of the 4KiB pages every 2577 period (see below). The default is 60. 2578 2579 kvm.nx_huge_pages_recovery_period_ms= 2580 [KVM] Controls the time period at which KVM zaps 4KiB pages 2581 back to huge pages. If the value is a non-zero N, KVM will 2582 zap a portion (see ratio above) of the pages every N msecs. 2583 If the value is 0 (the default), KVM will pick a period based 2584 on the ratio, such that a page is zapped after 1 hour on average. 2585 2586 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 2587 KVM/SVM. Default is 1 (enabled). 2588 2589 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 2590 a.k.a. Two-Dimensional Page Tables. Default is 1 2591 (enabled). Disable by KVM if hardware lacks support 2592 for NPT. 2593 2594 kvm-arm.mode= 2595 [KVM,ARM] Select one of KVM/arm64's modes of operation. 2596 2597 none: Forcefully disable KVM. 2598 2599 nvhe: Standard nVHE-based mode, without support for 2600 protected guests. 2601 2602 protected: nVHE-based mode with support for guests whose 2603 state is kept private from the host. 2604 2605 nested: VHE-based mode with support for nested 2606 virtualization. Requires at least ARMv8.3 2607 hardware. 2608 2609 Defaults to VHE/nVHE based on hardware support. Setting 2610 mode to "protected" will disable kexec and hibernation 2611 for the host. "nested" is experimental and should be 2612 used with extreme caution. 2613 2614 kvm-arm.vgic_v3_group0_trap= 2615 [KVM,ARM] Trap guest accesses to GICv3 group-0 2616 system registers 2617 2618 kvm-arm.vgic_v3_group1_trap= 2619 [KVM,ARM] Trap guest accesses to GICv3 group-1 2620 system registers 2621 2622 kvm-arm.vgic_v3_common_trap= 2623 [KVM,ARM] Trap guest accesses to GICv3 common 2624 system registers 2625 2626 kvm-arm.vgic_v4_enable= 2627 [KVM,ARM] Allow use of GICv4 for direct injection of 2628 LPIs. 2629 2630 kvm_cma_resv_ratio=n [PPC] 2631 Reserves given percentage from system memory area for 2632 contiguous memory allocation for KVM hash pagetable 2633 allocation. 2634 By default it reserves 5% of total system memory. 2635 Format: <integer> 2636 Default: 5 2637 2638 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 2639 a.k.a. Two-Dimensional Page Tables. Default is 1 2640 (enabled). Disable by KVM if hardware lacks support 2641 for EPT. 2642 2643 kvm-intel.emulate_invalid_guest_state= 2644 [KVM,Intel] Control whether to emulate invalid guest 2645 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 2646 as guest state is never invalid for unrestricted 2647 guests. This param doesn't apply to nested guests (L2), 2648 as KVM never emulates invalid L2 guest state. 2649 Default is 1 (enabled). 2650 2651 kvm-intel.flexpriority= 2652 [KVM,Intel] Control KVM's use of FlexPriority feature 2653 (TPR shadow). Default is 1 (enabled). Disable by KVM if 2654 hardware lacks support for it. 2655 2656 kvm-intel.nested= 2657 [KVM,Intel] Control nested virtualization feature in 2658 KVM/VMX. Default is 1 (enabled). 2659 2660 kvm-intel.unrestricted_guest= 2661 [KVM,Intel] Control KVM's use of unrestricted guest 2662 feature (virtualized real and unpaged mode). Default 2663 is 1 (enabled). Disable by KVM if EPT is disabled or 2664 hardware lacks support for it. 2665 2666 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2667 CVE-2018-3620. 2668 2669 Valid arguments: never, cond, always 2670 2671 always: L1D cache flush on every VMENTER. 2672 cond: Flush L1D on VMENTER only when the code between 2673 VMEXIT and VMENTER can leak host memory. 2674 never: Disables the mitigation 2675 2676 Default is cond (do L1 cache flush in specific instances) 2677 2678 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 2679 Identification feature (tagged TLBs). Default is 1 2680 (enabled). Disable by KVM if hardware lacks support 2681 for it. 2682 2683 l1d_flush= [X86,INTEL] 2684 Control mitigation for L1D based snooping vulnerability. 2685 2686 Certain CPUs are vulnerable to an exploit against CPU 2687 internal buffers which can forward information to a 2688 disclosure gadget under certain conditions. 2689 2690 In vulnerable processors, the speculatively 2691 forwarded data can be used in a cache side channel 2692 attack, to access data to which the attacker does 2693 not have direct access. 2694 2695 This parameter controls the mitigation. The 2696 options are: 2697 2698 on - enable the interface for the mitigation 2699 2700 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2701 affected CPUs 2702 2703 The kernel PTE inversion protection is unconditionally 2704 enabled and cannot be disabled. 2705 2706 full 2707 Provides all available mitigations for the 2708 L1TF vulnerability. Disables SMT and 2709 enables all mitigations in the 2710 hypervisors, i.e. unconditional L1D flush. 2711 2712 SMT control and L1D flush control via the 2713 sysfs interface is still possible after 2714 boot. Hypervisors will issue a warning 2715 when the first VM is started in a 2716 potentially insecure configuration, 2717 i.e. SMT enabled or L1D flush disabled. 2718 2719 full,force 2720 Same as 'full', but disables SMT and L1D 2721 flush runtime control. Implies the 2722 'nosmt=force' command line option. 2723 (i.e. sysfs control of SMT is disabled.) 2724 2725 flush 2726 Leaves SMT enabled and enables the default 2727 hypervisor mitigation, i.e. conditional 2728 L1D flush. 2729 2730 SMT control and L1D flush control via the 2731 sysfs interface is still possible after 2732 boot. Hypervisors will issue a warning 2733 when the first VM is started in a 2734 potentially insecure configuration, 2735 i.e. SMT enabled or L1D flush disabled. 2736 2737 flush,nosmt 2738 2739 Disables SMT and enables the default 2740 hypervisor mitigation. 2741 2742 SMT control and L1D flush control via the 2743 sysfs interface is still possible after 2744 boot. Hypervisors will issue a warning 2745 when the first VM is started in a 2746 potentially insecure configuration, 2747 i.e. SMT enabled or L1D flush disabled. 2748 2749 flush,nowarn 2750 Same as 'flush', but hypervisors will not 2751 warn when a VM is started in a potentially 2752 insecure configuration. 2753 2754 off 2755 Disables hypervisor mitigations and doesn't 2756 emit any warnings. 2757 It also drops the swap size and available 2758 RAM limit restriction on both hypervisor and 2759 bare metal. 2760 2761 Default is 'flush'. 2762 2763 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2764 2765 l2cr= [PPC] 2766 2767 l3cr= [PPC] 2768 2769 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2770 disabled it. 2771 2772 lapic= [X86,APIC] Do not use TSC deadline 2773 value for LAPIC timer one-shot implementation. Default 2774 back to the programmable timer unit in the LAPIC. 2775 Format: notscdeadline 2776 2777 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2778 in C2 power state. 2779 2780 libata.dma= [LIBATA] DMA control 2781 libata.dma=0 Disable all PATA and SATA DMA 2782 libata.dma=1 PATA and SATA Disk DMA only 2783 libata.dma=2 ATAPI (CDROM) DMA only 2784 libata.dma=4 Compact Flash DMA only 2785 Combinations also work, so libata.dma=3 enables DMA 2786 for disks and CDROMs, but not CFs. 2787 2788 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2789 libata.ignore_hpa=0 keep BIOS limits (default) 2790 libata.ignore_hpa=1 ignore limits, using full disk 2791 2792 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2793 when set. 2794 Format: <int> 2795 2796 libata.force= [LIBATA] Force configurations. The format is a comma- 2797 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 2798 PORT and DEVICE are decimal numbers matching port, link 2799 or device. Basically, it matches the ATA ID string 2800 printed on console by libata. If the whole ID part is 2801 omitted, the last PORT and DEVICE values are used. If 2802 ID hasn't been specified yet, the configuration applies 2803 to all ports, links and devices. 2804 2805 If only DEVICE is omitted, the parameter applies to 2806 the port and all links and devices behind it. DEVICE 2807 number of 0 either selects the first device or the 2808 first fan-out link behind PMP device. It does not 2809 select the host link. DEVICE number of 15 selects the 2810 host link and device attached to it. 2811 2812 The VAL specifies the configuration to force. As long 2813 as there is no ambiguity, shortcut notation is allowed. 2814 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2815 The following configurations can be forced. 2816 2817 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2818 Any ID with matching PORT is used. 2819 2820 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2821 2822 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2823 udma[/][16,25,33,44,66,100,133] notation is also 2824 allowed. 2825 2826 * nohrst, nosrst, norst: suppress hard, soft and both 2827 resets. 2828 2829 * rstonce: only attempt one reset during hot-unplug 2830 link recovery. 2831 2832 * [no]dbdelay: Enable or disable the extra 200ms delay 2833 before debouncing a link PHY and device presence 2834 detection. 2835 2836 * [no]ncq: Turn on or off NCQ. 2837 2838 * [no]ncqtrim: Enable or disable queued DSM TRIM. 2839 2840 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 2841 2842 * [no]trim: Enable or disable (unqueued) TRIM. 2843 2844 * trim_zero: Indicate that TRIM command zeroes data. 2845 2846 * max_trim_128m: Set 128M maximum trim size limit. 2847 2848 * [no]dma: Turn on or off DMA transfers. 2849 2850 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 2851 2852 * atapi_mod16_dma: Enable the use of ATAPI DMA for 2853 commands that are not a multiple of 16 bytes. 2854 2855 * [no]dmalog: Enable or disable the use of the 2856 READ LOG DMA EXT command to access logs. 2857 2858 * [no]iddevlog: Enable or disable access to the 2859 identify device data log. 2860 2861 * [no]logdir: Enable or disable access to the general 2862 purpose log directory. 2863 2864 * max_sec_128: Set transfer size limit to 128 sectors. 2865 2866 * max_sec_1024: Set or clear transfer size limit to 2867 1024 sectors. 2868 2869 * max_sec_lba48: Set or clear transfer size limit to 2870 65535 sectors. 2871 2872 * [no]lpm: Enable or disable link power management. 2873 2874 * [no]setxfer: Indicate if transfer speed mode setting 2875 should be skipped. 2876 2877 * [no]fua: Disable or enable FUA (Force Unit Access) 2878 support for devices supporting this feature. 2879 2880 * dump_id: Dump IDENTIFY data. 2881 2882 * disable: Disable this device. 2883 2884 If there are multiple matching configurations changing 2885 the same attribute, the last one is used. 2886 2887 load_ramdisk= [RAM] [Deprecated] 2888 2889 lockd.nlm_grace_period=P [NFS] Assign grace period. 2890 Format: <integer> 2891 2892 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2893 Format: <integer> 2894 2895 lockd.nlm_timeout=T [NFS] Assign timeout value. 2896 Format: <integer> 2897 2898 lockd.nlm_udpport=M [NFS] Assign UDP port. 2899 Format: <integer> 2900 2901 lockdown= [SECURITY] 2902 { integrity | confidentiality } 2903 Enable the kernel lockdown feature. If set to 2904 integrity, kernel features that allow userland to 2905 modify the running kernel are disabled. If set to 2906 confidentiality, kernel features that allow userland 2907 to extract confidential information from the kernel 2908 are also disabled. 2909 2910 locktorture.nreaders_stress= [KNL] 2911 Set the number of locking read-acquisition kthreads. 2912 Defaults to being automatically set based on the 2913 number of online CPUs. 2914 2915 locktorture.nwriters_stress= [KNL] 2916 Set the number of locking write-acquisition kthreads. 2917 2918 locktorture.onoff_holdoff= [KNL] 2919 Set time (s) after boot for CPU-hotplug testing. 2920 2921 locktorture.onoff_interval= [KNL] 2922 Set time (s) between CPU-hotplug operations, or 2923 zero to disable CPU-hotplug testing. 2924 2925 locktorture.shuffle_interval= [KNL] 2926 Set task-shuffle interval (jiffies). Shuffling 2927 tasks allows some CPUs to go into dyntick-idle 2928 mode during the locktorture test. 2929 2930 locktorture.shutdown_secs= [KNL] 2931 Set time (s) after boot system shutdown. This 2932 is useful for hands-off automated testing. 2933 2934 locktorture.stat_interval= [KNL] 2935 Time (s) between statistics printk()s. 2936 2937 locktorture.stutter= [KNL] 2938 Time (s) to stutter testing, for example, 2939 specifying five seconds causes the test to run for 2940 five seconds, wait for five seconds, and so on. 2941 This tests the locking primitive's ability to 2942 transition abruptly to and from idle. 2943 2944 locktorture.torture_type= [KNL] 2945 Specify the locking implementation to test. 2946 2947 locktorture.writer_fifo= [KNL] 2948 Run the write-side locktorture kthreads at 2949 sched_set_fifo() real-time priority. 2950 2951 locktorture.verbose= [KNL] 2952 Enable additional printk() statements. 2953 2954 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2955 Format: <irq> 2956 2957 loglevel= All Kernel Messages with a loglevel smaller than the 2958 console loglevel will be printed to the console. It can 2959 also be changed with klogd or other programs. The 2960 loglevels are defined as follows: 2961 2962 0 (KERN_EMERG) system is unusable 2963 1 (KERN_ALERT) action must be taken immediately 2964 2 (KERN_CRIT) critical conditions 2965 3 (KERN_ERR) error conditions 2966 4 (KERN_WARNING) warning conditions 2967 5 (KERN_NOTICE) normal but significant condition 2968 6 (KERN_INFO) informational 2969 7 (KERN_DEBUG) debug-level messages 2970 2971 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2972 in bytes. n must be a power of two and greater 2973 than the minimal size. The minimal size is defined 2974 by LOG_BUF_SHIFT kernel config parameter. There is 2975 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2976 that allows to increase the default size depending on 2977 the number of CPUs. See init/Kconfig for more details. 2978 2979 logo.nologo [FB] Disables display of the built-in Linux logo. 2980 This may be used to provide more screen space for 2981 kernel log messages and is useful when debugging 2982 kernel boot problems. 2983 2984 lp=0 [LP] Specify parallel ports to use, e.g, 2985 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2986 lp=reset first parallel port). 'lp=0' disables the 2987 lp=auto printer driver. 'lp=reset' (which can be 2988 specified in addition to the ports) causes 2989 attached printers to be reset. Using 2990 lp=port1,port2,... specifies the parallel ports 2991 to associate lp devices with, starting with 2992 lp0. A port specification may be 'none' to skip 2993 that lp device, or a parport name such as 2994 'parport0'. Specifying 'lp=auto' instead of a 2995 port specification list means that device IDs 2996 from each port should be examined, to see if 2997 an IEEE 1284-compliant printer is attached; if 2998 so, the driver will manage that printer. 2999 See also header of drivers/char/lp.c. 3000 3001 lpj=n [KNL] 3002 Sets loops_per_jiffy to given constant, thus avoiding 3003 time-consuming boot-time autodetection (up to 250 ms per 3004 CPU). 0 enables autodetection (default). To determine 3005 the correct value for your kernel, boot with normal 3006 autodetection and see what value is printed. Note that 3007 on SMP systems the preset will be applied to all CPUs, 3008 which is likely to cause problems if your CPUs need 3009 significantly divergent settings. An incorrect value 3010 will cause delays in the kernel to be wrong, leading to 3011 unpredictable I/O errors and other breakage. Although 3012 unlikely, in the extreme case this might damage your 3013 hardware. 3014 3015 ltpc= [NET] 3016 Format: <io>,<irq>,<dma> 3017 3018 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3019 3020 lsm=lsm1,...,lsmN 3021 [SECURITY] Choose order of LSM initialization. This 3022 overrides CONFIG_LSM, and the "security=" parameter. 3023 3024 machvec= [IA-64] Force the use of a particular machine-vector 3025 (machvec) in a generic kernel. 3026 Example: machvec=hpzx1 3027 3028 machtype= [Loongson] Share the same kernel image file between 3029 different yeeloong laptops. 3030 Example: machtype=lemote-yeeloong-2f-7inch 3031 3032 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater 3033 than or equal to this physical address is ignored. 3034 3035 maxcpus= [SMP] Maximum number of processors that an SMP kernel 3036 will bring up during bootup. maxcpus=n : n >= 0 limits 3037 the kernel to bring up 'n' processors. Surely after 3038 bootup you can bring up the other plugged cpu by executing 3039 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3040 only takes effect during system bootup. 3041 While n=0 is a special case, it is equivalent to "nosmp", 3042 which also disables the IO APIC. 3043 3044 max_loop= [LOOP] The number of loop block devices that get 3045 (loop.max_loop) unconditionally pre-created at init time. The default 3046 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3047 of statically allocating a predefined number, loop 3048 devices can be requested on-demand with the 3049 /dev/loop-control interface. 3050 3051 mce [X86-32] Machine Check Exception 3052 3053 mce=option [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 3054 3055 md= [HW] RAID subsystems devices and level 3056 See Documentation/admin-guide/md.rst. 3057 3058 mdacon= [MDA] 3059 Format: <first>,<last> 3060 Specifies range of consoles to be captured by the MDA. 3061 3062 mds= [X86,INTEL] 3063 Control mitigation for the Micro-architectural Data 3064 Sampling (MDS) vulnerability. 3065 3066 Certain CPUs are vulnerable to an exploit against CPU 3067 internal buffers which can forward information to a 3068 disclosure gadget under certain conditions. 3069 3070 In vulnerable processors, the speculatively 3071 forwarded data can be used in a cache side channel 3072 attack, to access data to which the attacker does 3073 not have direct access. 3074 3075 This parameter controls the MDS mitigation. The 3076 options are: 3077 3078 full - Enable MDS mitigation on vulnerable CPUs 3079 full,nosmt - Enable MDS mitigation and disable 3080 SMT on vulnerable CPUs 3081 off - Unconditionally disable MDS mitigation 3082 3083 On TAA-affected machines, mds=off can be prevented by 3084 an active TAA mitigation as both vulnerabilities are 3085 mitigated with the same mechanism so in order to disable 3086 this mitigation, you need to specify tsx_async_abort=off 3087 too. 3088 3089 Not specifying this option is equivalent to 3090 mds=full. 3091 3092 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3093 3094 mem=nn[KMG] [HEXAGON] Set the memory size. 3095 Must be specified, otherwise memory size will be 0. 3096 3097 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 3098 Amount of memory to be used in cases as follows: 3099 3100 1 for test; 3101 2 when the kernel is not able to see the whole system memory; 3102 3 memory that lies after 'mem=' boundary is excluded from 3103 the hypervisor, then assigned to KVM guests. 3104 4 to limit the memory available for kdump kernel. 3105 3106 [ARC,MICROBLAZE] - the limit applies only to low memory, 3107 high memory is not affected. 3108 3109 [ARM64] - only limits memory covered by the linear 3110 mapping. The NOMAP regions are not affected. 3111 3112 [X86] Work as limiting max address. Use together 3113 with memmap= to avoid physical address space collisions. 3114 Without memmap= PCI devices could be placed at addresses 3115 belonging to unused RAM. 3116 3117 Note that this only takes effects during boot time since 3118 in above case 3, memory may need be hot added after boot 3119 if system memory of hypervisor is not sufficient. 3120 3121 mem=nn[KMG]@ss[KMG] 3122 [ARM,MIPS] - override the memory layout reported by 3123 firmware. 3124 Define a memory region of size nn[KMG] starting at 3125 ss[KMG]. 3126 Multiple different regions can be specified with 3127 multiple mem= parameters on the command line. 3128 3129 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3130 memory. 3131 3132 memblock=debug [KNL] Enable memblock debug messages. 3133 3134 memchunk=nn[KMG] 3135 [KNL,SH] Allow user to override the default size for 3136 per-device physically contiguous DMA buffers. 3137 3138 memhp_default_state=online/offline/online_kernel/online_movable 3139 [KNL] Set the initial state for the memory hotplug 3140 onlining policy. If not specified, the default value is 3141 set according to the 3142 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 3143 option. 3144 See Documentation/admin-guide/mm/memory-hotplug.rst. 3145 3146 memmap=exactmap [KNL,X86] Enable setting of an exact 3147 E820 memory map, as specified by the user. 3148 Such memmap=exactmap lines can be constructed based on 3149 BIOS output or other requirements. See the memmap=nn@ss 3150 option description. 3151 3152 memmap=nn[KMG]@ss[KMG] 3153 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory. 3154 Region of memory to be used is from ss to ss+nn. 3155 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3156 which limits max address to nn[KMG]. 3157 Multiple different regions can be specified, 3158 comma delimited. 3159 Example: 3160 memmap=100M@2G,100M#3G,1G!1024G 3161 3162 memmap=nn[KMG]#ss[KMG] 3163 [KNL,ACPI] Mark specific memory as ACPI data. 3164 Region of memory to be marked is from ss to ss+nn. 3165 3166 memmap=nn[KMG]$ss[KMG] 3167 [KNL,ACPI] Mark specific memory as reserved. 3168 Region of memory to be reserved is from ss to ss+nn. 3169 Example: Exclude memory from 0x18690000-0x1869ffff 3170 memmap=64K$0x18690000 3171 or 3172 memmap=0x10000$0x18690000 3173 Some bootloaders may need an escape character before '$', 3174 like Grub2, otherwise '$' and the following number 3175 will be eaten. 3176 3177 memmap=nn[KMG]!ss[KMG] 3178 [KNL,X86] Mark specific memory as protected. 3179 Region of memory to be used, from ss to ss+nn. 3180 The memory region may be marked as e820 type 12 (0xc) 3181 and is NVDIMM or ADR memory. 3182 3183 memmap=<size>%<offset>-<oldtype>+<newtype> 3184 [KNL,ACPI] Convert memory within the specified region 3185 from <oldtype> to <newtype>. If "-<oldtype>" is left 3186 out, the whole region will be marked as <newtype>, 3187 even if previously unavailable. If "+<newtype>" is left 3188 out, matching memory will be removed. Types are 3189 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3190 3 = ACPI, 12 = PRAM. 3191 3192 memory_corruption_check=0/1 [X86] 3193 Some BIOSes seem to corrupt the first 64k of 3194 memory when doing things like suspend/resume. 