1 acpi= [HW,ACPI,X86,ARM64] 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] 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, ONLY "acpi=off", "acpi=on" or "acpi=force" 14 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/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. 326 327 amd_iommu_dump= [HW,X86-64] 328 Enable AMD IOMMU driver option to dump the ACPI table 329 for AMD IOMMU. With this option enabled, AMD IOMMU 330 driver will print ACPI tables for AMD IOMMU during 331 IOMMU initialization. 332 333 amd_iommu_intr= [HW,X86-64] 334 Specifies one of the following AMD IOMMU interrupt 335 remapping modes: 336 legacy - Use legacy interrupt remapping mode. 337 vapic - Use virtual APIC mode, which allows IOMMU 338 to inject interrupts directly into guest. 339 This mode requires kvm-amd.avic=1. 340 (Default when IOMMU HW support is present.) 341 342 amd_pstate= [X86] 343 disable 344 Do not enable amd_pstate as the default 345 scaling driver for the supported processors 346 passive 347 Use amd_pstate with passive mode as a scaling driver. 348 In this mode autonomous selection is disabled. 349 Driver requests a desired performance level and platform 350 tries to match the same performance level if it is 351 satisfied by guaranteed performance level. 352 active 353 Use amd_pstate_epp driver instance as the scaling driver, 354 driver provides a hint to the hardware if software wants 355 to bias toward performance (0x0) or energy efficiency (0xff) 356 to the CPPC firmware. then CPPC power algorithm will 357 calculate the runtime workload and adjust the realtime cores 358 frequency. 359 guided 360 Activate guided autonomous mode. Driver requests minimum and 361 maximum performance level and the platform autonomously 362 selects a performance level in this range and appropriate 363 to the current workload. 364 365 amijoy.map= [HW,JOY] Amiga joystick support 366 Map of devices attached to JOY0DAT and JOY1DAT 367 Format: <a>,<b> 368 See also Documentation/input/joydev/joystick.rst 369 370 analog.map= [HW,JOY] Analog joystick and gamepad support 371 Specifies type or capabilities of an analog joystick 372 connected to one of 16 gameports 373 Format: <type1>,<type2>,..<type16> 374 375 apc= [HW,SPARC] 376 Power management functions (SPARCstation-4/5 + deriv.) 377 Format: noidle 378 Disable APC CPU standby support. SPARCstation-Fox does 379 not play well with APC CPU idle - disable it if you have 380 APC and your system crashes randomly. 381 382 apic= [APIC,X86] Advanced Programmable Interrupt Controller 383 Change the output verbosity while booting 384 Format: { quiet (default) | verbose | debug } 385 Change the amount of debugging information output 386 when initialising the APIC and IO-APIC components. 387 For X86-32, this can also be used to specify an APIC 388 driver name. 389 Format: apic=driver_name 390 Examples: apic=bigsmp 391 392 apic_extnmi= [APIC,X86] External NMI delivery setting 393 Format: { bsp (default) | all | none } 394 bsp: External NMI is delivered only to CPU 0 395 all: External NMIs are broadcast to all CPUs as a 396 backup of CPU 0 397 none: External NMI is masked for all CPUs. This is 398 useful so that a dump capture kernel won't be 399 shot down by NMI 400 401 autoconf= [IPV6] 402 See Documentation/networking/ipv6.rst. 403 404 apm= [APM] Advanced Power Management 405 See header of arch/x86/kernel/apm_32.c. 406 407 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 408 Format: { "0" | "1" } 409 See security/apparmor/Kconfig help text 410 0 -- disable. 411 1 -- enable. 412 Default value is set via kernel config option. 413 414 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 415 Format: <io>,<irq>,<nodeID> 416 417 arm64.nobti [ARM64] Unconditionally disable Branch Target 418 Identification support 419 420 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 421 support 422 423 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 424 support 425 426 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 427 Extension support 428 429 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 430 Extension support 431 432 ataflop= [HW,M68k] 433 434 atarimouse= [HW,MOUSE] Atari Mouse 435 436 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 437 EzKey and similar keyboards 438 439 atkbd.reset= [HW] Reset keyboard during initialization 440 441 atkbd.set= [HW] Select keyboard code set 442 Format: <int> (2 = AT (default), 3 = PS/2) 443 444 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 445 keyboards 446 447 atkbd.softraw= [HW] Choose between synthetic and real raw mode 448 Format: <bool> (0 = real, 1 = synthetic (default)) 449 450 atkbd.softrepeat= [HW] 451 Use software keyboard repeat 452 453 audit= [KNL] Enable the audit sub-system 454 Format: { "0" | "1" | "off" | "on" } 455 0 | off - kernel audit is disabled and can not be 456 enabled until the next reboot 457 unset - kernel audit is initialized but disabled and 458 will be fully enabled by the userspace auditd. 459 1 | on - kernel audit is initialized and partially 460 enabled, storing at most audit_backlog_limit 461 messages in RAM until it is fully enabled by the 462 userspace auditd. 463 Default: unset 464 465 audit_backlog_limit= [KNL] Set the audit queue size limit. 466 Format: <int> (must be >=0) 467 Default: 64 468 469 bau= [X86_UV] Enable the BAU on SGI UV. The default 470 behavior is to disable the BAU (i.e. bau=0). 