3195 Setting this option will scan the memory 3196 looking for corruption. Enabling this will 3197 both detect corruption and prevent the kernel 3198 from using the memory being corrupted. 3199 However, its intended as a diagnostic tool; if 3200 repeatable BIOS-originated corruption always 3201 affects the same memory, you can use memmap= 3202 to prevent the kernel from using that memory. 3203 3204 memory_corruption_check_size=size [X86] 3205 By default it checks for corruption in the low 3206 64k, making this memory unavailable for normal 3207 use. Use this parameter to scan for 3208 corruption in more or less memory. 3209 3210 memory_corruption_check_period=seconds [X86] 3211 By default it checks for corruption every 60 3212 seconds. Use this parameter to check at some 3213 other rate. 0 disables periodic checking. 3214 3215 memory_hotplug.memmap_on_memory 3216 [KNL,X86,ARM] Boolean flag to enable this feature. 3217 Format: {on | off (default)} 3218 When enabled, runtime hotplugged memory will 3219 allocate its internal metadata (struct pages, 3220 those vmemmap pages cannot be optimized even 3221 if hugetlb_free_vmemmap is enabled) from the 3222 hotadded memory which will allow to hotadd a 3223 lot of memory without requiring additional 3224 memory to do so. 3225 This feature is disabled by default because it 3226 has some implication on large (e.g. GB) 3227 allocations in some configurations (e.g. small 3228 memory blocks). 3229 The state of the flag can be read in 3230 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3231 Note that even when enabled, there are a few cases where 3232 the feature is not effective. 3233 3234 memtest= [KNL,X86,ARM,M68K,PPC,RISCV] Enable memtest 3235 Format: <integer> 3236 default : 0 <disable> 3237 Specifies the number of memtest passes to be 3238 performed. Each pass selects another test 3239 pattern from a given set of patterns. Memtest 3240 fills the memory with this pattern, validates 3241 memory contents and reserves bad memory 3242 regions that are detected. 3243 3244 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3245 Valid arguments: on, off 3246 Default (depends on kernel configuration option): 3247 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 3248 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 3249 mem_encrypt=on: Activate SME 3250 mem_encrypt=off: Do not activate SME 3251 3252 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3253 for details on when memory encryption can be activated. 3254 3255 mem_sleep_default= [SUSPEND] Default system suspend mode: 3256 s2idle - Suspend-To-Idle 3257 shallow - Power-On Suspend or equivalent (if supported) 3258 deep - Suspend-To-RAM or equivalent (if supported) 3259 See Documentation/admin-guide/pm/sleep-states.rst. 3260 3261 mfgpt_irq= [IA-32] Specify the IRQ to use for the 3262 Multi-Function General Purpose Timers on AMD Geode 3263 platforms. 3264 3265 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3266 the BIOS has incorrectly applied a workaround. TinyBIOS 3267 version 0.98 is known to be affected, 0.99 fixes the 3268 problem by letting the user disable the workaround. 3269 3270 mga= [HW,DRM] 3271 3272 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this 3273 physical address is ignored. 3274 3275 mini2440= [ARM,HW,KNL] 3276 Format:[0..2][b][c][t] 3277 Default: "0tb" 3278 MINI2440 configuration specification: 3279 0 - The attached screen is the 3.5" TFT 3280 1 - The attached screen is the 7" TFT 3281 2 - The VGA Shield is attached (1024x768) 3282 Leaving out the screen size parameter will not load 3283 the TFT driver, and the framebuffer will be left 3284 unconfigured. 3285 b - Enable backlight. The TFT backlight pin will be 3286 linked to the kernel VESA blanking code and a GPIO 3287 LED. This parameter is not necessary when using the 3288 VGA shield. 3289 c - Enable the s3c camera interface. 3290 t - Reserved for enabling touchscreen support. The 3291 touchscreen support is not enabled in the mainstream 3292 kernel as of 2.6.30, a preliminary port can be found 3293 in the "bleeding edge" mini2440 support kernel at 3294 https://repo.or.cz/w/linux-2.6/mini2440.git 3295 3296 mitigations= 3297 [X86,PPC,S390,ARM64] Control optional mitigations for 3298 CPU vulnerabilities. This is a set of curated, 3299 arch-independent options, each of which is an 3300 aggregation of existing arch-specific options. 3301 3302 off 3303 Disable all optional CPU mitigations. This 3304 improves system performance, but it may also 3305 expose users to several CPU vulnerabilities. 3306 Equivalent to: if nokaslr then kpti=0 [ARM64] 3307 gather_data_sampling=off [X86] 3308 kvm.nx_huge_pages=off [X86] 3309 l1tf=off [X86] 3310 mds=off [X86] 3311 mmio_stale_data=off [X86] 3312 no_entry_flush [PPC] 3313 no_uaccess_flush [PPC] 3314 nobp=0 [S390] 3315 nopti [X86,PPC] 3316 nospectre_bhb [ARM64] 3317 nospectre_v1 [X86,PPC] 3318 nospectre_v2 [X86,PPC,S390,ARM64] 3319 retbleed=off [X86] 3320 spec_store_bypass_disable=off [X86,PPC] 3321 spectre_v2_user=off [X86] 3322 srbds=off [X86,INTEL] 3323 ssbd=force-off [ARM64] 3324 tsx_async_abort=off [X86] 3325 3326 Exceptions: 3327 This does not have any effect on 3328 kvm.nx_huge_pages when 3329 kvm.nx_huge_pages=force. 3330 3331 auto (default) 3332 Mitigate all CPU vulnerabilities, but leave SMT 3333 enabled, even if it's vulnerable. This is for 3334 users who don't want to be surprised by SMT 3335 getting disabled across kernel upgrades, or who 3336 have other ways of avoiding SMT-based attacks. 3337 Equivalent to: (default behavior) 3338 3339 auto,nosmt 3340 Mitigate all CPU vulnerabilities, disabling SMT 3341 if needed. This is for users who always want to 3342 be fully mitigated, even if it means losing SMT. 3343 Equivalent to: l1tf=flush,nosmt [X86] 3344 mds=full,nosmt [X86] 3345 tsx_async_abort=full,nosmt [X86] 3346 mmio_stale_data=full,nosmt [X86] 3347 retbleed=auto,nosmt [X86] 3348 3349 mminit_loglevel= 3350 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 3351 parameter allows control of the logging verbosity for 3352 the additional memory initialisation checks. A value 3353 of 0 disables mminit logging and a level of 4 will 3354 log everything. Information is printed at KERN_DEBUG 3355 so loglevel=8 may also need to be specified. 3356 3357 mmio_stale_data= 3358 [X86,INTEL] Control mitigation for the Processor 3359 MMIO Stale Data vulnerabilities. 3360 3361 Processor MMIO Stale Data is a class of 3362 vulnerabilities that may expose data after an MMIO 3363 operation. Exposed data could originate or end in 3364 the same CPU buffers as affected by MDS and TAA. 3365 Therefore, similar to MDS and TAA, the mitigation 3366 is to clear the affected CPU buffers. 3367 3368 This parameter controls the mitigation. The 3369 options are: 3370 3371 full - Enable mitigation on vulnerable CPUs 3372 3373 full,nosmt - Enable mitigation and disable SMT on 3374 vulnerable CPUs. 3375 3376 off - Unconditionally disable mitigation 3377 3378 On MDS or TAA affected machines, 3379 mmio_stale_data=off can be prevented by an active 3380 MDS or TAA mitigation as these vulnerabilities are 3381 mitigated with the same mechanism so in order to 3382 disable this mitigation, you need to specify 3383 mds=off and tsx_async_abort=off too. 3384 3385 Not specifying this option is equivalent to 3386 mmio_stale_data=full. 3387 3388 For details see: 3389 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3390 3391 <module>.async_probe[=<bool>] [KNL] 3392 If no <bool> value is specified or if the value 3393 specified is not a valid <bool>, enable asynchronous 3394 probe on this module. Otherwise, enable/disable 3395 asynchronous probe on this module as indicated by the 3396 <bool> value. See also: module.async_probe 3397 3398 module.async_probe=<bool> 3399 [KNL] When set to true, modules will use async probing 3400 by default. To enable/disable async probing for a 3401 specific module, use the module specific control that 3402 is documented under <module>.async_probe. When both 3403 module.async_probe and <module>.async_probe are 3404 specified, <module>.async_probe takes precedence for 3405 the specific module. 3406 3407 module.enable_dups_trace 3408 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 3409 this means that duplicate request_module() calls will 3410 trigger a WARN_ON() instead of a pr_warn(). Note that 3411 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 3412 will always be issued and this option does nothing. 3413 module.sig_enforce 3414 [KNL] When CONFIG_MODULE_SIG is set, this means that 3415 modules without (valid) signatures will fail to load. 3416 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3417 is always true, so this option does nothing. 3418 3419 module_blacklist= [KNL] Do not load a comma-separated list of 3420 modules. Useful for debugging problem modules. 3421 3422 mousedev.tap_time= 3423 [MOUSE] Maximum time between finger touching and 3424 leaving touchpad surface for touch to be considered 3425 a tap and be reported as a left button click (for 3426 touchpads working in absolute mode only). 3427 Format: <msecs> 3428 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 3429 reporting absolute coordinates, such as tablets 3430 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 3431 reporting absolute coordinates, such as tablets 3432 3433 movablecore= [KNL,X86,IA-64,PPC] 3434 Format: nn[KMGTPE] | nn% 3435 This parameter is the complement to kernelcore=, it 3436 specifies the amount of memory used for migratable 3437 allocations. If both kernelcore and movablecore is 3438 specified, then kernelcore will be at *least* the 3439 specified value but may be more. If movablecore on its 3440 own is specified, the administrator must be careful 3441 that the amount of memory usable for all allocations 3442 is not too small. 3443 3444 movable_node [KNL] Boot-time switch to make hotplugable memory 3445 NUMA nodes to be movable. This means that the memory 3446 of such nodes will be usable only for movable 3447 allocations which rules out almost all kernel 3448 allocations. Use with caution! 3449 3450 MTD_Partition= [MTD] 3451 Format: <name>,<region-number>,<size>,<offset> 3452 3453 MTD_Region= [MTD] Format: 3454 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 3455 3456 mtdparts= [MTD] 3457 See drivers/mtd/parsers/cmdlinepart.c 3458 3459 mtdset= [ARM] 3460 ARM/S3C2412 JIVE boot control 3461 3462 See arch/arm/mach-s3c/mach-jive.c 3463 3464 mtouchusb.raw_coordinates= 3465 [HW] Make the MicroTouch USB driver use raw coordinates 3466 ('y', default) or cooked coordinates ('n') 3467 3468 mtrr=debug [X86] 3469 Enable printing debug information related to MTRR 3470 registers at boot time. 3471 3472 mtrr_chunk_size=nn[KMG] [X86] 3473 used for mtrr cleanup. It is largest continuous chunk 3474 that could hold holes aka. UC entries. 3475 3476 mtrr_gran_size=nn[KMG] [X86] 3477 Used for mtrr cleanup. It is granularity of mtrr block. 3478 Default is 1. 3479 Large value could prevent small alignment from 3480 using up MTRRs. 3481 3482 mtrr_spare_reg_nr=n [X86] 3483 Format: <integer> 3484 Range: 0,7 : spare reg number 3485 Default : 1 3486 Used for mtrr cleanup. It is spare mtrr entries number. 3487 Set to 2 or more if your graphical card needs more. 3488 3489 multitce=off [PPC] This parameter disables the use of the pSeries 3490 firmware feature for updating multiple TCE entries 3491 at a time. 3492 3493 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 3494 3495 netdev= [NET] Network devices parameters 3496 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 3497 Note that mem_start is often overloaded to mean 3498 something different and driver-specific. 3499 This usage is only documented in each driver source 3500 file if at all. 3501 3502 netpoll.carrier_timeout= 3503 [NET] Specifies amount of time (in seconds) that 3504 netpoll should wait for a carrier. By default netpoll 3505 waits 4 seconds. 3506 3507 nf_conntrack.acct= 3508 [NETFILTER] Enable connection tracking flow accounting 3509 0 to disable accounting 3510 1 to enable accounting 3511 Default value is 0. 3512 3513 nfs.cache_getent= 3514 [NFS] sets the pathname to the program which is used 3515 to update the NFS client cache entries. 3516 3517 nfs.cache_getent_timeout= 3518 [NFS] sets the timeout after which an attempt to 3519 update a cache entry is deemed to have failed. 3520 3521 nfs.callback_nr_threads= 3522 [NFSv4] set the total number of threads that the 3523 NFS client will assign to service NFSv4 callback 3524 requests. 3525 3526 nfs.callback_tcpport= 3527 [NFS] set the TCP port on which the NFSv4 callback 3528 channel should listen. 3529 3530 nfs.enable_ino64= 3531 [NFS] enable 64-bit inode numbers. 3532 If zero, the NFS client will fake up a 32-bit inode 3533 number for the readdir() and stat() syscalls instead 3534 of returning the full 64-bit number. 3535 The default is to return 64-bit inode numbers. 3536 3537 nfs.idmap_cache_timeout= 3538 [NFS] set the maximum lifetime for idmapper cache 3539 entries. 3540 3541 nfs.max_session_cb_slots= 3542 [NFSv4.1] Sets the maximum number of session 3543 slots the client will assign to the callback 3544 channel. This determines the maximum number of 3545 callbacks the client will process in parallel for 3546 a particular server. 3547 3548 nfs.max_session_slots= 3549 [NFSv4.1] Sets the maximum number of session slots 3550 the client will attempt to negotiate with the server. 3551 This limits the number of simultaneous RPC requests 3552 that the client can send to the NFSv4.1 server. 3553 Note that there is little point in setting this 3554 value higher than the max_tcp_slot_table_limit. 3555 3556 nfs.nfs4_disable_idmapping= 3557 [NFSv4] When set to the default of '1', this option 3558 ensures that both the RPC level authentication 3559 scheme and the NFS level operations agree to use 3560 numeric uids/gids if the mount is using the 3561 'sec=sys' security flavour. In effect it is 3562 disabling idmapping, which can make migration from 3563 legacy NFSv2/v3 systems to NFSv4 easier. 3564 Servers that do not support this mode of operation 3565 will be autodetected by the client, and it will fall 3566 back to using the idmapper. 3567 To turn off this behaviour, set the value to '0'. 3568 3569 nfs.nfs4_unique_id= 3570 [NFS4] Specify an additional fixed unique ident- 3571 ification string that NFSv4 clients can insert into 3572 their nfs_client_id4 string. This is typically a 3573 UUID that is generated at system install time. 3574 3575 nfs.recover_lost_locks= 3576 [NFSv4] Attempt to recover locks that were lost due 3577 to a lease timeout on the server. Please note that 3578 doing this risks data corruption, since there are 3579 no guarantees that the file will remain unchanged 3580 after the locks are lost. 3581 If you want to enable the kernel legacy behaviour of 3582 attempting to recover these locks, then set this 3583 parameter to '1'. 3584 The default parameter value of '0' causes the kernel 3585 not to attempt recovery of lost locks. 3586 3587 nfs.send_implementation_id= 3588 [NFSv4.1] Send client implementation identification 3589 information in exchange_id requests. 3590 If zero, no implementation identification information 3591 will be sent. 3592 The default is to send the implementation identification 3593 information. 3594 3595 nfs4.layoutstats_timer= 3596 [NFSv4.2] Change the rate at which the kernel sends 3597 layoutstats to the pNFS metadata server. 3598 3599 Setting this to value to 0 causes the kernel to use 3600 whatever value is the default set by the layout 3601 driver. A non-zero value sets the minimum interval 3602 in seconds between layoutstats transmissions. 3603 3604 nfsd.inter_copy_offload_enable= 3605 [NFSv4.2] When set to 1, the server will support 3606 server-to-server copies for which this server is 3607 the destination of the copy. 3608 3609 nfsd.nfs4_disable_idmapping= 3610 [NFSv4] When set to the default of '1', the NFSv4 3611 server will return only numeric uids and gids to 3612 clients using auth_sys, and will accept numeric uids 3613 and gids from such clients. This is intended to ease 3614 migration from NFSv2/v3. 3615 3616 nfsd.nfsd4_ssc_umount_timeout= 3617 [NFSv4.2] When used as the destination of a 3618 server-to-server copy, knfsd temporarily mounts 3619 the source server. It caches the mount in case 3620 it will be needed again, and discards it if not 3621 used for the number of milliseconds specified by 3622 this parameter. 3623 3624 nfsaddrs= [NFS] Deprecated. Use ip= instead. 3625 See Documentation/admin-guide/nfs/nfsroot.rst. 3626 3627 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 3628 See Documentation/admin-guide/nfs/nfsroot.rst. 3629 3630 nfsrootdebug [NFS] enable nfsroot debugging messages. 3631 See Documentation/admin-guide/nfs/nfsroot.rst. 3632 3633 nmi_backtrace.backtrace_idle [KNL] 3634 Dump stacks even of idle CPUs in response to an 3635 NMI stack-backtrace request. 3636 3637 nmi_debug= [KNL,SH] Specify one or more actions to take 3638 when a NMI is triggered. 3639 Format: [state][,regs][,debounce][,die] 3640 3641 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3642 Format: [panic,][nopanic,][num] 3643 Valid num: 0 or 1 3644 0 - turn hardlockup detector in nmi_watchdog off 3645 1 - turn hardlockup detector in nmi_watchdog on 3646 When panic is specified, panic when an NMI watchdog 3647 timeout occurs (or 'nopanic' to not panic on an NMI 3648 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3649 To disable both hard and soft lockup detectors, 3650 please see 'nowatchdog'. 3651 This is useful when you use a panic=... timeout and 3652 need the box quickly up again. 3653 3654 These settings can be accessed at runtime via 3655 the nmi_watchdog and hardlockup_panic sysctls. 3656 3657 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3658 emulation library even if a 387 maths coprocessor 3659 is present. 3660 3661 no4lvl [RISCV] Disable 4-level and 5-level paging modes. Forces 3662 kernel to use 3-level paging instead. 3663 3664 no5lvl [X86-64,RISCV] Disable 5-level paging mode. Forces 3665 kernel to use 4-level paging instead. 3666 3667 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3668 caches in the slab allocator. Saves per-node memory, 3669 but will impact performance. 3670 3671 noalign [KNL,ARM] 3672 3673 noaltinstr [S390] Disables alternative instructions patching 3674 (CPU alternatives feature). 3675 3676 noapic [SMP,APIC] Tells the kernel to not make use of any 3677 IOAPICs that may be present in the system. 3678 3679 noautogroup Disable scheduler automatic task group creation. 3680 3681 nocache [ARM] 3682 3683 no_console_suspend 3684 [HW] Never suspend the console 3685 Disable suspending of consoles during suspend and 3686 hibernate operations. Once disabled, debugging 3687 messages can reach various consoles while the rest 3688 of the system is being put to sleep (ie, while 3689 debugging driver suspend/resume hooks). This may 3690 not work reliably with all consoles, but is known 3691 to work with serial and VGA consoles. 3692 To facilitate more flexible debugging, we also add 3693 console_suspend, a printk module parameter to control 3694 it. Users could use console_suspend (usually 3695 /sys/module/printk/parameters/console_suspend) to 3696 turn on/off it dynamically. 3697 3698 no_debug_objects 3699 [KNL] Disable object debugging 3700 3701 nodsp [SH] Disable hardware DSP at boot time. 3702 3703 noefi Disable EFI runtime services support. 3704 3705 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel. 3706 3707 noexec [IA-64] 3708 3709 noexec32 [X86-64] 3710 This affects only 32-bit executables. 3711 noexec32=on: enable non-executable mappings (default) 3712 read doesn't imply executable mappings 3713 noexec32=off: disable non-executable mappings 3714 read implies executable mappings 3715 3716 no_file_caps Tells the kernel not to honor file capabilities. The 3717 only way then for a file to be executed with privilege 3718 is to be setuid root or executed by root. 3719 3720 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3721 3722 nofsgsbase [X86] Disables FSGSBASE instructions. 3723 3724 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3725 register save and restore. The kernel will only save 3726 legacy floating-point registers on task switch. 3727 3728 nohalt [IA-64] Tells the kernel not to use the power saving 3729 function PAL_HALT_LIGHT when idle. This increases 3730 power-consumption. On the positive side, it reduces 3731 interrupt wake-up latency, which may improve performance 3732 in certain environments such as networked servers or 3733 real-time systems. 3734 3735 no_hash_pointers 3736 Force pointers printed to the console or buffers to be 3737 unhashed. By default, when a pointer is printed via %p 3738 format string, that pointer is "hashed", i.e. obscured 3739 by hashing the pointer value. This is a security feature 3740 that hides actual kernel addresses from unprivileged 3741 users, but it also makes debugging the kernel more 3742 difficult since unequal pointers can no longer be 3743 compared. However, if this command-line option is 3744 specified, then all normal pointers will have their true 3745 value printed. This option should only be specified when 3746 debugging the kernel. Please do not use on production 3747 kernels. 3748 3749 nohibernate [HIBERNATION] Disable hibernation and resume. 3750 3751 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,SH] Forces the kernel to 3752 busy wait in do_idle() and not use the arch_cpu_idle() 3753 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 3754 to be effective. This is useful on platforms where the 3755 sleep(SH) or wfi(ARM,ARM64) instructions do not work 3756 correctly or when doing power measurements to evaluate 3757 the impact of the sleep instructions. This is also 3758 useful when using JTAG debugger. 3759 3760 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings. 3761 3762 nohugevmalloc [KNL,X86,PPC,ARM64] Disable kernel huge vmalloc mappings. 3763 3764 nohz= [KNL] Boottime enable/disable dynamic ticks 3765 Valid arguments: on, off 3766 Default: on 3767 3768 nohz_full= [KNL,BOOT,SMP,ISOL] 3769 The argument is a cpu list, as described above. 3770 In kernels built with CONFIG_NO_HZ_FULL=y, set 3771 the specified list of CPUs whose tick will be stopped 3772 whenever possible. The boot CPU will be forced outside 3773 the range to maintain the timekeeping. Any CPUs 3774 in this list will have their RCU callbacks offloaded, 3775 just as if they had also been called out in the 3776 rcu_nocbs= boot parameter. 3777 3778 Note that this argument takes precedence over 3779 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 3780 3781 noinitrd [RAM] Tells the kernel not to load any configured 3782 initial RAM disk. 3783 3784 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3785 remapping. 3786 [Deprecated - use intremap=off] 3787 3788 nointroute [IA-64] 3789 3790 noinvpcid [X86] Disable the INVPCID cpu feature. 3791 3792 noiotrap [SH] Disables trapped I/O port accesses. 3793 3794 noirqdebug [X86-32] Disables the code which attempts to detect and 3795 disable unhandled interrupt sources. 3796 3797 noisapnp [ISAPNP] Disables ISA PnP code. 3798 3799 nojitter [IA-64] Disables jitter checking for ITC timers. 3800 3801 nokaslr [KNL] 3802 When CONFIG_RANDOMIZE_BASE is set, this disables 3803 kernel and module base offset ASLR (Address Space 3804 Layout Randomization). 3805 3806 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3807 fault handling. 3808 3809 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3810 3811 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3812 3813 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3814 3815 nomca [IA-64] Disable machine check abort handling 3816 3817 nomce [X86-32] Disable Machine Check Exception 3818 3819 nomfgpt [X86-32] Disable Multi-Function General Purpose 3820 Timer usage (for AMD Geode machines). 3821 3822 nomodeset Disable kernel modesetting. Most systems' firmware 3823 sets up a display mode and provides framebuffer memory 3824 for output. With nomodeset, DRM and fbdev drivers will 3825 not load if they could possibly displace the pre- 3826 initialized output. Only the system framebuffer will 3827 be available for use. The respective drivers will not 3828 perform display-mode changes or accelerated rendering. 3829 3830 Useful as error fallback, or for testing and debugging. 3831 3832 nomodule Disable module load 3833 3834 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3835 shutdown the other cpus. Instead use the REBOOT_VECTOR 3836 irq. 3837 3838 nopat [X86] Disable PAT (page attribute table extension of 3839 pagetables) support. 3840 3841 nopcid [X86-64] Disable the PCID cpu feature. 3842 3843 nopku [X86] Disable Memory Protection Keys CPU feature found 3844 in some Intel CPUs. 3845 3846 nopti [X86-64] 3847 Equivalent to pti=off 3848 3849 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 3850 Disables the PV optimizations forcing the guest to run 3851 as generic guest with no PV drivers. Currently support 3852 XEN HVM, KVM, HYPER_V and VMWARE guest. 3853 3854 nopvspin [X86,XEN,KVM] 3855 Disables the qspinlock slow path using PV optimizations 3856 which allow the hypervisor to 'idle' the guest on lock 3857 contention. 3858 3859 norandmaps Don't use address space randomization. Equivalent to 3860 echo 0 > /proc/sys/kernel/randomize_va_space 3861 3862 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3863 with UP alternatives 3864 3865 noresume [SWSUSP] Disables resume and restores original swap 3866 space. 3867 3868 nosbagart [IA-64] 3869 3870 no-scroll [VGA] Disables scrollback. 3871 This is required for the Braillex ib80-piezo Braille 3872 reader made by F.H. Papenmeier (Germany). 3873 3874 nosgx [X86-64,SGX] Disables Intel SGX kernel support. 3875 3876 nosmap [PPC] 3877 Disable SMAP (Supervisor Mode Access Prevention) 3878 even if it is supported by processor. 3879 3880 nosmep [PPC64s] 3881 Disable SMEP (Supervisor Mode Execution Prevention) 3882 even if it is supported by processor. 3883 3884 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3885 and disable the IO APIC. legacy for "maxcpus=0". 3886 3887 nosmt [KNL,MIPS,PPC,S390] Disable symmetric multithreading (SMT). 3888 Equivalent to smt=1. 3889 3890 [KNL,X86,PPC] Disable symmetric multithreading (SMT). 3891 nosmt=force: Force disable SMT, cannot be undone 3892 via the sysfs control file. 3893 3894 nosoftlockup [KNL] Disable the soft-lockup detector. 3895 3896 nospec_store_bypass_disable 3897 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3898 3899 nospectre_bhb [ARM64] Disable all mitigations for Spectre-BHB (branch 3900 history injection) vulnerability. System may allow data leaks 3901 with this option. 3902 3903 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3904 (bounds check bypass). With this option data leaks are 3905 possible in the system. 3906 3907 nospectre_v2 [X86,PPC_E500,ARM64] Disable all mitigations for 3908 the Spectre variant 2 (indirect branch prediction) 3909 vulnerability. System may allow data leaks with this 3910 option. 3911 3912 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES] Disable paravirtualized 3913 steal time accounting. steal time is computed, but 3914 won't influence scheduler behaviour 3915 3916 nosync [HW,M68K] Disables sync negotiation for all devices. 3917 3918 no_timer_check [X86,APIC] Disables the code which tests for 3919 broken timer IRQ sources. 3920 3921 no_uaccess_flush 3922 [PPC] Don't flush the L1-D cache after accessing user data. 3923 3924 novmcoredd [KNL,KDUMP] 3925 Disable device dump. Device dump allows drivers to 3926 append dump data to vmcore so you can collect driver 3927 specified debug info. Drivers can append the data 3928 without any limit and this data is stored in memory, 3929 so this may cause significant memory stress. Disabling 3930 device dump can help save memory but the driver debug 3931 data will be no longer available. This parameter 3932 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3933 is set. 3934 3935 no-vmw-sched-clock 3936 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3937 clock and use the default one. 3938 3939 nowatchdog [KNL] Disable both lockup detectors, i.e. 3940 soft-lockup and NMI watchdog (hard-lockup). 3941 3942 nowb [ARM] 3943 3944 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3945 3946 NOTE: this parameter will be ignored on systems with the 3947 LEGACY_XAPIC_DISABLED bit set in the 3948 IA32_XAPIC_DISABLE_STATUS MSR. 3949 3950 noxsave [BUGS=X86] Disables x86 extended register state save 3951 and restore using xsave. The kernel will fallback to 3952 enabling legacy floating-point and sse state. 3953 3954 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3955 register states. The kernel will fall back to use 3956 xsave to save the states. By using this parameter, 3957 performance of saving the states is degraded because 3958 xsave doesn't support modified optimization while 3959 xsaveopt supports it on xsaveopt enabled systems. 3960 3961 noxsaves [X86] Disables xsaves and xrstors used in saving and 3962 restoring x86 extended register state in compacted 3963 form of xsave area. The kernel will fall back to use 3964 xsaveopt and xrstor to save and restore the states 3965 in standard form of xsave area. By using this 3966 parameter, xsave area per process might occupy more 3967 memory on xsaves enabled systems. 3968 3969 nps_mtm_hs_ctr= [KNL,ARC] 3970 This parameter sets the maximum duration, in 3971 cycles, each HW thread of the CTOP can run 3972 without interruptions, before HW switches it. 3973 The actual maximum duration is 16 times this 3974 parameter's value. 3975 Format: integer between 1 and 255 3976 Default: 255 3977 3978 nptcg= [IA-64] Override max number of concurrent global TLB 3979 purges which is reported from either PAL_VM_SUMMARY or 3980 SAL PALO. 3981 3982 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3983 could support. nr_cpus=n : n >= 1 limits the kernel to 3984 support 'n' processors. It could be larger than the 3985 number of already plugged CPU during bootup, later in 3986 runtime you can physically add extra cpu until it reaches 3987 n. So during boot up some boot time memory for per-cpu 3988 variables need be pre-allocated for later physical cpu 3989 hot plugging. 3990 3991 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3992 3993 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only 3994 set up a single NUMA node spanning all memory. 3995 3996 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 3997 NUMA balancing. 3998 Allowed values are enable and disable 3999 4000 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4001 'node', 'default' can be specified 4002 This can be set from sysctl after boot. 4003 See Documentation/admin-guide/sysctl/vm.rst for details. 4004 4005 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 4006 See Documentation/core-api/debugging-via-ohci1394.rst for more 4007 info. 4008 4009 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4010 Rather than timing out after 20 ms if an EC 4011 command is not properly ACKed, override the length 4012 of the timeout. We have interrupts disabled while 4013 waiting for the ACK, so if this is set too high 4014 interrupts *may* be lost! 4015 4016 omap_mux= [OMAP] Override bootloader pin multiplexing. 4017 Format: <mux_mode0.mode_name=value>... 4018 For example, to override I2C bus2: 4019 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4020 4021 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4022 4023 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4024 4025 boundary - index of last SLC block on Flex-OneNAND. 4026 The remaining blocks are configured as MLC blocks. 4027 lock - Configure if Flex-OneNAND boundary should be locked. 4028 Once locked, the boundary cannot be changed. 4029 1 indicates lock status, 0 indicates unlock status. 4030 4031 oops=panic Always panic on oopses. Default is to just kill the 4032 process, but there is a small probability of 4033 deadlocking the machine. 4034 This will also cause panics on machine check exceptions. 4035 Useful together with panic=30 to trigger a reboot. 4036 4037 page_alloc.shuffle= 4038 [KNL] Boolean flag to control whether the page allocator 4039 should randomize its free lists. The randomization may 4040 be automatically enabled if the kernel detects it is 4041 running on a platform with a direct-mapped memory-side 4042 cache, and this parameter can be used to 4043 override/disable that behavior. The state of the flag 4044 can be read from sysfs at: 4045 /sys/module/page_alloc/parameters/shuffle. 4046 4047 page_owner= [KNL] Boot-time page_owner enabling option. 4048 Storage of the information about who allocated 4049 each page is disabled in default. With this switch, 4050 we can turn it on. 4051 on: enable the feature 4052 4053 page_poison= [KNL] Boot-time parameter changing the state of 4054 poisoning on the buddy allocator, available with 4055 CONFIG_PAGE_POISONING=y. 4056 off: turn off poisoning (default) 4057 on: turn on poisoning 4058 4059 page_reporting.page_reporting_order= 4060 [KNL] Minimal page reporting order 4061 Format: <integer> 4062 Adjust the minimal page reporting order. The page 4063 reporting is disabled when it exceeds MAX_ORDER. 4064 4065 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4066 timeout > 0: seconds before rebooting 4067 timeout = 0: wait forever 4068 timeout < 0: reboot immediately 4069 Format: <timeout> 4070 4071 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 4072 Format: <hex>[,nousertaint] 4073 Hexadecimal bitmask representing the set of TAINT flags 4074 that will cause the kernel to panic when add_taint() is 4075 called with any of the flags in this set. 4076 The optional switch "nousertaint" can be utilized to 4077 prevent userspace forced crashes by writing to sysctl 4078 /proc/sys/kernel/tainted any flagset matching with the 4079 bitmask set on panic_on_taint. 4080 See Documentation/admin-guide/tainted-kernels.rst for 4081 extra details on the taint flags that users can pick 4082 to compose the bitmask to assign to panic_on_taint. 4083 4084 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4085 on a WARN(). 4086 4087 panic_print= Bitmask for printing system info when panic happens. 4088 User can chose combination of the following bits: 4089 bit 0: print all tasks info 4090 bit 1: print system memory info 4091 bit 2: print timer info 4092 bit 3: print locks info if CONFIG_LOCKDEP is on 4093 bit 4: print ftrace buffer 4094 bit 5: print all printk messages in buffer 4095 bit 6: print all CPUs backtrace (if available in the arch) 4096 *Be aware* that this option may print a _lot_ of lines, 4097 so there are risks of losing older messages in the log. 4098 Use this option carefully, maybe worth to setup a 4099 bigger log buffer with "log_buf_len" along with this. 4100 4101 parkbd.port= [HW] Parallel port number the keyboard adapter is 4102 connected to, default is 0. 4103 Format: <parport#> 4104 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4105 0 for XT, 1 for AT (default is AT). 4106 Format: <mode> 4107 4108 parport= [HW,PPT] Specify parallel ports. 0 disables. 4109 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4110 Use 'auto' to force the driver to use any 4111 IRQ/DMA settings detected (the default is to 4112 ignore detected IRQ/DMA settings because of 4113 possible conflicts). You can specify the base 4114 address, IRQ, and DMA settings; IRQ and DMA 4115 should be numbers, or 'auto' (for using detected 4116 settings on that particular port), or 'nofifo' 4117 (to avoid using a FIFO even if it is detected). 4118 Parallel ports are assigned in the order they 4119 are specified on the command line, starting 4120 with parport0. 4121 4122 parport_init_mode= [HW,PPT] 4123 Configure VIA parallel port to operate in 4124 a specific mode. This is necessary on Pegasos 4125 computer where firmware has no options for setting 4126 up parallel port mode and sets it to spp. 4127 Currently this function knows 686a and 8231 chips. 4128 Format: [spp|ps2|epp|ecp|ecpepp] 4129 4130 pata_legacy.all= [HW,LIBATA] 4131 Format: <int> 4132 Set to non-zero to probe primary and secondary ISA 4133 port ranges on PCI systems where no PCI PATA device 4134 has been found at either range. Disabled by default. 4135 4136 pata_legacy.autospeed= [HW,LIBATA] 4137 Format: <int> 4138 Set to non-zero if a chip is present that snoops speed 4139 changes. Disabled by default. 4140 4141 pata_legacy.ht6560a= [HW,LIBATA] 4142 Format: <int> 4143 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4144 the secondary channel, or both channels respectively. 4145 Disabled by default. 4146 4147 pata_legacy.ht6560b= [HW,LIBATA] 4148 Format: <int> 4149 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4150 the secondary channel, or both channels respectively. 4151 Disabled by default. 4152 4153 pata_legacy.iordy_mask= [HW,LIBATA] 4154 Format: <int> 4155 IORDY enable mask. Set individual bits to allow IORDY 4156 for the respective channel. Bit 0 is for the first 4157 legacy channel handled by this driver, bit 1 is for 4158 the second channel, and so on. The sequence will often 4159 correspond to the primary legacy channel, the secondary 4160 legacy channel, and so on, but the handling of a PCI 4161 bus and the use of other driver options may interfere 4162 with the sequence. By default IORDY is allowed across 4163 all channels. 4164 4165 pata_legacy.opti82c46x= [HW,LIBATA] 4166 Format: <int> 4167 Set to 1, 2, or 3 for Opti 82c611A on the primary 4168 channel, the secondary channel, or both channels 4169 respectively. Disabled by default. 4170 4171 pata_legacy.opti82c611a= [HW,LIBATA] 4172 Format: <int> 4173 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4174 channel, the secondary channel, or both channels 4175 respectively. Disabled by default. 4176 4177 pata_legacy.pio_mask= [HW,LIBATA] 4178 Format: <int> 4179 PIO mode mask for autospeed devices. Set individual 4180 bits to allow the use of the respective PIO modes. 4181 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4182 All modes allowed by default. 4183 4184 pata_legacy.probe_all= [HW,LIBATA] 4185 Format: <int> 4186 Set to non-zero to probe tertiary and further ISA 4187 port ranges on PCI systems. Disabled by default. 4188 4189 pata_legacy.probe_mask= [HW,LIBATA] 4190 Format: <int> 4191 Probe mask for legacy ISA PATA ports. Depending on 4192 platform configuration and the use of other driver 4193 options up to 6 legacy ports are supported: 0x1f0, 4194 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4195 of individual ports can be disabled by setting the 4196 corresponding bits in the mask to 1. Bit 0 is for 4197 the first port in the list above (0x1f0), and so on. 4198 By default all supported ports are probed. 4199 4200 pata_legacy.qdi= [HW,LIBATA] 4201 Format: <int> 4202 Set to non-zero to probe QDI controllers. By default 4203 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4204 4205 pata_legacy.winbond= [HW,LIBATA] 4206 Format: <int> 4207 Set to non-zero to probe Winbond controllers. Use 4208 the standard I/O port (0x130) if 1, otherwise the 4209 value given is the I/O port to use (typically 0x1b0). 4210 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4211 0 otherwise. 4212 4213 pata_platform.pio_mask= [HW,LIBATA] 4214 Format: <int> 4215 Supported PIO mode mask. Set individual bits to allow 4216 the use of the respective PIO modes. Bit 0 is for 4217 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4218 allowed by default. 4219 4220 pause_on_oops=<int> 4221 Halt all CPUs after the first oops has been printed for 4222 the specified number of seconds. This is to be used if 4223 your oopses keep scrolling off the screen. 4224 4225 pcbit= [HW,ISDN] 4226 4227 pci=option[,option...] [PCI] various PCI subsystem options. 4228 4229 Some options herein operate on a specific device 4230 or a set of devices (<pci_dev>). These are 4231 specified in one of the following formats: 4232 4233 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4234 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4235 4236 Note: the first format specifies a PCI 4237 bus/device/function address which may change 4238 if new hardware is inserted, if motherboard 4239 firmware changes, or due to changes caused 4240 by other kernel parameters. If the 4241 domain is left unspecified, it is 4242 taken to be zero. Optionally, a path 4243 to a device through multiple device/function 4244 addresses can be specified after the base 4245 address (this is more robust against 4246 renumbering issues). The second format 4247 selects devices using IDs from the 4248 configuration space which may match multiple 4249 devices in the system. 4250 4251 earlydump dump PCI config space before the kernel 4252 changes anything 4253 off [X86] don't probe for the PCI bus 4254 bios [X86-32] force use of PCI BIOS, don't access 4255 the hardware directly. Use this if your machine 4256 has a non-standard PCI host bridge. 4257 nobios [X86-32] disallow use of PCI BIOS, only direct 4258 hardware access methods are allowed. Use this 4259 if you experience crashes upon bootup and you 4260 suspect they are caused by the BIOS. 4261 conf1 [X86] Force use of PCI Configuration Access 4262 Mechanism 1 (config address in IO port 0xCF8, 4263 data in IO port 0xCFC, both 32-bit). 4264 conf2 [X86] Force use of PCI Configuration Access 4265 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4266 the function, IO port 0xCFA, also 8-bit, sets 4267 bus number. The config space is then accessed 4268 through ports 0xC000-0xCFFF). 4269 See http://wiki.osdev.org/PCI for more info 4270 on the configuration access mechanisms. 4271 noaer [PCIE] If the PCIEAER kernel config parameter is 4272 enabled, this kernel boot option can be used to 4273 disable the use of PCIE advanced error reporting. 4274 nodomains [PCI] Disable support for multiple PCI 4275 root domains (aka PCI segments, in ACPI-speak). 4276 nommconf [X86] Disable use of MMCONFIG for PCI 4277 Configuration 4278 check_enable_amd_mmconf [X86] check for and enable 4279 properly configured MMIO access to PCI 4280 config space on AMD family 10h CPU 4281 nomsi [MSI] If the PCI_MSI kernel config parameter is 4282 enabled, this kernel boot option can be used to 4283 disable the use of MSI interrupts system-wide. 4284 noioapicquirk [APIC] Disable all boot interrupt quirks. 4285 Safety option to keep boot IRQs enabled. This 4286 should never be necessary. 4287 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4288 primary IO-APIC for bridges that cannot disable 4289 boot IRQs. This fixes a source of spurious IRQs 4290 when the system masks IRQs. 4291 noioapicreroute [APIC] Disable workaround that uses the 4292 boot IRQ equivalent of an IRQ that connects to 4293 a chipset where boot IRQs cannot be disabled. 4294 The opposite of ioapicreroute. 4295 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4296 routing table. These calls are known to be buggy 4297 on several machines and they hang the machine 4298 when used, but on other computers it's the only 4299 way to get the interrupt routing table. Try 4300 this option if the kernel is unable to allocate 4301 IRQs or discover secondary PCI buses on your 4302 motherboard. 4303 rom [X86] Assign address space to expansion ROMs. 4304 Use with caution as certain devices share 4305 address decoders between ROMs and other 4306 resources. 4307 norom [X86] Do not assign address space to 4308 expansion ROMs that do not already have 4309 BIOS assigned address ranges. 4310 nobar [X86] Do not assign address space to the 4311 BARs that weren't assigned by the BIOS. 4312 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4313 assigned automatically to PCI devices. You can 4314 make the kernel exclude IRQs of your ISA cards 4315 this way. 4316 pirqaddr=0xAAAAA [X86] Specify the physical address 4317 of the PIRQ table (normally generated 4318 by the BIOS) if it is outside the 4319 F0000h-100000h range. 4320 lastbus=N [X86] Scan all buses thru bus #N. Can be 4321 useful if the kernel is unable to find your 4322 secondary buses and you want to tell it 4323 explicitly which ones they are. 4324 assign-busses [X86] Always assign all PCI bus 4325 numbers ourselves, overriding 4326 whatever the firmware may have done. 4327 usepirqmask [X86] Honor the possible IRQ mask stored 4328 in the BIOS $PIR table. This is needed on 4329 some systems with broken BIOSes, notably 4330 some HP Pavilion N5400 and Omnibook XE3 4331 notebooks. This will have no effect if ACPI 4332 IRQ routing is enabled. 4333 noacpi [X86] Do not use ACPI for IRQ routing 4334 or for PCI scanning. 4335 use_crs [X86] Use PCI host bridge window information 4336 from ACPI. On BIOSes from 2008 or later, this 4337 is enabled by default. If you need to use this, 4338 please report a bug. 4339 nocrs [X86] Ignore PCI host bridge windows from ACPI. 4340 If you need to use this, please report a bug. 4341 use_e820 [X86] Use E820 reservations to exclude parts of 4342 PCI host bridge windows. This is a workaround 4343 for BIOS defects in host bridge _CRS methods. 4344 If you need to use this, please report a bug to 4345 <[email protected]>. 4346 no_e820 [X86] Ignore E820 reservations for PCI host 4347 bridge windows. This is the default on modern 4348 hardware. If you need to use this, please report 4349 a bug to <[email protected]>. 4350 routeirq Do IRQ routing for all PCI devices. 4351 This is normally done in pci_enable_device(), 4352 so this option is a temporary workaround 4353 for broken drivers that don't call it. 4354 skip_isa_align [X86] do not align io start addr, so can 4355 handle more pci cards 4356 noearly [X86] Don't do any early type 1 scanning. 4357 This might help on some broken boards which 4358 machine check when some devices' config space 4359 is read. But various workarounds are disabled 4360 and some IOMMU drivers will not work. 4361 bfsort Sort PCI devices into breadth-first order. 4362 This sorting is done to get a device 4363 order compatible with older (<= 2.4) kernels. 4364 nobfsort Don't sort PCI devices into breadth-first order. 4365 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 4366 tuning and use the BIOS-configured MPS defaults. 4367 pcie_bus_safe Set every device's MPS to the largest value 4368 supported by all devices below the root complex. 4369 pcie_bus_perf Set device MPS to the largest allowable MPS 4370 based on its parent bus. Also set MRRS (Max 4371 Read Request Size) to the largest supported 4372 value (no larger than the MPS that the device 4373 or bus can support) for best performance. 4374 pcie_bus_peer2peer Set every device's MPS to 128B, which 4375 every device is guaranteed to support. This 4376 configuration allows peer-to-peer DMA between 4377 any pair of devices, possibly at the cost of 4378 reduced performance. This also guarantees 4379 that hot-added devices will work. 4380 cbiosize=nn[KMG] The fixed amount of bus space which is 4381 reserved for the CardBus bridge's IO window. 4382 The default value is 256 bytes. 4383 cbmemsize=nn[KMG] The fixed amount of bus space which is 4384 reserved for the CardBus bridge's memory 4385 window. The default value is 64 megabytes. 4386 resource_alignment= 4387 Format: 4388 [<order of align>@]<pci_dev>[; ...] 4389 Specifies alignment and device to reassign 4390 aligned memory resources. How to 4391 specify the device is described above. 4392 If <order of align> is not specified, 4393 PAGE_SIZE is used as alignment. 4394 A PCI-PCI bridge can be specified if resource 4395 windows need to be expanded. 4396 To specify the alignment for several 4397 instances of a device, the PCI vendor, 4398 device, subvendor, and subdevice may be 4399 specified, e.g., 12@pci:8086:9c22:103c:198f 4400 for 4096-byte alignment. 4401 ecrc= Enable/disable PCIe ECRC (transaction layer 4402 end-to-end CRC checking). Only effective if 4403 OS has native AER control (either granted by 4404 ACPI _OSC or forced via "pcie_ports=native") 4405 bios: Use BIOS/firmware settings. This is the 4406 the default. 4407 off: Turn ECRC off 4408 on: Turn ECRC on. 4409 hpiosize=nn[KMG] The fixed amount of bus space which is 4410 reserved for hotplug bridge's IO window. 4411 Default size is 256 bytes. 4412 hpmmiosize=nn[KMG] The fixed amount of bus space which is 4413 reserved for hotplug bridge's MMIO window. 4414 Default size is 2 megabytes. 4415 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 4416 reserved for hotplug bridge's MMIO_PREF window. 4417 Default size is 2 megabytes. 4418 hpmemsize=nn[KMG] The fixed amount of bus space which is 4419 reserved for hotplug bridge's MMIO and 4420 MMIO_PREF window. 4421 Default size is 2 megabytes. 4422 hpbussize=nn The minimum amount of additional bus numbers 4423 reserved for buses below a hotplug bridge. 4424 Default is 1. 4425 realloc= Enable/disable reallocating PCI bridge resources 4426 if allocations done by BIOS are too small to 4427 accommodate resources required by all child 4428 devices. 4429 off: Turn realloc off 4430 on: Turn realloc on 4431 realloc same as realloc=on 4432 noari do not use PCIe ARI. 4433 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 4434 do not use PCIe ATS (and IOMMU device IOTLB). 4435 pcie_scan_all Scan all possible PCIe devices. Otherwise we 4436 only look for one device below a PCIe downstream 4437 port. 4438 big_root_window Try to add a big 64bit memory window to the PCIe 4439 root complex on AMD CPUs. Some GFX hardware 4440 can resize a BAR to allow access to all VRAM. 4441 Adding the window is slightly risky (it may 4442 conflict with unreported devices), so this 4443 taints the kernel. 4444 disable_acs_redir=<pci_dev>[; ...] 4445 Specify one or more PCI devices (in the format 4446 specified above) separated by semicolons. 4447 Each device specified will have the PCI ACS 4448 redirect capabilities forced off which will 4449 allow P2P traffic between devices through 4450 bridges without forcing it upstream. Note: 4451 this removes isolation between devices and 4452 may put more devices in an IOMMU group. 4453 force_floating [S390] Force usage of floating interrupts. 4454 nomio [S390] Do not use MIO instructions. 4455 norid [S390] ignore the RID field and force use of 4456 one PCI domain per PCI function 4457 4458 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 4459 Management. 4460 off Disable ASPM. 4461 force Enable ASPM even on devices that claim not to support it. 4462 WARNING: Forcing ASPM on may cause system lockups. 4463 4464 pcie_ports= [PCIE] PCIe port services handling: 4465 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 4466 even if the platform doesn't give the OS permission to 4467 use them. This may cause conflicts if the platform 4468 also tries to use these services. 4469 dpc-native Use native PCIe service for DPC only. May 4470 cause conflicts if firmware uses AER or DPC. 4471 compat Disable native PCIe services (PME, AER, DPC, PCIe 4472 hotplug). 4473 4474 pcie_port_pm= [PCIE] PCIe port power management handling: 4475 off Disable power management of all PCIe ports 4476 force Forcibly enable power management of all PCIe ports 4477 4478 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 4479 nomsi Do not use MSI for native PCIe PME signaling (this makes 4480 all PCIe root ports use INTx for all services). 4481 4482 pcmv= [HW,PCMCIA] BadgePAD 4 4483 4484 pd_ignore_unused 4485 [PM] 4486 Keep all power-domains already enabled by bootloader on, 4487 even if no driver has claimed them. This is useful 4488 for debug and development, but should not be 4489 needed on a platform with proper driver support. 4490 4491 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 4492 boot time. 4493 Format: { 0 | 1 } 4494 See arch/parisc/kernel/pdc_chassis.c 4495 4496 percpu_alloc= Select which percpu first chunk allocator to use. 4497 Currently supported values are "embed" and "page". 4498 Archs may support subset or none of the selections. 4499 See comments in mm/percpu.c for details on each 4500 allocator. This parameter is primarily for debugging 4501 and performance comparison. 4502 4503 pirq= [SMP,APIC] Manual mp-table setup 4504 See Documentation/arch/x86/i386/IO-APIC.rst. 4505 4506 plip= [PPT,NET] Parallel port network link 4507 Format: { parport<nr> | timid | 0 } 4508 See also Documentation/admin-guide/parport.rst. 4509 4510 pmtmr= [X86] Manual setup of pmtmr I/O Port. 4511 Override pmtimer IOPort with a hex value. 4512 e.g. pmtmr=0x508 4513 4514 pmu_override= [PPC] Override the PMU. 4515 This option takes over the PMU facility, so it is no 4516 longer usable by perf. Setting this option starts the 4517 PMU counters by setting MMCR0 to 0 (the FC bit is 4518 cleared). If a number is given, then MMCR1 is set to 4519 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 4520 remains 0. 4521 4522 pm_debug_messages [SUSPEND,KNL] 4523 Enable suspend/resume debug messages during boot up. 4524 4525 pnp.debug=1 [PNP] 4526 Enable PNP debug messages (depends on the 4527 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 4528 via /sys/module/pnp/parameters/debug. We always show 4529 current resource usage; turning this on also shows 4530 possible settings and some assignment information. 4531 4532 pnpacpi= [ACPI] 4533 { off } 4534 4535 pnpbios= [ISAPNP] 4536 { on | off | curr | res | no-curr | no-res } 4537 4538 pnp_reserve_irq= 4539 [ISAPNP] Exclude IRQs for the autoconfiguration 4540 4541 pnp_reserve_dma= 4542 [ISAPNP] Exclude DMAs for the autoconfiguration 4543 4544 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 4545 Ranges are in pairs (I/O port base and size). 4546 4547 pnp_reserve_mem= 4548 [ISAPNP] Exclude memory regions for the 4549 autoconfiguration. 4550 Ranges are in pairs (memory base and size). 4551 4552 ports= [IP_VS_FTP] IPVS ftp helper module 4553 Default is 21. 4554 Up to 8 (IP_VS_APP_MAX_PORTS) ports 4555 may be specified. 4556 Format: <port>,<port>.... 4557 4558 powersave=off [PPC] This option disables power saving features. 4559 It specifically disables cpuidle and sets the 4560 platform machine description specific power_save 4561 function to NULL. On Idle the CPU just reduces 4562 execution priority. 4563 4564 ppc_strict_facility_enable 4565 [PPC] This option catches any kernel floating point, 4566 Altivec, VSX and SPE outside of regions specifically 4567 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 4568 There is some performance impact when enabling this. 4569 4570 ppc_tm= [PPC] 4571 Format: {"off"} 4572 Disable Hardware Transactional Memory 4573 4574 preempt= [KNL] 4575 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 4576 none - Limited to cond_resched() calls 4577 voluntary - Limited to cond_resched() and might_sleep() calls 4578 full - Any section that isn't explicitly preempt disabled 4579 can be preempted anytime. 4580 4581 print-fatal-signals= 4582 [KNL] debug: print fatal signals 4583 4584 If enabled, warn about various signal handling 4585 related application anomalies: too many signals, 4586 too many POSIX.1 timers, fatal signals causing a 4587 coredump - etc. 4588 4589 If you hit the warning due to signal overflow, 4590 you might want to try "ulimit -i unlimited". 4591 4592 default: off. 4593 4594 printk.always_kmsg_dump= 4595 Trigger kmsg_dump for cases other than kernel oops or 4596 panics 4597 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4598 default: disabled 4599 4600 printk.console_no_auto_verbose= 4601 Disable console loglevel raise on oops, panic 4602 or lockdep-detected issues (only if lock debug is on). 4603 With an exception to setups with low baudrate on 4604 serial console, keeping this 0 is a good choice 4605 in order to provide more debug information. 4606 Format: <bool> 4607 default: 0 (auto_verbose is enabled) 4608 4609 printk.devkmsg={on,off,ratelimit} 4610 Control writing to /dev/kmsg. 4611 on - unlimited logging to /dev/kmsg from userspace 4612 off - logging to /dev/kmsg disabled 4613 ratelimit - ratelimit the logging 4614 Default: ratelimit 4615 4616 printk.time= Show timing data prefixed to each printk message line 4617 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4618 4619 processor.max_cstate= [HW,ACPI] 4620 Limit processor to maximum C-state 4621 max_cstate=9 overrides any DMI blacklist limit. 4622 4623 processor.nocst [HW,ACPI] 4624 Ignore the _CST method to determine C-states, 4625 instead using the legacy FADT method 4626 4627 profile= [KNL] Enable kernel profiling via /proc/profile 4628 Format: [<profiletype>,]<number> 4629 Param: <profiletype>: "schedule", "sleep", or "kvm" 4630 [defaults to kernel profiling] 4631 Param: "schedule" - profile schedule points. 4632 Param: "sleep" - profile D-state sleeping (millisecs). 4633 Requires CONFIG_SCHEDSTATS 4634 Param: "kvm" - profile VM exits. 4635 Param: <number> - step/bucket size as a power of 2 for 4636 statistical time based profiling. 4637 4638 prompt_ramdisk= [RAM] [Deprecated] 4639 4640 prot_virt= [S390] enable hosting protected virtual machines 4641 isolated from the hypervisor (if hardware supports 4642 that). 4643 Format: <bool> 4644 4645 psi= [KNL] Enable or disable pressure stall information 4646 tracking. 4647 Format: <bool> 4648 4649 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 4650 probe for; one of (bare|imps|exps|lifebook|any). 4651 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 4652 per second. 4653 psmouse.resetafter= [HW,MOUSE] 4654 Try to reset the device after so many bad packets 4655 (0 = never). 4656 psmouse.resolution= 4657 [HW,MOUSE] Set desired mouse resolution, in dpi. 4658 psmouse.smartscroll= 4659 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 4660 0 = disabled, 1 = enabled (default). 4661 4662 pstore.backend= Specify the name of the pstore backend to use 4663 4664 pti= [X86-64] Control Page Table Isolation of user and 4665 kernel address spaces. Disabling this feature 4666 removes hardening, but improves performance of 4667 system calls and interrupts. 4668 4669 on - unconditionally enable 4670 off - unconditionally disable 4671 auto - kernel detects whether your CPU model is 4672 vulnerable to issues that PTI mitigates 4673 4674 Not specifying this option is equivalent to pti=auto. 4675 4676 pty.legacy_count= 4677 [KNL] Number of legacy pty's. Overwrites compiled-in 4678 default number. 4679 4680 quiet [KNL] Disable most log messages 4681 4682 r128= [HW,DRM] 4683 4684 radix_hcall_invalidate=on [PPC/PSERIES] 4685 Disable RADIX GTSE feature and use hcall for TLB 4686 invalidate. 4687 4688 raid= [HW,RAID] 4689 See Documentation/admin-guide/md.rst. 4690 4691 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 4692 See Documentation/admin-guide/blockdev/ramdisk.rst. 4693 4694 ramdisk_start= [RAM] RAM disk image start address 4695 4696 random.trust_cpu=off 4697 [KNL] Disable trusting the use of the CPU's 4698 random number generator (if available) to 4699 initialize the kernel's RNG. 4700 4701 random.trust_bootloader=off 4702 [KNL] Disable trusting the use of the a seed 4703 passed by the bootloader (if available) to 4704 initialize the kernel's RNG. 4705 4706 randomize_kstack_offset= 4707 [KNL] Enable or disable kernel stack offset 4708 randomization, which provides roughly 5 bits of 4709 entropy, frustrating memory corruption attacks 4710 that depend on stack address determinism or 4711 cross-syscall address exposures. This is only 4712 available on architectures that have defined 4713 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 4714 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4715 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 4716 4717 ras=option[,option,...] [KNL] RAS-specific options 4718 4719 cec_disable [X86] 4720 Disable the Correctable Errors Collector, 4721 see CONFIG_RAS_CEC help text. 4722 4723 rcu_nocbs[=cpu-list] 4724 [KNL] The optional argument is a cpu list, 4725 as described above. 4726 4727 In kernels built with CONFIG_RCU_NOCB_CPU=y, 4728 enable the no-callback CPU mode, which prevents 4729 such CPUs' callbacks from being invoked in 4730 softirq context. Invocation of such CPUs' RCU 4731 callbacks will instead be offloaded to "rcuox/N" 4732 kthreads created for that purpose, where "x" is 4733 "p" for RCU-preempt, "s" for RCU-sched, and "g" 4734 for the kthreads that mediate grace periods; and 4735 "N" is the CPU number. This reduces OS jitter on 4736 the offloaded CPUs, which can be useful for HPC 4737 and real-time workloads. It can also improve 4738 energy efficiency for asymmetric multiprocessors. 4739 4740 If a cpulist is passed as an argument, the specified 4741 list of CPUs is set to no-callback mode from boot. 4742 4743 Otherwise, if the '=' sign and the cpulist 4744 arguments are omitted, no CPU will be set to 4745 no-callback mode from boot but the mode may be 4746 toggled at runtime via cpusets. 4747 4748 Note that this argument takes precedence over 4749 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4750 4751 rcu_nocb_poll [KNL] 4752 Rather than requiring that offloaded CPUs 4753 (specified by rcu_nocbs= above) explicitly 4754 awaken the corresponding "rcuoN" kthreads, 4755 make these kthreads poll for callbacks. 4756 This improves the real-time response for the 4757 offloaded CPUs by relieving them of the need to 4758 wake up the corresponding kthread, but degrades 4759 energy efficiency by requiring that the kthreads 4760 periodically wake up to do the polling. 4761 4762 rcutree.blimit= [KNL] 4763 Set maximum number of finished RCU callbacks to 4764 process in one batch. 4765 4766 rcutree.dump_tree= [KNL] 4767 Dump the structure of the rcu_node combining tree 4768 out at early boot. This is used for diagnostic 4769 purposes, to verify correct tree setup. 4770 4771 rcutree.gp_cleanup_delay= [KNL] 4772 Set the number of jiffies to delay each step of 4773 RCU grace-period cleanup. 4774 4775 rcutree.gp_init_delay= [KNL] 4776 Set the number of jiffies to delay each step of 4777 RCU grace-period initialization. 4778 4779 rcutree.gp_preinit_delay= [KNL] 4780 Set the number of jiffies to delay each step of 4781 RCU grace-period pre-initialization, that is, 4782 the propagation of recent CPU-hotplug changes up 4783 the rcu_node combining tree. 4784 4785 rcutree.jiffies_till_first_fqs= [KNL] 4786 Set delay from grace-period initialization to 4787 first attempt to force quiescent states. 4788 Units are jiffies, minimum value is zero, 4789 and maximum value is HZ. 4790 4791 rcutree.jiffies_till_next_fqs= [KNL] 4792 Set delay between subsequent attempts to force 4793 quiescent states. Units are jiffies, minimum 4794 value is one, and maximum value is HZ. 4795 4796 rcutree.jiffies_till_sched_qs= [KNL] 4797 Set required age in jiffies for a 4798 given grace period before RCU starts 4799 soliciting quiescent-state help from 4800 rcu_note_context_switch() and cond_resched(). 4801 If not specified, the kernel will calculate 4802 a value based on the most recent settings 4803 of rcutree.jiffies_till_first_fqs 4804 and rcutree.jiffies_till_next_fqs. 4805 This calculated value may be viewed in 4806 rcutree.jiffies_to_sched_qs. Any attempt to set 4807 rcutree.jiffies_to_sched_qs will be cheerfully 4808 overwritten. 4809 4810 rcutree.kthread_prio= [KNL,BOOT] 4811 Set the SCHED_FIFO priority of the RCU per-CPU 4812 kthreads (rcuc/N). This value is also used for 4813 the priority of the RCU boost threads (rcub/N) 4814 and for the RCU grace-period kthreads (rcu_bh, 4815 rcu_preempt, and rcu_sched). If RCU_BOOST is 4816 set, valid values are 1-99 and the default is 1 4817 (the least-favored priority). Otherwise, when 4818 RCU_BOOST is not set, valid values are 0-99 and 4819 the default is zero (non-realtime operation). 4820 When RCU_NOCB_CPU is set, also adjust the 4821 priority of NOCB callback kthreads. 4822 4823 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 4824 On callback-offloaded (rcu_nocbs) CPUs, 4825 RCU reduces the lock contention that would 4826 otherwise be caused by callback floods through 4827 use of the ->nocb_bypass list. However, in the 4828 common non-flooded case, RCU queues directly to 4829 the main ->cblist in order to avoid the extra 4830 overhead of the ->nocb_bypass list and its lock. 4831 But if there are too many callbacks queued during 4832 a single jiffy, RCU pre-queues the callbacks into 4833 the ->nocb_bypass queue. The definition of "too 4834 many" is supplied by this kernel boot parameter. 4835 4836 rcutree.qhimark= [KNL] 4837 Set threshold of queued RCU callbacks beyond which 4838 batch limiting is disabled. 4839 4840 rcutree.qlowmark= [KNL] 4841 Set threshold of queued RCU callbacks below which 4842 batch limiting is re-enabled. 4843 4844 rcutree.qovld= [KNL] 4845 Set threshold of queued RCU callbacks beyond which 4846 RCU's force-quiescent-state scan will aggressively 4847 enlist help from cond_resched() and sched IPIs to 4848 help CPUs more quickly reach quiescent states. 4849 Set to less than zero to make this be set based 4850 on rcutree.qhimark at boot time and to zero to 4851 disable more aggressive help enlistment. 4852 4853 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 4854 Set the page-cache refill delay (in milliseconds) 4855 in response to low-memory conditions. The range 4856 of permitted values is in the range 0:100000. 4857 4858 rcutree.rcu_divisor= [KNL] 4859 Set the shift-right count to use to compute 4860 the callback-invocation batch limit bl from 4861 the number of callbacks queued on this CPU. 4862 The result will be bounded below by the value of 4863 the rcutree.blimit kernel parameter. Every bl 4864 callbacks, the softirq handler will exit in 4865 order to allow the CPU to do other work. 4866 4867 Please note that this callback-invocation batch 4868 limit applies only to non-offloaded callback 4869 invocation. Offloaded callbacks are instead 4870 invoked in the context of an rcuoc kthread, which 4871 scheduler will preempt as it does any other task. 4872 4873 rcutree.rcu_fanout_exact= [KNL] 4874 Disable autobalancing of the rcu_node combining 4875 tree. This is used by rcutorture, and might 4876 possibly be useful for architectures having high 4877 cache-to-cache transfer latencies. 4878 4879 rcutree.rcu_fanout_leaf= [KNL] 4880 Change the number of CPUs assigned to each 4881 leaf rcu_node structure. Useful for very 4882 large systems, which will choose the value 64, 4883 and for NUMA systems with large remote-access 4884 latencies, which will choose a value aligned 4885 with the appropriate hardware boundaries. 4886 4887 rcutree.rcu_min_cached_objs= [KNL] 4888 Minimum number of objects which are cached and 4889 maintained per one CPU. Object size is equal 4890 to PAGE_SIZE. The cache allows to reduce the 4891 pressure to page allocator, also it makes the 4892 whole algorithm to behave better in low memory 4893 condition. 4894 4895 rcutree.rcu_nocb_gp_stride= [KNL] 4896 Set the number of NOCB callback kthreads in 4897 each group, which defaults to the square root 4898 of the number of CPUs. Larger numbers reduce 4899 the wakeup overhead on the global grace-period 4900 kthread, but increases that same overhead on 4901 each group's NOCB grace-period kthread. 4902 4903 rcutree.rcu_kick_kthreads= [KNL] 4904 Cause the grace-period kthread to get an extra 4905 wake_up() if it sleeps three times longer than 4906 it should at force-quiescent-state time. 4907 This wake_up() will be accompanied by a 4908 WARN_ONCE() splat and an ftrace_dump(). 4909 4910 rcutree.rcu_resched_ns= [KNL] 4911 Limit the time spend invoking a batch of RCU 4912 callbacks to the specified number of nanoseconds. 4913 By default, this limit is checked only once 4914 every 32 callbacks in order to limit the pain 4915 inflicted by local_clock() overhead. 4916 4917 rcutree.rcu_unlock_delay= [KNL] 4918 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 4919 this specifies an rcu_read_unlock()-time delay 4920 in microseconds. This defaults to zero. 4921 Larger delays increase the probability of 4922 catching RCU pointer leaks, that is, buggy use 4923 of RCU-protected pointers after the relevant 4924 rcu_read_unlock() has completed. 4925 4926 rcutree.sysrq_rcu= [KNL] 4927 Commandeer a sysrq key to dump out Tree RCU's 4928 rcu_node tree with an eye towards determining 4929 why a new grace period has not yet started. 4930 4931 rcutree.use_softirq= [KNL] 4932 If set to zero, move all RCU_SOFTIRQ processing to 4933 per-CPU rcuc kthreads. Defaults to a non-zero 4934 value, meaning that RCU_SOFTIRQ is used by default. 4935 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4936 4937 But note that CONFIG_PREEMPT_RT=y kernels disable 4938 this kernel boot parameter, forcibly setting it 4939 to zero. 4940 4941 rcuscale.gp_async= [KNL] 4942 Measure performance of asynchronous 4943 grace-period primitives such as call_rcu(). 4944 4945 rcuscale.gp_async_max= [KNL] 4946 Specify the maximum number of outstanding 4947 callbacks per writer thread. When a writer 4948 thread exceeds this limit, it invokes the 4949 corresponding flavor of rcu_barrier() to allow 4950 previously posted callbacks to drain. 4951 4952 rcuscale.gp_exp= [KNL] 4953 Measure performance of expedited synchronous 4954 grace-period primitives. 4955 4956 rcuscale.holdoff= [KNL] 4957 Set test-start holdoff period. The purpose of 4958 this parameter is to delay the start of the 4959 test until boot completes in order to avoid 4960 interference. 4961 4962 rcuscale.kfree_by_call_rcu= [KNL] 4963 In kernels built with CONFIG_RCU_LAZY=y, test 4964 call_rcu() instead of kfree_rcu(). 4965 4966 rcuscale.kfree_mult= [KNL] 4967 Instead of allocating an object of size kfree_obj, 4968 allocate one of kfree_mult * sizeof(kfree_obj). 4969 Defaults to 1. 4970 4971 rcuscale.kfree_rcu_test= [KNL] 4972 Set to measure performance of kfree_rcu() flooding. 4973 4974 rcuscale.kfree_rcu_test_double= [KNL] 4975 Test the double-argument variant of kfree_rcu(). 4976 If this parameter has the same value as 4977 rcuscale.kfree_rcu_test_single, both the single- 4978 and double-argument variants are tested. 4979 4980 rcuscale.kfree_rcu_test_single= [KNL] 4981 Test the single-argument variant of kfree_rcu(). 4982 If this parameter has the same value as 4983 rcuscale.kfree_rcu_test_double, both the single- 4984 and double-argument variants are tested. 4985 4986 rcuscale.kfree_nthreads= [KNL] 4987 The number of threads running loops of kfree_rcu(). 4988 4989 rcuscale.kfree_alloc_num= [KNL] 4990 Number of allocations and frees done in an iteration. 4991 4992 rcuscale.kfree_loops= [KNL] 4993 Number of loops doing rcuscale.kfree_alloc_num number 4994 of allocations and frees. 4995 4996 rcuscale.minruntime= [KNL] 4997 Set the minimum test run time in seconds. This 4998 does not affect the data-collection interval, 4999 but instead allows better measurement of things 5000 like CPU consumption. 5001 5002 rcuscale.nreaders= [KNL] 5003 Set number of RCU readers. The value -1 selects 5004 N, where N is the number of CPUs. A value 5005 "n" less than -1 selects N-n+1, where N is again 5006 the number of CPUs. For example, -2 selects N 5007 (the number of CPUs), -3 selects N+1, and so on. 5008 A value of "n" less than or equal to -N selects 5009 a single reader. 5010 5011 rcuscale.nwriters= [KNL] 5012 Set number of RCU writers. The values operate 5013 the same as for rcuscale.nreaders. 5014 N, where N is the number of CPUs 5015 5016 rcuscale.scale_type= [KNL] 5017 Specify the RCU implementation to test. 5018 5019 rcuscale.shutdown= [KNL] 5020 Shut the system down after performance tests 5021 complete. This is useful for hands-off automated 5022 testing. 5023 5024 rcuscale.verbose= [KNL] 5025 Enable additional printk() statements. 5026 5027 rcuscale.writer_holdoff= [KNL] 5028 Write-side holdoff between grace periods, 5029 in microseconds. The default of zero says 5030 no holdoff. 5031 5032 rcuscale.writer_holdoff_jiffies= [KNL] 5033 Additional write-side holdoff between grace 5034 periods, but in jiffies. The default of zero 5035 says no holdoff. 5036 5037 rcutorture.fqs_duration= [KNL] 5038 Set duration of force_quiescent_state bursts 5039 in microseconds. 5040 5041 rcutorture.fqs_holdoff= [KNL] 5042 Set holdoff time within force_quiescent_state bursts 5043 in microseconds. 5044 5045 rcutorture.fqs_stutter= [KNL] 5046 Set wait time between force_quiescent_state bursts 5047 in seconds. 5048 5049 rcutorture.fwd_progress= [KNL] 5050 Specifies the number of kthreads to be used 5051 for RCU grace-period forward-progress testing 5052 for the types of RCU supporting this notion. 5053 Defaults to 1 kthread, values less than zero or 5054 greater than the number of CPUs cause the number 5055 of CPUs to be used. 5056 5057 rcutorture.fwd_progress_div= [KNL] 5058 Specify the fraction of a CPU-stall-warning 5059 period to do tight-loop forward-progress testing. 5060 5061 rcutorture.fwd_progress_holdoff= [KNL] 5062 Number of seconds to wait between successive 5063 forward-progress tests. 5064 5065 rcutorture.fwd_progress_need_resched= [KNL] 5066 Enclose cond_resched() calls within checks for 5067 need_resched() during tight-loop forward-progress 5068 testing. 5069 5070 rcutorture.gp_cond= [KNL] 5071 Use conditional/asynchronous update-side 5072 primitives, if available. 5073 5074 rcutorture.gp_exp= [KNL] 5075 Use expedited update-side primitives, if available. 5076 5077 rcutorture.gp_normal= [KNL] 5078 Use normal (non-expedited) asynchronous 5079 update-side primitives, if available. 5080 5081 rcutorture.gp_sync= [KNL] 5082 Use normal (non-expedited) synchronous 5083 update-side primitives, if available. If all 5084 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5085 rcutorture.gp_normal=, and rcutorture.gp_sync= 5086 are zero, rcutorture acts as if is interpreted 5087 they are all non-zero. 5088 5089 rcutorture.irqreader= [KNL] 5090 Run RCU readers from irq handlers, or, more 5091 accurately, from a timer handler. Not all RCU 5092 flavors take kindly to this sort of thing. 5093 5094 rcutorture.leakpointer= [KNL] 5095 Leak an RCU-protected pointer out of the reader. 5096 This can of course result in splats, and is 5097 intended to test the ability of things like 5098 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5099 such leaks. 5100 5101 rcutorture.n_barrier_cbs= [KNL] 5102 Set callbacks/threads for rcu_barrier() testing. 5103 5104 rcutorture.nfakewriters= [KNL] 5105 Set number of concurrent RCU writers. These just 5106 stress RCU, they don't participate in the actual 5107 test, hence the "fake". 5108 5109 rcutorture.nocbs_nthreads= [KNL] 5110 Set number of RCU callback-offload togglers. 5111 Zero (the default) disables toggling. 5112 5113 rcutorture.nocbs_toggle= [KNL] 5114 Set the delay in milliseconds between successive 5115 callback-offload toggling attempts. 5116 5117 rcutorture.nreaders= [KNL] 5118 Set number of RCU readers. The value -1 selects 5119 N-1, where N is the number of CPUs. A value 5120 "n" less than -1 selects N-n-2, where N is again 5121 the number of CPUs. For example, -2 selects N 5122 (the number of CPUs), -3 selects N+1, and so on. 5123 5124 rcutorture.object_debug= [KNL] 5125 Enable debug-object double-call_rcu() testing. 5126 5127 rcutorture.onoff_holdoff= [KNL] 5128 Set time (s) after boot for CPU-hotplug testing. 5129 5130 rcutorture.onoff_interval= [KNL] 5131 Set time (jiffies) between CPU-hotplug operations, 5132 or zero to disable CPU-hotplug testing. 5133 5134 rcutorture.read_exit= [KNL] 5135 Set the number of read-then-exit kthreads used 5136 to test the interaction of RCU updaters and 5137 task-exit processing. 5138 5139 rcutorture.read_exit_burst= [KNL] 5140 The number of times in a given read-then-exit 5141 episode that a set of read-then-exit kthreads 5142 is spawned. 5143 5144 rcutorture.read_exit_delay= [KNL] 5145 The delay, in seconds, between successive 5146 read-then-exit testing episodes. 5147 5148 rcutorture.shuffle_interval= [KNL] 5149 Set task-shuffle interval (s). Shuffling tasks 5150 allows some CPUs to go into dyntick-idle mode 5151 during the rcutorture test. 5152 5153 rcutorture.shutdown_secs= [KNL] 5154 Set time (s) after boot system shutdown. This 5155 is useful for hands-off automated testing. 5156 5157 rcutorture.stall_cpu= [KNL] 5158 Duration of CPU stall (s) to test RCU CPU stall 5159 warnings, zero to disable. 5160 5161 rcutorture.stall_cpu_block= [KNL] 5162 Sleep while stalling if set. This will result 5163 in warnings from preemptible RCU in addition to 5164 any other stall-related activity. Note that 5165 in kernels built with CONFIG_PREEMPTION=n and 5166 CONFIG_PREEMPT_COUNT=y, this parameter will 5167 cause the CPU to pass through a quiescent state. 5168 Given CONFIG_PREEMPTION=n, this will suppress 5169 RCU CPU stall warnings, but will instead result 5170 in scheduling-while-atomic splats. 5171 5172 Use of this module parameter results in splats. 5173 5174 5175 rcutorture.stall_cpu_holdoff= [KNL] 5176 Time to wait (s) after boot before inducing stall. 5177 5178 rcutorture.stall_cpu_irqsoff= [KNL] 5179 Disable interrupts while stalling if set. 5180 5181 rcutorture.stall_gp_kthread= [KNL] 5182 Duration (s) of forced sleep within RCU 5183 grace-period kthread to test RCU CPU stall 5184 warnings, zero to disable. If both stall_cpu 5185 and stall_gp_kthread are specified, the 5186 kthread is starved first, then the CPU. 5187 5188 rcutorture.stat_interval= [KNL] 5189 Time (s) between statistics printk()s. 5190 5191 rcutorture.stutter= [KNL] 5192 Time (s) to stutter testing, for example, specifying 5193 five seconds causes the test to run for five seconds, 5194 wait for five seconds, and so on. This tests RCU's 5195 ability to transition abruptly to and from idle. 5196 5197 rcutorture.test_boost= [KNL] 5198 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 5199 "Maybe" means test if the RCU implementation 5200 under test support RCU priority boosting. 5201 5202 rcutorture.test_boost_duration= [KNL] 5203 Duration (s) of each individual boost test. 5204 5205 rcutorture.test_boost_interval= [KNL] 5206 Interval (s) between each boost test. 5207 5208 rcutorture.test_no_idle_hz= [KNL] 5209 Test RCU's dyntick-idle handling. See also the 5210 rcutorture.shuffle_interval parameter. 5211 5212 rcutorture.torture_type= [KNL] 5213 Specify the RCU implementation to test. 5214 5215 rcutorture.verbose= [KNL] 5216 Enable additional printk() statements. 5217 5218 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 5219 Dump ftrace buffer after reporting RCU CPU 5220 stall warning. 5221 5222 rcupdate.rcu_cpu_stall_suppress= [KNL] 5223 Suppress RCU CPU stall warning messages. 5224 5225 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 5226 Suppress RCU CPU stall warning messages and 5227 rcutorture writer stall warnings that occur 5228 during early boot, that is, during the time 5229 before the init task is spawned. 5230 5231 rcupdate.rcu_cpu_stall_timeout= [KNL] 5232 Set timeout for RCU CPU stall warning messages. 5233 The value is in seconds and the maximum allowed 5234 value is 300 seconds. 5235 5236 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 5237 Set timeout for expedited RCU CPU stall warning 5238 messages. The value is in milliseconds 5239 and the maximum allowed value is 21000 5240 milliseconds. Please note that this value is 5241 adjusted to an arch timer tick resolution. 5242 Setting this to zero causes the value from 5243 rcupdate.rcu_cpu_stall_timeout to be used (after 5244 conversion from seconds to milliseconds). 5245 5246 rcupdate.rcu_cpu_stall_cputime= [KNL] 5247 Provide statistics on the cputime and count of 5248 interrupts and tasks during the sampling period. For 5249 multiple continuous RCU stalls, all sampling periods 5250 begin at half of the first RCU stall timeout. 5251 5252 rcupdate.rcu_exp_stall_task_details= [KNL] 5253 Print stack dumps of any tasks blocking the 5254 current expedited RCU grace period during an 5255 expedited RCU CPU stall warning. 5256 5257 rcupdate.rcu_expedited= [KNL] 5258 Use expedited grace-period primitives, for 5259 example, synchronize_rcu_expedited() instead 5260 of synchronize_rcu(). This reduces latency, 5261 but can increase CPU utilization, degrade 5262 real-time latency, and degrade energy efficiency. 5263 No effect on CONFIG_TINY_RCU kernels. 5264 5265 rcupdate.rcu_normal= [KNL] 5266 Use only normal grace-period primitives, 5267 for example, synchronize_rcu() instead of 5268 synchronize_rcu_expedited(). This improves 5269 real-time latency, CPU utilization, and 5270 energy efficiency, but can expose users to 5271 increased grace-period latency. This parameter 5272 overrides rcupdate.rcu_expedited. No effect on 5273 CONFIG_TINY_RCU kernels. 5274 5275 rcupdate.rcu_normal_after_boot= [KNL] 5276 Once boot has completed (that is, after 5277 rcu_end_inkernel_boot() has been invoked), use 5278 only normal grace-period primitives. No effect 5279 on CONFIG_TINY_RCU kernels. 5280 5281 But note that CONFIG_PREEMPT_RT=y kernels enables 5282 this kernel boot parameter, forcibly setting 5283 it to the value one, that is, converting any 5284 post-boot attempt at an expedited RCU grace 5285 period to instead use normal non-expedited 5286 grace-period processing. 5287 5288 rcupdate.rcu_task_collapse_lim= [KNL] 5289 Set the maximum number of callbacks present 5290 at the beginning of a grace period that allows 5291 the RCU Tasks flavors to collapse back to using 5292 a single callback queue. This switching only 5293 occurs when rcupdate.rcu_task_enqueue_lim is 5294 set to the default value of -1. 5295 5296 rcupdate.rcu_task_contend_lim= [KNL] 5297 Set the minimum number of callback-queuing-time 5298 lock-contention events per jiffy required to 5299 cause the RCU Tasks flavors to switch to per-CPU 5300 callback queuing. This switching only occurs 5301 when rcupdate.rcu_task_enqueue_lim is set to 5302 the default value of -1. 5303 5304 rcupdate.rcu_task_enqueue_lim= [KNL] 5305 Set the number of callback queues to use for the 5306 RCU Tasks family of RCU flavors. The default 5307 of -1 allows this to be automatically (and 5308 dynamically) adjusted. This parameter is intended 5309 for use in testing. 5310 5311 rcupdate.rcu_task_ipi_delay= [KNL] 5312 Set time in jiffies during which RCU tasks will 5313 avoid sending IPIs, starting with the beginning 5314 of a given grace period. Setting a large 5315 number avoids disturbing real-time workloads, 5316 but lengthens grace periods. 5317 5318 rcupdate.rcu_task_lazy_lim= [KNL] 5319 Number of callbacks on a given CPU that will 5320 cancel laziness on that CPU. Use -1 to disable 5321 cancellation of laziness, but be advised that 5322 doing so increases the danger of OOM due to 5323 callback flooding. 5324 5325 rcupdate.rcu_task_stall_info= [KNL] 5326 Set initial timeout in jiffies for RCU task stall 5327 informational messages, which give some indication 5328 of the problem for those not patient enough to 5329 wait for ten minutes. Informational messages are 5330 only printed prior to the stall-warning message 5331 for a given grace period. Disable with a value 5332 less than or equal to zero. Defaults to ten 5333 seconds. A change in value does not take effect 5334 until the beginning of the next grace period. 5335 5336 rcupdate.rcu_task_stall_info_mult= [KNL] 5337 Multiplier for time interval between successive 5338 RCU task stall informational messages for a given 5339 RCU tasks grace period. This value is clamped 5340 to one through ten, inclusive. It defaults to 5341 the value three, so that the first informational 5342 message is printed 10 seconds into the grace 5343 period, the second at 40 seconds, the third at 5344 160 seconds, and then the stall warning at 600 5345 seconds would prevent a fourth at 640 seconds. 5346 5347 rcupdate.rcu_task_stall_timeout= [KNL] 5348 Set timeout in jiffies for RCU task stall 5349 warning messages. Disable with a value less 5350 than or equal to zero. Defaults to ten minutes. 5351 A change in value does not take effect until 5352 the beginning of the next grace period. 5353 5354 rcupdate.rcu_tasks_lazy_ms= [KNL] 5355 Set timeout in milliseconds RCU Tasks asynchronous 5356 callback batching for call_rcu_tasks(). 5357 A negative value will take the default. A value 5358 of zero will disable batching. Batching is 5359 always disabled for synchronize_rcu_tasks(). 5360 5361 rcupdate.rcu_tasks_rude_lazy_ms= [KNL] 5362 Set timeout in milliseconds RCU Tasks 5363 Rude asynchronous callback batching for 5364 call_rcu_tasks_rude(). A negative value 5365 will take the default. A value of zero will 5366 disable batching. Batching is always disabled 5367 for synchronize_rcu_tasks_rude(). 5368 5369 rcupdate.rcu_tasks_trace_lazy_ms= [KNL] 5370 Set timeout in milliseconds RCU Tasks 5371 Trace asynchronous callback batching for 5372 call_rcu_tasks_trace(). A negative value 5373 will take the default. A value of zero will 5374 disable batching. Batching is always disabled 5375 for synchronize_rcu_tasks_trace(). 5376 5377 rcupdate.rcu_self_test= [KNL] 5378 Run the RCU early boot self tests 5379 5380 rdinit= [KNL] 5381 Format: <full_path> 5382 Run specified binary instead of /init from the ramdisk, 5383 used for early userspace startup. See initrd. 5384 5385 rdrand= [X86] 5386 force - Override the decision by the kernel to hide the 5387 advertisement of RDRAND support (this affects 5388 certain AMD processors because of buggy BIOS 5389 support, specifically around the suspend/resume 5390 path). 5391 5392 rdt= [HW,X86,RDT] 5393 Turn on/off individual RDT features. List is: 5394 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 5395 mba, smba, bmec. 5396 E.g. to turn on cmt and turn off mba use: 5397 rdt=cmt,!mba 5398 5399 reboot= [KNL] 5400 Format (x86 or x86_64): 5401 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 5402 [[,]s[mp]#### \ 5403 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 5404 [[,]f[orce] 5405 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 5406 (prefix with 'panic_' to set mode for panic 5407 reboot only), 5408 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 5409 reboot_force is either force or not specified, 5410 reboot_cpu is s[mp]#### with #### being the processor 5411 to be used for rebooting. 5412 5413 refscale.holdoff= [KNL] 5414 Set test-start holdoff period. The purpose of 5415 this parameter is to delay the start of the 5416 test until boot completes in order to avoid 5417 interference. 5418 5419 refscale.loops= [KNL] 5420 Set the number of loops over the synchronization 5421 primitive under test. Increasing this number 5422 reduces noise due to loop start/end overhead, 5423 but the default has already reduced the per-pass 5424 noise to a handful of picoseconds on ca. 2020 5425 x86 laptops. 5426 5427 refscale.nreaders= [KNL] 5428 Set number of readers. The default value of -1 5429 selects N, where N is roughly 75% of the number 5430 of CPUs. A value of zero is an interesting choice. 5431 5432 refscale.nruns= [KNL] 5433 Set number of runs, each of which is dumped onto 5434 the console log. 5435 5436 refscale.readdelay= [KNL] 5437 Set the read-side critical-section duration, 5438 measured in microseconds. 5439 5440 refscale.scale_type= [KNL] 5441 Specify the read-protection implementation to test. 5442 5443 refscale.shutdown= [KNL] 5444 Shut down the system at the end of the performance 5445 test. This defaults to 1 (shut it down) when 5446 refscale is built into the kernel and to 0 (leave 5447 it running) when refscale is built as a module. 5448 5449 refscale.verbose= [KNL] 5450 Enable additional printk() statements. 5451 5452 refscale.verbose_batched= [KNL] 5453 Batch the additional printk() statements. If zero 5454 (the default) or negative, print everything. Otherwise, 5455 print every Nth verbose statement, where N is the value 5456 specified. 5457 5458 relax_domain_level= 5459 [KNL, SMP] Set scheduler's default relax_domain_level. 5460 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 5461 5462 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 5463 Format: <base1>,<size1>[,<base2>,<size2>,...] 5464 Reserve I/O ports or memory so the kernel won't use 5465 them. If <base> is less than 0x10000, the region 5466 is assumed to be I/O ports; otherwise it is memory. 5467 5468 reservetop= [X86-32] 5469 Format: nn[KMG] 5470 Reserves a hole at the top of the kernel virtual 5471 address space. 5472 5473 reset_devices [KNL] Force drivers to reset the underlying device 5474 during initialization. 5475 5476 resume= [SWSUSP] 5477 Specify the partition device for software suspend 5478 Format: 5479 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 5480 5481 resume_offset= [SWSUSP] 5482 Specify the offset from the beginning of the partition 5483 given by "resume=" at which the swap header is located, 5484 in <PAGE_SIZE> units (needed only for swap files). 5485 See Documentation/power/swsusp-and-swap-files.rst 5486 5487 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 5488 read the resume files 5489 5490 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 5491 Useful for devices that are detected asynchronously 5492 (e.g. USB and MMC devices). 5493 5494 retain_initrd [RAM] Keep initrd memory after extraction 5495 5496 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 5497 Speculative Code Execution with Return Instructions) 5498 vulnerability. 5499 5500 AMD-based UNRET and IBPB mitigations alone do not stop 5501 sibling threads from influencing the predictions of other 5502 sibling threads. For that reason, STIBP is used on pro- 5503 cessors that support it, and mitigate SMT on processors 5504 that don't. 5505 5506 off - no mitigation 5507 auto - automatically select a migitation 5508 auto,nosmt - automatically select a mitigation, 5509 disabling SMT if necessary for 5510 the full mitigation (only on Zen1 5511 and older without STIBP). 5512 ibpb - On AMD, mitigate short speculation 5513 windows on basic block boundaries too. 5514 Safe, highest perf impact. It also 5515 enables STIBP if present. Not suitable 5516 on Intel. 5517 ibpb,nosmt - Like "ibpb" above but will disable SMT 5518 when STIBP is not available. This is 5519 the alternative for systems which do not 5520 have STIBP. 5521 unret - Force enable untrained return thunks, 5522 only effective on AMD f15h-f17h based 5523 systems. 5524 unret,nosmt - Like unret, but will disable SMT when STIBP 5525 is not available. This is the alternative for 5526 systems which do not have STIBP. 5527 5528 Selecting 'auto' will choose a mitigation method at run 5529 time according to the CPU. 5530 5531 Not specifying this option is equivalent to retbleed=auto. 5532 5533 rfkill.default_state= 5534 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 5535 etc. communication is blocked by default. 5536 1 Unblocked. 5537 5538 rfkill.master_switch_mode= 5539 0 The "airplane mode" button does nothing. 5540 1 The "airplane mode" button toggles between everything 5541 blocked and the previous configuration. 5542 2 The "airplane mode" button toggles between everything 5543 blocked and everything unblocked. 5544 5545 rhash_entries= [KNL,NET] 5546 Set number of hash buckets for route cache 5547 5548 ring3mwait=disable 5549 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 5550 CPUs. 5551 5552 riscv_isa_fallback [RISCV] 5553 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 5554 falling back to detecting extension support by parsing 5555 "riscv,isa" property on devicetree systems when the 5556 replacement properties are not found. See the Kconfig 5557 entry for RISCV_ISA_FALLBACK. 5558 5559 ro [KNL] Mount root device read-only on boot 5560 5561 rodata= [KNL] 5562 on Mark read-only kernel memory as read-only (default). 5563 off Leave read-only kernel memory writable for debugging. 5564 full Mark read-only kernel memory and aliases as read-only 5565 [arm64] 5566 5567 rockchip.usb_uart 5568 Enable the uart passthrough on the designated usb port 5569 on Rockchip SoCs. When active, the signals of the 5570 debug-uart get routed to the D+ and D- pins of the usb 5571 port and the regular usb controller gets disabled. 5572 5573 root= [KNL] Root filesystem 5574 Usually this a a block device specifier of some kind, 5575 see the early_lookup_bdev comment in 5576 block/early-lookup.c for details. 5577 Alternatively this can be "ram" for the legacy initial 5578 ramdisk, "nfs" and "cifs" for root on a network file 5579 system, or "mtd" and "ubi" for mounting from raw flash. 5580 5581 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 5582 mount the root filesystem 5583 5584 rootflags= [KNL] Set root filesystem mount option string 5585 5586 rootfstype= [KNL] Set root filesystem type 5587 5588 rootwait [KNL] Wait (indefinitely) for root device to show up. 5589 Useful for devices that are detected asynchronously 5590 (e.g. USB and MMC devices). 5591 5592 rootwait= [KNL] Maximum time (in seconds) to wait for root device 5593 to show up before attempting to mount the root 5594 filesystem. 5595 5596 rproc_mem=nn[KMG][@address] 5597 [KNL,ARM,CMA] Remoteproc physical memory block. 5598 Memory area to be used by remote processor image, 5599 managed by CMA. 5600 5601 rw [KNL] Mount root device read-write on boot 5602 5603 S [KNL] Run init in single mode 5604 5605 s390_iommu= [HW,S390] 5606 Set s390 IOTLB flushing mode 5607 strict 5608 With strict flushing every unmap operation will result in 5609 an IOTLB flush. Default is lazy flushing before reuse, 5610 which is faster. 5611 5612 s390_iommu_aperture= [KNL,S390] 5613 Specifies the size of the per device DMA address space 5614 accessible through the DMA and IOMMU APIs as a decimal 5615 factor of the size of main memory. 5616 The default is 1 meaning that one can concurrently use 5617 as many DMA addresses as physical memory is installed, 5618 if supported by hardware, and thus map all of memory 5619 once. With a value of 2 one can map all of memory twice 5620 and so on. As a special case a factor of 0 imposes no 5621 restrictions other than those given by hardware at the 5622 cost of significant additional memory use for tables. 5623 5624 sa1100ir [NET] 5625 See drivers/net/irda/sa1100_ir.c. 5626 5627 sched_verbose [KNL] Enables verbose scheduler debug messages. 5628 5629 schedstats= [KNL,X86] Enable or disable scheduled statistics. 5630 Allowed values are enable and disable. This feature 5631 incurs a small amount of overhead in the scheduler 5632 but is useful for debugging and performance tuning. 5633 5634 sched_thermal_decay_shift= 5635 [KNL, SMP] Set a decay shift for scheduler thermal 5636 pressure signal. Thermal pressure signal follows the 5637 default decay period of other scheduler pelt 5638 signals(usually 32 ms but configurable). Setting 5639 sched_thermal_decay_shift will left shift the decay 5640 period for the thermal pressure signal by the shift 5641 value. 5642 i.e. with the default pelt decay period of 32 ms 5643 sched_thermal_decay_shift thermal pressure decay pr 5644 1 64 ms 5645 2 128 ms 5646 and so on. 5647 Format: integer between 0 and 10 5648 Default is 0. 5649 5650 scftorture.holdoff= [KNL] 5651 Number of seconds to hold off before starting 5652 test. Defaults to zero for module insertion and 5653 to 10 seconds for built-in smp_call_function() 5654 tests. 5655 5656 scftorture.longwait= [KNL] 5657 Request ridiculously long waits randomly selected 5658 up to the chosen limit in seconds. Zero (the 5659 default) disables this feature. Please note 5660 that requesting even small non-zero numbers of 5661 seconds can result in RCU CPU stall warnings, 5662 softlockup complaints, and so on. 5663 5664 scftorture.nthreads= [KNL] 5665 Number of kthreads to spawn to invoke the 5666 smp_call_function() family of functions. 5667 The default of -1 specifies a number of kthreads 5668 equal to the number of CPUs. 5669 5670 scftorture.onoff_holdoff= [KNL] 5671 Number seconds to wait after the start of the 5672 test before initiating CPU-hotplug operations. 5673 5674 scftorture.onoff_interval= [KNL] 5675 Number seconds to wait between successive 5676 CPU-hotplug operations. Specifying zero (which 5677 is the default) disables CPU-hotplug operations. 5678 5679 scftorture.shutdown_secs= [KNL] 5680 The number of seconds following the start of the 5681 test after which to shut down the system. The 5682 default of zero avoids shutting down the system. 5683 Non-zero values are useful for automated tests. 5684 5685 scftorture.stat_interval= [KNL] 5686 The number of seconds between outputting the 5687 current test statistics to the console. A value 5688 of zero disables statistics output. 5689 5690 scftorture.stutter_cpus= [KNL] 5691 The number of jiffies to wait between each change 5692 to the set of CPUs under test. 5693 5694 scftorture.use_cpus_read_lock= [KNL] 5695 Use use_cpus_read_lock() instead of the default 5696 preempt_disable() to disable CPU hotplug 5697 while invoking one of the smp_call_function*() 5698 functions. 5699 5700 scftorture.verbose= [KNL] 5701 Enable additional printk() statements. 5702 5703 scftorture.weight_single= [KNL] 5704 The probability weighting to use for the 5705 smp_call_function_single() function with a zero 5706 "wait" parameter. A value of -1 selects the 5707 default if all other weights are -1. However, 5708 if at least one weight has some other value, a 5709 value of -1 will instead select a weight of zero. 5710 5711 scftorture.weight_single_wait= [KNL] 5712 The probability weighting to use for the 5713 smp_call_function_single() function with a 5714 non-zero "wait" parameter. See weight_single. 5715 5716 scftorture.weight_many= [KNL] 5717 The probability weighting to use for the 5718 smp_call_function_many() function with a zero 5719 "wait" parameter. See weight_single. 5720 Note well that setting a high probability for 5721 this weighting can place serious IPI load 5722 on the system. 5723 5724 scftorture.weight_many_wait= [KNL] 5725 The probability weighting to use for the 5726 smp_call_function_many() function with a 5727 non-zero "wait" parameter. See weight_single 5728 and weight_many. 5729 5730 scftorture.weight_all= [KNL] 5731 The probability weighting to use for the 5732 smp_call_function_all() function with a zero 5733 "wait" parameter. See weight_single and 5734 weight_many. 5735 5736 scftorture.weight_all_wait= [KNL] 5737 The probability weighting to use for the 5738 smp_call_function_all() function with a 5739 non-zero "wait" parameter. See weight_single 5740 and weight_many. 5741 5742 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 5743 xtime_lock contention on larger systems, and/or RCU lock 5744 contention on all systems with CONFIG_MAXSMP set. 5745 Format: { "0" | "1" } 5746 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 5747 1 -- enable. 5748 Note: increases power consumption, thus should only be 5749 enabled if running jitter sensitive (HPC/RT) workloads. 5750 5751 security= [SECURITY] Choose a legacy "major" security module to 5752 enable at boot. This has been deprecated by the 5753 "lsm=" parameter. 5754 5755 selinux= [SELINUX] Disable or enable SELinux at boot time. 5756 Format: { "0" | "1" } 5757 See security/selinux/Kconfig help text. 5758 0 -- disable. 5759 1 -- enable. 5760 Default value is 1. 5761 5762 serialnumber [BUGS=X86-32] 5763 5764 sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 5765 5766 shapers= [NET] 5767 Maximal number of shapers. 5768 5769 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 5770 Limit apic dumping. The parameter defines the maximal 5771 number of local apics being dumped. Also it is possible 5772 to set it to "all" by meaning -- no limit here. 5773 Format: { 1 (default) | 2 | ... | all }. 5774 The parameter valid if only apic=debug or 5775 apic=verbose is specified. 5776 Example: apic=debug show_lapic=all 5777 5778 simeth= [IA-64] 5779 simscsi= 5780 5781 slram= [HW,MTD] 5782 5783 slab_merge [MM] 5784 Enable merging of slabs with similar size when the 5785 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 5786 5787 slab_nomerge [MM] 5788 Disable merging of slabs with similar size. May be 5789 necessary if there is some reason to distinguish 5790 allocs to different slabs, especially in hardened 5791 environments where the risk of heap overflows and 5792 layout control by attackers can usually be 5793 frustrated by disabling merging. This will reduce 5794 most of the exposure of a heap attack to a single 5795 cache (risks via metadata attacks are mostly 5796 unchanged). Debug options disable merging on their 5797 own. 5798 For more information see Documentation/mm/slub.rst. 5799 5800 slab_max_order= [MM, SLAB] 5801 Determines the maximum allowed order for slabs. 5802 A high setting may cause OOMs due to memory 5803 fragmentation. Defaults to 1 for systems with 5804 more than 32MB of RAM, 0 otherwise. 5805 5806 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB] 5807 Enabling slub_debug allows one to determine the 5808 culprit if slab objects become corrupted. Enabling 5809 slub_debug can create guard zones around objects and 5810 may poison objects when not in use. Also tracks the 5811 last alloc / free. For more information see 5812 Documentation/mm/slub.rst. 5813 5814 slub_max_order= [MM, SLUB] 5815 Determines the maximum allowed order for slabs. 5816 A high setting may cause OOMs due to memory 5817 fragmentation. For more information see 5818 Documentation/mm/slub.rst. 5819 5820 slub_min_objects= [MM, SLUB] 5821 The minimum number of objects per slab. SLUB will 5822 increase the slab order up to slub_max_order to 5823 generate a sufficiently large slab able to contain 5824 the number of objects indicated. The higher the number 5825 of objects the smaller the overhead of tracking slabs 5826 and the less frequently locks need to be acquired. 5827 For more information see Documentation/mm/slub.rst. 5828 5829 slub_min_order= [MM, SLUB] 5830 Determines the minimum page order for slabs. Must be 5831 lower than slub_max_order. 5832 For more information see Documentation/mm/slub.rst. 5833 5834 slub_merge [MM, SLUB] 5835 Same with slab_merge. 5836 5837 slub_nomerge [MM, SLUB] 5838 Same with slab_nomerge. This is supported for legacy. 5839 See slab_nomerge for more information. 5840 5841 smart2= [HW] 5842 Format: <io1>[,<io2>[,...,<io8>]] 5843 5844 smp.csd_lock_timeout= [KNL] 5845 Specify the period of time in milliseconds 5846 that smp_call_function() and friends will wait 5847 for a CPU to release the CSD lock. This is 5848 useful when diagnosing bugs involving CPUs 5849 disabling interrupts for extended periods 5850 of time. Defaults to 5,000 milliseconds, and 5851 setting a value of zero disables this feature. 5852 This feature may be more efficiently disabled 5853 using the csdlock_debug- kernel parameter. 5854 5855 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 5856 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 5857 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 5858 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 5859 smsc-ircc2.ircc_irq= [HW] IRQ line 5860 smsc-ircc2.ircc_dma= [HW] DMA channel 5861 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 5862 0: Toshiba Satellite 1800 (GP data pin select) 5863 1: Fast pin select (default) 5864 2: ATC IRMode 5865 5866 smt= [KNL,MIPS,S390] Set the maximum number of threads (logical 5867 CPUs) to use per physical CPU on systems capable of 5868 symmetric multithreading (SMT). Will be capped to the 5869 actual hardware limit. 5870 Format: <integer> 5871 Default: -1 (no limit) 5872 5873 softlockup_panic= 5874 [KNL] Should the soft-lockup detector generate panics. 5875 Format: 0 | 1 5876 5877 A value of 1 instructs the soft-lockup detector 5878 to panic the machine when a soft-lockup occurs. It is 5879 also controlled by the kernel.softlockup_panic sysctl 5880 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 5881 respective build-time switch to that functionality. 5882 5883 softlockup_all_cpu_backtrace= 5884 [KNL] Should the soft-lockup detector generate 5885 backtraces on all cpus. 5886 Format: 0 | 1 5887 5888 sonypi.*= [HW] Sony Programmable I/O Control Device driver 5889 See Documentation/admin-guide/laptops/sonypi.rst 5890 5891 spectre_v2= [X86] Control mitigation of Spectre variant 2 5892 (indirect branch speculation) vulnerability. 5893 The default operation protects the kernel from 5894 user space attacks. 5895 5896 on - unconditionally enable, implies 5897 spectre_v2_user=on 5898 off - unconditionally disable, implies 5899 spectre_v2_user=off 5900 auto - kernel detects whether your CPU model is 5901 vulnerable 5902 5903 Selecting 'on' will, and 'auto' may, choose a 5904 mitigation method at run time according to the 5905 CPU, the available microcode, the setting of the 5906 CONFIG_RETPOLINE configuration option, and the 5907 compiler with which the kernel was built. 5908 5909 Selecting 'on' will also enable the mitigation 5910 against user space to user space task attacks. 5911 5912 Selecting 'off' will disable both the kernel and 5913 the user space protections. 5914 5915 Specific mitigations can also be selected manually: 5916 5917 retpoline - replace indirect branches 5918 retpoline,generic - Retpolines 5919 retpoline,lfence - LFENCE; indirect branch 5920 retpoline,amd - alias for retpoline,lfence 5921 eibrs - Enhanced/Auto IBRS 5922 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 5923 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 5924 ibrs - use IBRS to protect kernel 5925 5926 Not specifying this option is equivalent to 5927 spectre_v2=auto. 5928 5929 spectre_v2_user= 5930 [X86] Control mitigation of Spectre variant 2 5931 (indirect branch speculation) vulnerability between 5932 user space tasks 5933 5934 on - Unconditionally enable mitigations. Is 5935 enforced by spectre_v2=on 5936 5937 off - Unconditionally disable mitigations. Is 5938 enforced by spectre_v2=off 5939 5940 prctl - Indirect branch speculation is enabled, 5941 but mitigation can be enabled via prctl 5942 per thread. The mitigation control state 5943 is inherited on fork. 5944 5945 prctl,ibpb 5946 - Like "prctl" above, but only STIBP is 5947 controlled per thread. IBPB is issued 5948 always when switching between different user 5949 space processes. 5950 5951 seccomp 5952 - Same as "prctl" above, but all seccomp 5953 threads will enable the mitigation unless 5954 they explicitly opt out. 5955 5956 seccomp,ibpb 5957 - Like "seccomp" above, but only STIBP is 5958 controlled per thread. IBPB is issued 5959 always when switching between different 5960 user space processes. 5961 5962 auto - Kernel selects the mitigation depending on 5963 the available CPU features and vulnerability. 5964 5965 Default mitigation: "prctl" 5966 5967 Not specifying this option is equivalent to 5968 spectre_v2_user=auto. 5969 5970 spec_rstack_overflow= 5971 [X86] Control RAS overflow mitigation on AMD Zen CPUs 5972 5973 off - Disable mitigation 5974 microcode - Enable microcode mitigation only 5975 safe-ret - Enable sw-only safe RET mitigation (default) 5976 ibpb - Enable mitigation by issuing IBPB on 5977 kernel entry 5978 ibpb-vmexit - Issue IBPB only on VMEXIT 5979 (cloud-specific mitigation) 5980 5981 spec_store_bypass_disable= 5982 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 5983 (Speculative Store Bypass vulnerability) 5984 5985 Certain CPUs are vulnerable to an exploit against a 5986 a common industry wide performance optimization known 5987 as "Speculative Store Bypass" in which recent stores 5988 to the same memory location may not be observed by 5989 later loads during speculative execution. The idea 5990 is that such stores are unlikely and that they can 5991 be detected prior to instruction retirement at the 5992 end of a particular speculation execution window. 5993 5994 In vulnerable processors, the speculatively forwarded 5995 store can be used in a cache side channel attack, for 5996 example to read memory to which the attacker does not 5997 directly have access (e.g. inside sandboxed code). 5998 5999 This parameter controls whether the Speculative Store 6000 Bypass optimization is used. 6001 6002 On x86 the options are: 6003 6004 on - Unconditionally disable Speculative Store Bypass 6005 off - Unconditionally enable Speculative Store Bypass 6006 auto - Kernel detects whether the CPU model contains an 6007 implementation of Speculative Store Bypass and 6008 picks the most appropriate mitigation. If the 6009 CPU is not vulnerable, "off" is selected. If the 6010 CPU is vulnerable the default mitigation is 6011 architecture and Kconfig dependent. See below. 6012 prctl - Control Speculative Store Bypass per thread 6013 via prctl. Speculative Store Bypass is enabled 6014 for a process by default. The state of the control 6015 is inherited on fork. 6016 seccomp - Same as "prctl" above, but all seccomp threads 6017 will disable SSB unless they explicitly opt out. 6018 6019 Default mitigations: 6020 X86: "prctl" 6021 6022 On powerpc the options are: 6023 6024 on,auto - On Power8 and Power9 insert a store-forwarding 6025 barrier on kernel entry and exit. On Power7 6026 perform a software flush on kernel entry and 6027 exit. 6028 off - No action. 6029 6030 Not specifying this option is equivalent to 6031 spec_store_bypass_disable=auto. 6032 6033 spia_io_base= [HW,MTD] 6034 spia_fio_base= 6035 spia_pedr= 6036 spia_peddr= 6037 6038 split_lock_detect= 6039 [X86] Enable split lock detection or bus lock detection 6040 6041 When enabled (and if hardware support is present), atomic 6042 instructions that access data across cache line 6043 boundaries will result in an alignment check exception 6044 for split lock detection or a debug exception for 6045 bus lock detection. 6046 6047 off - not enabled 6048 6049 warn - the kernel will emit rate-limited warnings 6050 about applications triggering the #AC 6051 exception or the #DB exception. This mode is 6052 the default on CPUs that support split lock 6053 detection or bus lock detection. Default 6054 behavior is by #AC if both features are 6055 enabled in hardware. 6056 6057 fatal - the kernel will send SIGBUS to applications 6058 that trigger the #AC exception or the #DB 6059 exception. Default behavior is by #AC if 6060 both features are enabled in hardware. 6061 6062 ratelimit:N - 6063 Set system wide rate limit to N bus locks 6064 per second for bus lock detection. 6065 0 < N <= 1000. 6066 6067 N/A for split lock detection. 6068 6069 6070 If an #AC exception is hit in the kernel or in 6071 firmware (i.e. not while executing in user mode) 6072 the kernel will oops in either "warn" or "fatal" 6073 mode. 6074 6075 #DB exception for bus lock is triggered only when 6076 CPL > 0. 6077 6078 srbds= [X86,INTEL] 6079 Control the Special Register Buffer Data Sampling 6080 (SRBDS) mitigation. 6081 6082 Certain CPUs are vulnerable to an MDS-like 6083 exploit which can leak bits from the random 6084 number generator. 6085 6086 By default, this issue is mitigated by 6087 microcode. However, the microcode fix can cause 6088 the RDRAND and RDSEED instructions to become 6089 much slower. Among other effects, this will 6090 result in reduced throughput from /dev/urandom. 6091 6092 The microcode mitigation can be disabled with 6093 the following option: 6094 6095 off: Disable mitigation and remove 6096 performance impact to RDRAND and RDSEED 6097 6098 srcutree.big_cpu_lim [KNL] 6099 Specifies the number of CPUs constituting a 6100 large system, such that srcu_struct structures 6101 should immediately allocate an srcu_node array. 6102 This kernel-boot parameter defaults to 128, 6103 but takes effect only when the low-order four 6104 bits of srcutree.convert_to_big is equal to 3 6105 (decide at boot). 6106 6107 srcutree.convert_to_big [KNL] 6108 Specifies under what conditions an SRCU tree 6109 srcu_struct structure will be converted to big 6110 form, that is, with an rcu_node tree: 6111 6112 0: Never. 6113 1: At init_srcu_struct() time. 6114 2: When rcutorture decides to. 6115 3: Decide at boot time (default). 6116 0x1X: Above plus if high contention. 6117 6118 Either way, the srcu_node tree will be sized based 6119 on the actual runtime number of CPUs (nr_cpu_ids) 6120 instead of the compile-time CONFIG_NR_CPUS. 6121 6122 srcutree.counter_wrap_check [KNL] 6123 Specifies how frequently to check for 6124 grace-period sequence counter wrap for the 6125 srcu_data structure's ->srcu_gp_seq_needed field. 6126 The greater the number of bits set in this kernel 6127 parameter, the less frequently counter wrap will 6128 be checked for. Note that the bottom two bits 6129 are ignored. 6130 6131 srcutree.exp_holdoff [KNL] 6132 Specifies how many nanoseconds must elapse 6133 since the end of the last SRCU grace period for 6134 a given srcu_struct until the next normal SRCU 6135 grace period will be considered for automatic 6136 expediting. Set to zero to disable automatic 6137 expediting. 6138 6139 srcutree.srcu_max_nodelay [KNL] 6140 Specifies the number of no-delay instances 6141 per jiffy for which the SRCU grace period 6142 worker thread will be rescheduled with zero 6143 delay. Beyond this limit, worker thread will 6144 be rescheduled with a sleep delay of one jiffy. 6145 6146 srcutree.srcu_max_nodelay_phase [KNL] 6147 Specifies the per-grace-period phase, number of 6148 non-sleeping polls of readers. Beyond this limit, 6149 grace period worker thread will be rescheduled 6150 with a sleep delay of one jiffy, between each 6151 rescan of the readers, for a grace period phase. 6152 6153 srcutree.srcu_retry_check_delay [KNL] 6154 Specifies number of microseconds of non-sleeping 6155 delay between each non-sleeping poll of readers. 6156 6157 srcutree.small_contention_lim [KNL] 6158 Specifies the number of update-side contention 6159 events per jiffy will be tolerated before 6160 initiating a conversion of an srcu_struct 6161 structure to big form. Note that the value of 6162 srcutree.convert_to_big must have the 0x10 bit 6163 set for contention-based conversions to occur. 6164 6165 ssbd= [ARM64,HW] 6166 Speculative Store Bypass Disable control 6167 6168 On CPUs that are vulnerable to the Speculative 6169 Store Bypass vulnerability and offer a 6170 firmware based mitigation, this parameter 6171 indicates how the mitigation should be used: 6172 6173 force-on: Unconditionally enable mitigation for 6174 for both kernel and userspace 6175 force-off: Unconditionally disable mitigation for 6176 for both kernel and userspace 6177 kernel: Always enable mitigation in the 6178 kernel, and offer a prctl interface 6179 to allow userspace to register its 6180 interest in being mitigated too. 6181 6182 stack_guard_gap= [MM] 6183 override the default stack gap protection. The value 6184 is in page units and it defines how many pages prior 6185 to (for stacks growing down) resp. after (for stacks 6186 growing up) the main stack are reserved for no other 6187 mapping. Default value is 256 pages. 6188 6189 stack_depot_disable= [KNL] 6190 Setting this to true through kernel command line will 6191 disable the stack depot thereby saving the static memory 6192 consumed by the stack hash table. By default this is set 6193 to false. 6194 6195 stacktrace [FTRACE] 6196 Enabled the stack tracer on boot up. 6197 6198 stacktrace_filter=[function-list] 6199 [FTRACE] Limit the functions that the stack tracer 6200 will trace at boot up. function-list is a comma-separated 6201 list of functions. This list can be changed at run 6202 time by the stack_trace_filter file in the debugfs 6203 tracing directory. Note, this enables stack tracing 6204 and the stacktrace above is not needed. 6205 6206 sti= [PARISC,HW] 6207 Format: <num> 6208 Set the STI (builtin display/keyboard on the HP-PARISC 6209 machines) console (graphic card) which should be used 6210 as the initial boot-console. 6211 See also comment in drivers/video/console/sticore.c. 6212 6213 sti_font= [HW] 6214 See comment in drivers/video/console/sticore.c. 6215 6216 stifb= [HW] 6217 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 6218 6219 strict_sas_size= 6220 [X86] 6221 Format: <bool> 6222 Enable or disable strict sigaltstack size checks 6223 against the required signal frame size which 6224 depends on the supported FPU features. This can 6225 be used to filter out binaries which have 6226 not yet been made aware of AT_MINSIGSTKSZ. 6227 6228 stress_hpt [PPC] 6229 Limits the number of kernel HPT entries in the hash 6230 page table to increase the rate of hash page table 6231 faults on kernel addresses. 6232 6233 stress_slb [PPC] 6234 Limits the number of kernel SLB entries, and flushes 6235 them frequently to increase the rate of SLB faults 6236 on kernel addresses. 6237 6238 sunrpc.min_resvport= 6239 sunrpc.max_resvport= 6240 [NFS,SUNRPC] 6241 SunRPC servers often require that client requests 6242 originate from a privileged port (i.e. a port in the 6243 range 0 < portnr < 1024). 6244 An administrator who wishes to reserve some of these 6245 ports for other uses may adjust the range that the 6246 kernel's sunrpc client considers to be privileged 6247 using these two parameters to set the minimum and 6248 maximum port values. 6249 6250 sunrpc.svc_rpc_per_connection_limit= 6251 [NFS,SUNRPC] 6252 Limit the number of requests that the server will 6253 process in parallel from a single connection. 6254 The default value is 0 (no limit). 6255 6256 sunrpc.pool_mode= 6257 [NFS] 6258 Control how the NFS server code allocates CPUs to 6259 service thread pools. Depending on how many NICs 6260 you have and where their interrupts are bound, this 6261 option will affect which CPUs will do NFS serving. 6262 Note: this parameter cannot be changed while the 6263 NFS server is running. 6264 6265 auto the server chooses an appropriate mode 6266 automatically using heuristics 6267 global a single global pool contains all CPUs 6268 percpu one pool for each CPU 6269 pernode one pool for each NUMA node (equivalent 6270 to global on non-NUMA machines) 6271 6272 sunrpc.tcp_slot_table_entries= 6273 sunrpc.udp_slot_table_entries= 6274 [NFS,SUNRPC] 6275 Sets the upper limit on the number of simultaneous 6276 RPC calls that can be sent from the client to a 6277 server. Increasing these values may allow you to 6278 improve throughput, but will also increase the 6279 amount of memory reserved for use by the client. 6280 6281 suspend.pm_test_delay= 6282 [SUSPEND] 6283 Sets the number of seconds to remain in a suspend test 6284 mode before resuming the system (see 6285 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 6286 is set. Default value is 5. 6287 6288 svm= [PPC] 6289 Format: { on | off | y | n | 1 | 0 } 6290 This parameter controls use of the Protected 6291 Execution Facility on pSeries. 6292 6293 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 6294 Format: { <int> [,<int>] | force | noforce } 6295 <int> -- Number of I/O TLB slabs 6296 <int> -- Second integer after comma. Number of swiotlb 6297 areas with their own lock. Will be rounded up 6298 to a power of 2. 6299 force -- force using of bounce buffers even if they 6300 wouldn't be automatically used by the kernel 6301 noforce -- Never use bounce buffers (for debugging) 6302 6303 switches= [HW,M68k] 6304 6305 sysctl.*= [KNL] 6306 Set a sysctl parameter, right before loading the init 6307 process, as if the value was written to the respective 6308 /proc/sys/... file. Both '.' and '/' are recognized as 6309 separators. Unrecognized parameters and invalid values 6310 are reported in the kernel log. Sysctls registered 6311 later by a loaded module cannot be set this way. 6312 Example: sysctl.vm.swappiness=40 6313 6314 sysrq_always_enabled 6315 [KNL] 6316 Ignore sysrq setting - this boot parameter will 6317 neutralize any effect of /proc/sys/kernel/sysrq. 6318 Useful for debugging. 6319 6320 tcpmhash_entries= [KNL,NET] 6321 Set the number of tcp_metrics_hash slots. 6322 Default value is 8192 or 16384 depending on total 6323 ram pages. This is used to specify the TCP metrics 6324 cache size. See Documentation/networking/ip-sysctl.rst 6325 "tcp_no_metrics_save" section for more details. 6326 6327 tdfx= [HW,DRM] 6328 6329 test_suspend= [SUSPEND] 6330 Format: { "mem" | "standby" | "freeze" }[,N] 6331 Specify "mem" (for Suspend-to-RAM) or "standby" (for 6332 standby suspend) or "freeze" (for suspend type freeze) 6333 as the system sleep state during system startup with 6334 the optional capability to repeat N number of times. 6335 The system is woken from this state using a 6336 wakeup-capable RTC alarm. 6337 6338 thash_entries= [KNL,NET] 6339 Set number of hash buckets for TCP connection 6340 6341 thermal.act= [HW,ACPI] 6342 -1: disable all active trip points in all thermal zones 6343 <degrees C>: override all lowest active trip points 6344 6345 thermal.crt= [HW,ACPI] 6346 -1: disable all critical trip points in all thermal zones 6347 <degrees C>: override all critical trip points 6348 6349 thermal.off= [HW,ACPI] 6350 1: disable ACPI thermal control 6351 6352 thermal.psv= [HW,ACPI] 6353 -1: disable all passive trip points 6354 <degrees C>: override all passive trip points to this 6355 value 6356 6357 thermal.tzp= [HW,ACPI] 6358 Specify global default ACPI thermal zone polling rate 6359 <deci-seconds>: poll all this frequency 6360 0: no polling (default) 6361 6362 threadirqs [KNL] 6363 Force threading of all interrupt handlers except those 6364 marked explicitly IRQF_NO_THREAD. 6365 6366 topology= [S390] 6367 Format: {off | on} 6368 Specify if the kernel should make use of the cpu 6369 topology information if the hardware supports this. 6370 The scheduler will make use of this information and 6371 e.g. base its process migration decisions on it. 6372 Default is on. 6373 6374 topology_updates= [KNL, PPC, NUMA] 6375 Format: {off} 6376 Specify if the kernel should ignore (off) 6377 topology updates sent by the hypervisor to this 6378 LPAR. 6379 6380 torture.disable_onoff_at_boot= [KNL] 6381 Prevent the CPU-hotplug component of torturing 6382 until after init has spawned. 6383 6384 torture.ftrace_dump_at_shutdown= [KNL] 6385 Dump the ftrace buffer at torture-test shutdown, 6386 even if there were no errors. This can be a 6387 very costly operation when many torture tests 6388 are running concurrently, especially on systems 6389 with rotating-rust storage. 6390 6391 torture.verbose_sleep_frequency= [KNL] 6392 Specifies how many verbose printk()s should be 6393 emitted between each sleep. The default of zero 6394 disables verbose-printk() sleeping. 6395 6396 torture.verbose_sleep_duration= [KNL] 6397 Duration of each verbose-printk() sleep in jiffies. 6398 6399 tp720= [HW,PS2] 6400 6401 tpm_suspend_pcr=[HW,TPM] 6402 Format: integer pcr id 6403 Specify that at suspend time, the tpm driver 6404 should extend the specified pcr with zeros, 6405 as a workaround for some chips which fail to 6406 flush the last written pcr on TPM_SaveState. 6407 This will guarantee that all the other pcrs 6408 are saved. 6409 6410 tpm_tis.interrupts= [HW,TPM] 6411 Enable interrupts for the MMIO based physical layer 6412 for the FIFO interface. By default it is set to false 6413 (0). For more information about TPM hardware interfaces 6414 defined by Trusted Computing Group (TCG) see 6415 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 6416 6417 tp_printk [FTRACE] 6418 Have the tracepoints sent to printk as well as the 6419 tracing ring buffer. This is useful for early boot up 6420 where the system hangs or reboots and does not give the 6421 option for reading the tracing buffer or performing a 6422 ftrace_dump_on_oops. 6423 6424 To turn off having tracepoints sent to printk, 6425 echo 0 > /proc/sys/kernel/tracepoint_printk 6426 Note, echoing 1 into this file without the 6427 tracepoint_printk kernel cmdline option has no effect. 6428 6429 The tp_printk_stop_on_boot (see below) can also be used 6430 to stop the printing of events to console at 6431 late_initcall_sync. 6432 6433 ** CAUTION ** 6434 6435 Having tracepoints sent to printk() and activating high 6436 frequency tracepoints such as irq or sched, can cause 6437 the system to live lock. 6438 6439 tp_printk_stop_on_boot [FTRACE] 6440 When tp_printk (above) is set, it can cause a lot of noise 6441 on the console. It may be useful to only include the 6442 printing of events during boot up, as user space may 6443 make the system inoperable. 6444 6445 This command line option will stop the printing of events 6446 to console at the late_initcall_sync() time frame. 6447 6448 trace_buf_size=nn[KMG] 6449 [FTRACE] will set tracing buffer size on each cpu. 6450 6451 trace_clock= [FTRACE] Set the clock used for tracing events 6452 at boot up. 6453 local - Use the per CPU time stamp counter 6454 (converted into nanoseconds). Fast, but 6455 depending on the architecture, may not be 6456 in sync between CPUs. 6457 global - Event time stamps are synchronize across 6458 CPUs. May be slower than the local clock, 6459 but better for some race conditions. 6460 counter - Simple counting of events (1, 2, ..) 6461 note, some counts may be skipped due to the 6462 infrastructure grabbing the clock more than 6463 once per event. 6464 uptime - Use jiffies as the time stamp. 6465 perf - Use the same clock that perf uses. 6466 mono - Use ktime_get_mono_fast_ns() for time stamps. 6467 mono_raw - Use ktime_get_raw_fast_ns() for time 6468 stamps. 6469 boot - Use ktime_get_boot_fast_ns() for time stamps. 6470 Architectures may add more clocks. See 6471 Documentation/trace/ftrace.rst for more details. 6472 6473 trace_event=[event-list] 6474 [FTRACE] Set and start specified trace events in order 6475 to facilitate early boot debugging. The event-list is a 6476 comma-separated list of trace events to enable. See 6477 also Documentation/trace/events.rst 6478 6479 trace_instance=[instance-info] 6480 [FTRACE] Create a ring buffer instance early in boot up. 6481 This will be listed in: 6482 6483 /sys/kernel/tracing/instances 6484 6485 Events can be enabled at the time the instance is created 6486 via: 6487 6488 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 6489 6490 Note, the "<system*>:" portion is optional if the event is 6491 unique. 6492 6493 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 6494 6495 will enable the "sched_switch" event (note, the "sched:" is optional, and 6496 the same thing would happen if it was left off). The irq_handler_entry 6497 event, and all events under the "initcall" system. 6498 6499 trace_options=[option-list] 6500 [FTRACE] Enable or disable tracer options at boot. 6501 The option-list is a comma delimited list of options 6502 that can be enabled or disabled just as if you were 6503 to echo the option name into 6504 6505 /sys/kernel/tracing/trace_options 6506 6507 For example, to enable stacktrace option (to dump the 6508 stack trace of each event), add to the command line: 6509 6510 trace_options=stacktrace 6511 6512 See also Documentation/trace/ftrace.rst "trace options" 6513 section. 6514 6515 trace_trigger=[trigger-list] 6516 [FTRACE] Add a event trigger on specific events. 6517 Set a trigger on top of a specific event, with an optional 6518 filter. 6519 6520 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..." 6521 Where more than one trigger may be specified that are comma deliminated. 6522 6523 For example: 6524 6525 trace_trigger="sched_switch.stacktrace if prev_state == 2" 6526 6527 The above will enable the "stacktrace" trigger on the "sched_switch" 6528 event but only trigger it if the "prev_state" of the "sched_switch" 6529 event is "2" (TASK_UNINTERUPTIBLE). 6530 6531 See also "Event triggers" in Documentation/trace/events.rst 6532 6533 6534 traceoff_on_warning 6535 [FTRACE] enable this option to disable tracing when a 6536 warning is hit. This turns off "tracing_on". Tracing can 6537 be enabled again by echoing '1' into the "tracing_on" 6538 file located in /sys/kernel/tracing/ 6539 6540 This option is useful, as it disables the trace before 6541 the WARNING dump is called, which prevents the trace to 6542 be filled with content caused by the warning output. 6543 6544 This option can also be set at run time via the sysctl 6545 option: kernel/traceoff_on_warning 6546 6547 transparent_hugepage= 6548 [KNL] 6549 Format: [always|madvise|never] 6550 Can be used to control the default behavior of the system 6551 with respect to transparent hugepages. 6552 See Documentation/admin-guide/mm/transhuge.rst 6553 for more details. 6554 6555 trusted.source= [KEYS] 6556 Format: <string> 6557 This parameter identifies the trust source as a backend 6558 for trusted keys implementation. Supported trust 6559 sources: 6560 - "tpm" 6561 - "tee" 6562 - "caam" 6563 If not specified then it defaults to iterating through 6564 the trust source list starting with TPM and assigns the 6565 first trust source as a backend which is initialized 6566 successfully during iteration. 6567 6568 trusted.rng= [KEYS] 6569 Format: <string> 6570 The RNG used to generate key material for trusted keys. 6571 Can be one of: 6572 - "kernel" 6573 - the same value as trusted.source: "tpm" or "tee" 6574 - "default" 6575 If not specified, "default" is used. In this case, 6576 the RNG's choice is left to each individual trust source. 6577 6578 tsc= Disable clocksource stability checks for TSC. 6579 Format: <string> 6580 [x86] reliable: mark tsc clocksource as reliable, this 6581 disables clocksource verification at runtime, as well 6582 as the stability checks done at bootup. Used to enable 6583 high-resolution timer mode on older hardware, and in 6584 virtualized environment. 6585 [x86] noirqtime: Do not use TSC to do irq accounting. 6586 Used to run time disable IRQ_TIME_ACCOUNTING on any 6587 platforms where RDTSC is slow and this accounting 6588 can add overhead. 6589 [x86] unstable: mark the TSC clocksource as unstable, this 6590 marks the TSC unconditionally unstable at bootup and 6591 avoids any further wobbles once the TSC watchdog notices. 6592 [x86] nowatchdog: disable clocksource watchdog. Used 6593 in situations with strict latency requirements (where 6594 interruptions from clocksource watchdog are not 6595 acceptable). 6596 [x86] recalibrate: force recalibration against a HW timer 6597 (HPET or PM timer) on systems whose TSC frequency was 6598 obtained from HW or FW using either an MSR or CPUID(0x15). 6599 Warn if the difference is more than 500 ppm. 6600 [x86] watchdog: Use TSC as the watchdog clocksource with 6601 which to check other HW timers (HPET or PM timer), but 6602 only on systems where TSC has been deemed trustworthy. 6603 This will be suppressed by an earlier tsc=nowatchdog and 6604 can be overridden by a later tsc=nowatchdog. A console 6605 message will flag any such suppression or overriding. 6606 6607 tsc_early_khz= [X86] Skip early TSC calibration and use the given 6608 value instead. Useful when the early TSC frequency discovery 6609 procedure is not reliable, such as on overclocked systems 6610 with CPUID.16h support and partial CPUID.15h support. 6611 Format: <unsigned int> 6612 6613 tsx= [X86] Control Transactional Synchronization 6614 Extensions (TSX) feature in Intel processors that 6615 support TSX control. 6616 6617 This parameter controls the TSX feature. The options are: 6618 6619 on - Enable TSX on the system. Although there are 6620 mitigations for all known security vulnerabilities, 6621 TSX has been known to be an accelerator for 6622 several previous speculation-related CVEs, and 6623 so there may be unknown security risks associated 6624 with leaving it enabled. 6625 6626 off - Disable TSX on the system. (Note that this 6627 option takes effect only on newer CPUs which are 6628 not vulnerable to MDS, i.e., have 6629 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 6630 the new IA32_TSX_CTRL MSR through a microcode 6631 update. This new MSR allows for the reliable 6632 deactivation of the TSX functionality.) 6633 6634 auto - Disable TSX if X86_BUG_TAA is present, 6635 otherwise enable TSX on the system. 6636 6637 Not specifying this option is equivalent to tsx=off. 6638 6639 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6640 for more details. 6641 6642 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 6643 Abort (TAA) vulnerability. 6644 6645 Similar to Micro-architectural Data Sampling (MDS) 6646 certain CPUs that support Transactional 6647 Synchronization Extensions (TSX) are vulnerable to an 6648 exploit against CPU internal buffers which can forward 6649 information to a disclosure gadget under certain 6650 conditions. 6651 6652 In vulnerable processors, the speculatively forwarded 6653 data can be used in a cache side channel attack, to 6654 access data to which the attacker does not have direct 6655 access. 6656 6657 This parameter controls the TAA mitigation. The 6658 options are: 6659 6660 full - Enable TAA mitigation on vulnerable CPUs 6661 if TSX is enabled. 6662 6663 full,nosmt - Enable TAA mitigation and disable SMT on 6664 vulnerable CPUs. If TSX is disabled, SMT 6665 is not disabled because CPU is not 6666 vulnerable to cross-thread TAA attacks. 6667 off - Unconditionally disable TAA mitigation 6668 6669 On MDS-affected machines, tsx_async_abort=off can be 6670 prevented by an active MDS mitigation as both vulnerabilities 6671 are mitigated with the same mechanism so in order to disable 6672 this mitigation, you need to specify mds=off too. 6673 6674 Not specifying this option is equivalent to 6675 tsx_async_abort=full. On CPUs which are MDS affected 6676 and deploy MDS mitigation, TAA mitigation is not 6677 required and doesn't provide any additional 6678 mitigation. 6679 6680 For details see: 6681 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6682 6683 turbografx.map[2|3]= [HW,JOY] 6684 TurboGraFX parallel port interface 6685 Format: 6686 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 6687 See also Documentation/input/devices/joystick-parport.rst 6688 6689 udbg-immortal [PPC] When debugging early kernel crashes that 6690 happen after console_init() and before a proper 6691 console driver takes over, this boot options might 6692 help "seeing" what's going on. 6693 6694 uhash_entries= [KNL,NET] 6695 Set number of hash buckets for UDP/UDP-Lite connections 6696 6697 uhci-hcd.ignore_oc= 6698 [USB] Ignore overcurrent events (default N). 6699 Some badly-designed motherboards generate lots of 6700 bogus events, for ports that aren't wired to 6701 anything. Set this parameter to avoid log spamming. 6702 Note that genuine overcurrent events won't be 6703 reported either. 6704 6705 unknown_nmi_panic 6706 [X86] Cause panic on unknown NMI. 6707 6708 unwind_debug [X86-64] 6709 Enable unwinder debug output. This can be 6710 useful for debugging certain unwinder error 6711 conditions, including corrupt stacks and 6712 bad/missing unwinder metadata. 6713 6714 usbcore.authorized_default= 6715 [USB] Default USB device authorization: 6716 (default -1 = authorized (same as 1), 6717 0 = not authorized, 1 = authorized, 2 = authorized 6718 if device connected to internal port) 6719 6720 usbcore.autosuspend= 6721 [USB] The autosuspend time delay (in seconds) used 6722 for newly-detected USB devices (default 2). This 6723 is the time required before an idle device will be 6724 autosuspended. Devices for which the delay is set 6725 to a negative value won't be autosuspended at all. 6726 6727 usbcore.usbfs_snoop= 6728 [USB] Set to log all usbfs traffic (default 0 = off). 6729 6730 usbcore.usbfs_snoop_max= 6731 [USB] Maximum number of bytes to snoop in each URB 6732 (default = 65536). 6733 6734 usbcore.blinkenlights= 6735 [USB] Set to cycle leds on hubs (default 0 = off). 6736 6737 usbcore.old_scheme_first= 6738 [USB] Start with the old device initialization 6739 scheme (default 0 = off). 6740 6741 usbcore.usbfs_memory_mb= 6742 [USB] Memory limit (in MB) for buffers allocated by 6743 usbfs (default = 16, 0 = max = 2047). 6744 6745 usbcore.use_both_schemes= 6746 [USB] Try the other device initialization scheme 6747 if the first one fails (default 1 = enabled). 6748 6749 usbcore.initial_descriptor_timeout= 6750 [USB] Specifies timeout for the initial 64-byte 6751 USB_REQ_GET_DESCRIPTOR request in milliseconds 6752 (default 5000 = 5.0 seconds). 6753 6754 usbcore.nousb [USB] Disable the USB subsystem 6755 6756 usbcore.quirks= 6757 [USB] A list of quirk entries to augment the built-in 6758 usb core quirk list. List entries are separated by 6759 commas. Each entry has the form 6760 VendorID:ProductID:Flags. The IDs are 4-digit hex 6761 numbers and Flags is a set of letters. Each letter 6762 will change the built-in quirk; setting it if it is 6763 clear and clearing it if it is set. The letters have 6764 the following meanings: 6765 a = USB_QUIRK_STRING_FETCH_255 (string 6766 descriptors must not be fetched using 6767 a 255-byte read); 6768 b = USB_QUIRK_RESET_RESUME (device can't resume 6769 correctly so reset it instead); 6770 c = USB_QUIRK_NO_SET_INTF (device can't handle 6771 Set-Interface requests); 6772 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 6773 handle its Configuration or Interface 6774 strings); 6775 e = USB_QUIRK_RESET (device can't be reset 6776 (e.g morph devices), don't use reset); 6777 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 6778 more interface descriptions than the 6779 bNumInterfaces count, and can't handle 6780 talking to these interfaces); 6781 g = USB_QUIRK_DELAY_INIT (device needs a pause 6782 during initialization, after we read 6783 the device descriptor); 6784 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 6785 high speed and super speed interrupt 6786 endpoints, the USB 2.0 and USB 3.0 spec 6787 require the interval in microframes (1 6788 microframe = 125 microseconds) to be 6789 calculated as interval = 2 ^ 6790 (bInterval-1). 6791 Devices with this quirk report their 6792 bInterval as the result of this 6793 calculation instead of the exponent 6794 variable used in the calculation); 6795 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 6796 handle device_qualifier descriptor 6797 requests); 6798 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 6799 generates spurious wakeup, ignore 6800 remote wakeup capability); 6801 k = USB_QUIRK_NO_LPM (device can't handle Link 6802 Power Management); 6803 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 6804 (Device reports its bInterval as linear 6805 frames instead of the USB 2.0 6806 calculation); 6807 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 6808 to be disconnected before suspend to 6809 prevent spurious wakeup); 6810 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 6811 pause after every control message); 6812 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 6813 delay after resetting its port); 6814 Example: quirks=0781:5580:bk,0a5c:5834:gij 6815 6816 usbhid.mousepoll= 6817 [USBHID] The interval which mice are to be polled at. 6818 6819 usbhid.jspoll= 6820 [USBHID] The interval which joysticks are to be polled at. 6821 6822 usbhid.kbpoll= 6823 [USBHID] The interval which keyboards are to be polled at. 6824 6825 usb-storage.delay_use= 6826 [UMS] The delay in seconds before a new device is 6827 scanned for Logical Units (default 1). 6828 6829 usb-storage.quirks= 6830 [UMS] A list of quirks entries to supplement or 6831 override the built-in unusual_devs list. List 6832 entries are separated by commas. Each entry has 6833 the form VID:PID:Flags where VID and PID are Vendor 6834 and Product ID values (4-digit hex numbers) and 6835 Flags is a set of characters, each corresponding 6836 to a common usb-storage quirk flag as follows: 6837 a = SANE_SENSE (collect more than 18 bytes 6838 of sense data, not on uas); 6839 b = BAD_SENSE (don't collect more than 18 6840 bytes of sense data, not on uas); 6841 c = FIX_CAPACITY (decrease the reported 6842 device capacity by one sector); 6843 d = NO_READ_DISC_INFO (don't use 6844 READ_DISC_INFO command, not on uas); 6845 e = NO_READ_CAPACITY_16 (don't use 6846 READ_CAPACITY_16 command); 6847 f = NO_REPORT_OPCODES (don't use report opcodes 6848 command, uas only); 6849 g = MAX_SECTORS_240 (don't transfer more than 6850 240 sectors at a time, uas only); 6851 h = CAPACITY_HEURISTICS (decrease the 6852 reported device capacity by one 6853 sector if the number is odd); 6854 i = IGNORE_DEVICE (don't bind to this 6855 device); 6856 j = NO_REPORT_LUNS (don't use report luns 6857 command, uas only); 6858 k = NO_SAME (do not use WRITE_SAME, uas only) 6859 l = NOT_LOCKABLE (don't try to lock and 6860 unlock ejectable media, not on uas); 6861 m = MAX_SECTORS_64 (don't transfer more 6862 than 64 sectors = 32 KB at a time, 6863 not on uas); 6864 n = INITIAL_READ10 (force a retry of the 6865 initial READ(10) command, not on uas); 6866 o = CAPACITY_OK (accept the capacity 6867 reported by the device, not on uas); 6868 p = WRITE_CACHE (the device cache is ON 6869 by default, not on uas); 6870 r = IGNORE_RESIDUE (the device reports 6871 bogus residue values, not on uas); 6872 s = SINGLE_LUN (the device has only one 6873 Logical Unit); 6874 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 6875 commands, uas only); 6876 u = IGNORE_UAS (don't bind to the uas driver); 6877 w = NO_WP_DETECT (don't test whether the 6878 medium is write-protected). 6879 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 6880 even if the device claims no cache, 6881 not on uas) 6882 Example: quirks=0419:aaf5:rl,0421:0433:rc 6883 6884 user_debug= [KNL,ARM] 6885 Format: <int> 6886 See arch/arm/Kconfig.debug help text. 6887 1 - undefined instruction events 6888 2 - system calls 6889 4 - invalid data aborts 6890 8 - SIGSEGV faults 6891 16 - SIGBUS faults 6892 Example: user_debug=31 6893 6894 userpte= 6895 [X86] Flags controlling user PTE allocations. 6896 6897 nohigh = do not allocate PTE pages in 6898 HIGHMEM regardless of setting 6899 of CONFIG_HIGHPTE. 6900 6901 vdso= [X86,SH,SPARC] 6902 On X86_32, this is an alias for vdso32=. Otherwise: 6903 6904 vdso=1: enable VDSO (the default) 6905 vdso=0: disable VDSO mapping 6906 6907 vdso32= [X86] Control the 32-bit vDSO 6908 vdso32=1: enable 32-bit VDSO 6909 vdso32=0 or vdso32=2: disable 32-bit VDSO 6910 6911 See the help text for CONFIG_COMPAT_VDSO for more 6912 details. If CONFIG_COMPAT_VDSO is set, the default is 6913 vdso32=0; otherwise, the default is vdso32=1. 6914 6915 For compatibility with older kernels, vdso32=2 is an 6916 alias for vdso32=0. 6917 6918 Try vdso32=0 if you encounter an error that says: 6919 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 6920 6921 vector= [IA-64,SMP] 6922 vector=percpu: enable percpu vector domain 6923 6924 video= [FB] Frame buffer configuration 6925 See Documentation/fb/modedb.rst. 6926 6927 video.brightness_switch_enabled= [ACPI] 6928 Format: [0|1] 6929 If set to 1, on receiving an ACPI notify event 6930 generated by hotkey, video driver will adjust brightness 6931 level and then send out the event to user space through 6932 the allocated input device. If set to 0, video driver 6933 will only send out the event without touching backlight 6934 brightness level. 6935 default: 1 6936 6937 virtio_mmio.device= 6938 [VMMIO] Memory mapped virtio (platform) device. 6939 6940 <size>@<baseaddr>:<irq>[:<id>] 6941 where: 6942 <size> := size (can use standard suffixes 6943 like K, M and G) 6944 <baseaddr> := physical base address 6945 <irq> := interrupt number (as passed to 6946 request_irq()) 6947 <id> := (optional) platform device id 6948 example: 6949 virtio_mmio.device=1K@0x100b0000:48:7 6950 6951 Can be used multiple times for multiple devices. 6952 6953 vga= [BOOT,X86-32] Select a particular video mode 6954 See Documentation/arch/x86/boot.rst and 6955 Documentation/admin-guide/svga.rst. 6956 Use vga=ask for menu. 6957 This is actually a boot loader parameter; the value is 6958 passed to the kernel using a special protocol. 6959 6960 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 6961 May slow down system boot speed, especially when 6962 enabled on systems with a large amount of memory. 6963 All options are enabled by default, and this 6964 interface is meant to allow for selectively 6965 enabling or disabling specific virtual memory 6966 debugging features. 6967 6968 Available options are: 6969 P Enable page structure init time poisoning 6970 - Disable all of the above options 6971 6972 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 6973 size of <nn>. This can be used to increase the 6974 minimum size (128MB on x86). It can also be used to 6975 decrease the size and leave more room for directly 6976 mapped kernel RAM. 6977 6978 vmcp_cma=nn[MG] [KNL,S390] 6979 Sets the memory size reserved for contiguous memory 6980 allocations for the vmcp device driver. 6981 6982 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 6983 Format: <command> 6984 6985 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 6986 Format: <command> 6987 6988 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 6989 Format: <command> 6990 6991 vsyscall= [X86-64] 6992 Controls the behavior of vsyscalls (i.e. calls to 6993 fixed addresses of 0xffffffffff600x00 from legacy 6994 code). Most statically-linked binaries and older 6995 versions of glibc use these calls. Because these 6996 functions are at fixed addresses, they make nice 6997 targets for exploits that can control RIP. 6998 6999 emulate Vsyscalls turn into traps and are emulated 7000 reasonably safely. The vsyscall page is 7001 readable. 7002 7003 xonly [default] Vsyscalls turn into traps and are 7004 emulated reasonably safely. The vsyscall 7005 page is not readable. 7006 7007 none Vsyscalls don't work at all. This makes 7008 them quite hard to use for exploits but 7009 might break your system. 7010 7011 vt.color= [VT] Default text color. 7012 Format: 0xYX, X = foreground, Y = background. 7013 Default: 0x07 = light gray on black. 7014 7015 vt.cur_default= [VT] Default cursor shape. 7016 Format: 0xCCBBAA, where AA, BB, and CC are the same as 7017 the parameters of the <Esc>[?A;B;Cc escape sequence; 7018 see VGA-softcursor.txt. Default: 2 = underline. 7019 7020 vt.default_blu= [VT] 7021 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 7022 Change the default blue palette of the console. 7023 This is a 16-member array composed of values 7024 ranging from 0-255. 7025 7026 vt.default_grn= [VT] 7027 Format: <green0>,<green1>,<green2>,...,<green15> 7028 Change the default green palette of the console. 7029 This is a 16-member array composed of values 7030 ranging from 0-255. 7031 7032 vt.default_red= [VT] 7033 Format: <red0>,<red1>,<red2>,...,<red15> 7034 Change the default red palette of the console. 7035 This is a 16-member array composed of values 7036 ranging from 0-255. 7037 7038 vt.default_utf8= 7039 [VT] 7040 Format=<0|1> 7041 Set system-wide default UTF-8 mode for all tty's. 7042 Default is 1, i.e. UTF-8 mode is enabled for all 7043 newly opened terminals. 7044 7045 vt.global_cursor_default= 7046 [VT] 7047 Format=<-1|0|1> 7048 Set system-wide default for whether a cursor 7049 is shown on new VTs. Default is -1, 7050 i.e. cursors will be created by default unless 7051 overridden by individual drivers. 0 will hide 7052 cursors, 1 will display them. 7053 7054 vt.italic= [VT] Default color for italic text; 0-15. 7055 Default: 2 = green. 7056 7057 vt.underline= [VT] Default color for underlined text; 0-15. 7058 Default: 3 = cyan. 7059 7060 watchdog timers [HW,WDT] For information on watchdog timers, 7061 see Documentation/watchdog/watchdog-parameters.rst 7062 or other driver-specific files in the 7063 Documentation/watchdog/ directory. 7064 7065 watchdog_thresh= 7066 [KNL] 7067 Set the hard lockup detector stall duration 7068 threshold in seconds. The soft lockup detector 7069 threshold is set to twice the value. A value of 0 7070 disables both lockup detectors. Default is 10 7071 seconds. 7072 7073 workqueue.unbound_cpus= 7074 [KNL,SMP] Specify to constrain one or some CPUs 7075 to use in unbound workqueues. 7076 Format: <cpu-list> 7077 By default, all online CPUs are available for 7078 unbound workqueues. 7079 7080 workqueue.watchdog_thresh= 7081 If CONFIG_WQ_WATCHDOG is configured, workqueue can 7082 warn stall conditions and dump internal state to 7083 help debugging. 0 disables workqueue stall 7084 detection; otherwise, it's the stall threshold 7085 duration in seconds. The default value is 30 and 7086 it can be updated at runtime by writing to the 7087 corresponding sysfs file. 7088 7089 workqueue.cpu_intensive_thresh_us= 7090 Per-cpu work items which run for longer than this 7091 threshold are automatically considered CPU intensive 7092 and excluded from concurrency management to prevent 7093 them from noticeably delaying other per-cpu work 7094 items. Default is 10000 (10ms). 7095 7096 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 7097 will report the work functions which violate this 7098 threshold repeatedly. They are likely good 7099 candidates for using WQ_UNBOUND workqueues instead. 7100 7101 workqueue.power_efficient 7102 Per-cpu workqueues are generally preferred because 7103 they show better performance thanks to cache 7104 locality; unfortunately, per-cpu workqueues tend to 7105 be more power hungry than unbound workqueues. 7106 7107 Enabling this makes the per-cpu workqueues which 7108 were observed to contribute significantly to power 7109 consumption unbound, leading to measurably lower 7110 power usage at the cost of small performance 7111 overhead. 7112 7113 The default value of this parameter is determined by 7114 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 7115 7116 workqueue.default_affinity_scope= 7117 Select the default affinity scope to use for unbound 7118 workqueues. Can be one of "cpu", "smt", "cache", 7119 "numa" and "system". Default is "cache". For more 7120 information, see the Affinity Scopes section in 7121 Documentation/core-api/workqueue.rst. 7122 7123 This can be changed after boot by writing to the 7124 matching /sys/module/workqueue/parameters file. All 7125 workqueues with the "default" affinity scope will be 7126 updated accordignly. 7127 7128 workqueue.debug_force_rr_cpu 7129 Workqueue used to implicitly guarantee that work 7130 items queued without explicit CPU specified are put 7131 on the local CPU. This guarantee is no longer true 7132 and while local CPU is still preferred work items 7133 may be put on foreign CPUs. This debug option 7134 forces round-robin CPU selection to flush out 7135 usages which depend on the now broken guarantee. 7136 When enabled, memory and cache locality will be 7137 impacted. 7138 7139 writecombine= [LOONGARCH] Control the MAT (Memory Access Type) of 7140 ioremap_wc(). 7141 7142 on - Enable writecombine, use WUC for ioremap_wc() 7143 off - Disable writecombine, use SUC for ioremap_wc() 7144 7145 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 7146 default x2apic cluster mode on platforms 7147 supporting x2apic. 7148 7149 xen_512gb_limit [KNL,X86-64,XEN] 7150 Restricts the kernel running paravirtualized under Xen 7151 to use only up to 512 GB of RAM. The reason to do so is 7152 crash analysis tools and Xen tools for doing domain 7153 save/restore/migration must be enabled to handle larger 7154 domains. 7155 7156 xen_emul_unplug= [HW,X86,XEN] 7157 Unplug Xen emulated devices 7158 Format: [unplug0,][unplug1] 7159 ide-disks -- unplug primary master IDE devices 7160 aux-ide-disks -- unplug non-primary-master IDE devices 7161 nics -- unplug network devices 7162 all -- unplug all emulated devices (NICs and IDE disks) 7163 unnecessary -- unplugging emulated devices is 7164 unnecessary even if the host did not respond to 7165 the unplug protocol 7166 never -- do not unplug even if version check succeeds 7167 7168 xen_legacy_crash [X86,XEN] 7169 Crash from Xen panic notifier, without executing late 7170 panic() code such as dumping handler. 7171 7172 xen_msr_safe= [X86,XEN] 7173 Format: <bool> 7174 Select whether to always use non-faulting (safe) MSR 7175 access functions when running as Xen PV guest. The 7176 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 7177 7178 xen_nopvspin [X86,XEN] 7179 Disables the qspinlock slowpath using Xen PV optimizations. 7180 This parameter is obsoleted by "nopvspin" parameter, which 7181 has equivalent effect for XEN platform. 7182 7183 xen_nopv [X86] 7184 Disables the PV optimizations forcing the HVM guest to 7185 run as generic HVM guest with no PV drivers. 7186 This option is obsoleted by the "nopv" option, which 7187 has equivalent effect for XEN platform. 7188 7189 xen_no_vector_callback 7190 [KNL,X86,XEN] Disable the vector callback for Xen 7191 event channel interrupts. 7192 7193 xen_scrub_pages= [XEN] 7194 Boolean option to control scrubbing pages before giving them back 7195 to Xen, for use by other domains. Can be also changed at runtime 7196 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 7197 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 7198 7199 xen_timer_slop= [X86-64,XEN] 7200 Set the timer slop (in nanoseconds) for the virtual Xen 7201 timers (default is 100000). This adjusts the minimum 7202 delta of virtualized Xen timers, where lower values 7203 improve timer resolution at the expense of processing 7204 more timer interrupts. 7205 7206 xen.balloon_boot_timeout= [XEN] 7207 The time (in seconds) to wait before giving up to boot 7208 in case initial ballooning fails to free enough memory. 7209 Applies only when running as HVM or PVH guest and 7210 started with less memory configured than allowed at 7211 max. Default is 180. 7212 7213 xen.event_eoi_delay= [XEN] 7214 How long to delay EOI handling in case of event 7215 storms (jiffies). Default is 10. 7216 7217 xen.event_loop_timeout= [XEN] 7218 After which time (jiffies) the event handling loop 7219 should start to delay EOI handling. Default is 2. 7220 7221 xen.fifo_events= [XEN] 7222 Boolean parameter to disable using fifo event handling 7223 even if available. Normally fifo event handling is 7224 preferred over the 2-level event handling, as it is 7225 fairer and the number of possible event channels is 7226 much higher. Default is on (use fifo events). 7227 7228 xirc2ps_cs= [NET,PCMCIA] 7229 Format: 7230 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 7231 7232 xive= [PPC] 7233 By default on POWER9 and above, the kernel will 7234 natively use the XIVE interrupt controller. This option 7235 allows the fallback firmware mode to be used: 7236 7237 off Fallback to firmware control of XIVE interrupt 7238 controller on both pseries and powernv 7239 platforms. Only useful on POWER9 and above. 7240 7241 xive.store-eoi=off [PPC] 7242 By default on POWER10 and above, the kernel will use 7243 stores for EOI handling when the XIVE interrupt mode 7244 is active. This option allows the XIVE driver to use 7245 loads instead, as on POWER9. 7246 7247 xhci-hcd.quirks [USB,KNL] 7248 A hex value specifying bitmask with supplemental xhci 7249 host controller quirks. Meaning of each bit can be 7250 consulted in header drivers/usb/host/xhci.h. 7251 7252 xmon [PPC] 7253 Format: { early | on | rw | ro | off } 7254 Controls if xmon debugger is enabled. Default is off. 7255 Passing only "xmon" is equivalent to "xmon=early". 7256 early Call xmon as early as possible on boot; xmon 7257 debugger is called from setup_arch(). 7258 on xmon debugger hooks will be installed so xmon 7259 is only called on a kernel crash. Default mode, 7260 i.e. either "ro" or "rw" mode, is controlled 7261 with CONFIG_XMON_DEFAULT_RO_MODE. 7262 rw xmon debugger hooks will be installed so xmon 7263 is called only on a kernel crash, mode is write, 7264 meaning SPR registers, memory and, other data 7265 can be written using xmon commands. 7266 ro same as "rw" option above but SPR registers, 7267 memory, and other data can't be written using 7268 xmon commands. 7269 off xmon is disabled. 7270 7271