471 Format: { "0" | "1" } 472 0 - Disable the BAU. 473 1 - Enable the BAU. 474 unset - Disable the BAU. 475 476 baycom_epp= [HW,AX25] 477 Format: <io>,<mode> 478 479 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 480 Format: <io>,<mode> 481 See header of drivers/net/hamradio/baycom_par.c. 482 483 baycom_ser_fdx= [HW,AX25] 484 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 485 Format: <io>,<irq>,<mode>[,<baud>] 486 See header of drivers/net/hamradio/baycom_ser_fdx.c. 487 488 baycom_ser_hdx= [HW,AX25] 489 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 490 Format: <io>,<irq>,<mode> 491 See header of drivers/net/hamradio/baycom_ser_hdx.c. 492 493 bert_disable [ACPI] 494 Disable BERT OS support on buggy BIOSes. 495 496 bgrt_disable [ACPI][X86] 497 Disable BGRT to avoid flickering OEM logo. 498 499 blkdevparts= Manual partition parsing of block device(s) for 500 embedded devices based on command line input. 501 See Documentation/block/cmdline-partition.rst 502 503 boot_delay= Milliseconds to delay each printk during boot. 504 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 505 and you may also have to specify "lpj=". Boot_delay 506 values larger than 10 seconds (10000) are assumed 507 erroneous and ignored. 508 Format: integer 509 510 bootconfig [KNL] 511 Extended command line options can be added to an initrd 512 and this will cause the kernel to look for it. 513 514 See Documentation/admin-guide/bootconfig.rst 515 516 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 517 bttv.radio= Most important insmod options are available as 518 kernel args too. 519 bttv.pll= See Documentation/admin-guide/media/bttv.rst 520 bttv.tuner= 521 522 bulk_remove=off [PPC] This parameter disables the use of the pSeries 523 firmware feature for flushing multiple hpte entries 524 at a time. 525 526 c101= [NET] Moxa C101 synchronous serial card 527 528 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 529 Sometimes CPU hardware bugs make them report the cache 530 size incorrectly. The kernel will attempt work arounds 531 to fix known problems, but for some CPUs it is not 532 possible to determine what the correct size should be. 533 This option provides an override for these situations. 534 535 carrier_timeout= 536 [NET] Specifies amount of time (in seconds) that 537 the kernel should wait for a network carrier. By default 538 it waits 120 seconds. 539 540 ca_keys= [KEYS] This parameter identifies a specific key(s) on 541 the system trusted keyring to be used for certificate 542 trust validation. 543 format: { id:<keyid> | builtin } 544 545 cca= [MIPS] Override the kernel pages' cache coherency 546 algorithm. Accepted values range from 0 to 7 547 inclusive. See arch/mips/include/asm/pgtable-bits.h 548 for platform specific values (SB1, Loongson3 and 549 others). 550 551 ccw_timeout_log [S390] 552 See Documentation/s390/common_io.rst for details. 553 554 cgroup_disable= [KNL] Disable a particular controller or optional feature 555 Format: {name of the controller(s) or feature(s) to disable} 556 The effects of cgroup_disable=foo are: 557 - foo isn't auto-mounted if you mount all cgroups in 558 a single hierarchy 559 - foo isn't visible as an individually mountable 560 subsystem 561 - if foo is an optional feature then the feature is 562 disabled and corresponding cgroup files are not 563 created 564 {Currently only "memory" controller deal with this and 565 cut the overhead, others just disable the usage. So 566 only cgroup_disable=memory is actually worthy} 567 Specifying "pressure" disables per-cgroup pressure 568 stall information accounting feature 569 570 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 571 Format: { { controller | "all" | "named" } 572 [,{ controller | "all" | "named" }...] } 573 Like cgroup_disable, but only applies to cgroup v1; 574 the blacklisted controllers remain available in cgroup2. 575 "all" blacklists all controllers and "named" disables 576 named mounts. Specifying both "all" and "named" disables 577 all v1 hierarchies. 578 579 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 580 Format: <string> 581 nosocket -- Disable socket memory accounting. 582 nokmem -- Disable kernel memory accounting. 583 nobpf -- Disable BPF memory accounting. 584 585 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 586 Format: { "0" | "1" } 587 See security/selinux/Kconfig help text. 588 0 -- check protection applied by kernel (includes 589 any implied execute protection). 590 1 -- check protection requested by application. 591 Default value is set via a kernel config option. 592 Value can be changed at runtime via 593 /sys/fs/selinux/checkreqprot. 594 Setting checkreqprot to 1 is deprecated. 595 596 cio_ignore= [S390] 597 See Documentation/s390/common_io.rst for details. 598 599 clearcpuid=X[,X...] [X86] 600 Disable CPUID feature X for the kernel. See 601 arch/x86/include/asm/cpufeatures.h for the valid bit 602 numbers X. Note the Linux-specific bits are not necessarily 603 stable over kernel options, but the vendor-specific 604 ones should be. 605 X can also be a string as appearing in the flags: line 606 in /proc/cpuinfo which does not have the above 607 instability issue. However, not all features have names 608 in /proc/cpuinfo. 609 Note that using this option will taint your kernel. 610 Also note that user programs calling CPUID directly 611 or using the feature without checking anything 612 will still see it. This just prevents it from 613 being used by the kernel or shown in /proc/cpuinfo. 614 Also note the kernel might malfunction if you disable 615 some critical bits. 616 617 clk_ignore_unused 618 [CLK] 619 Prevents the clock framework from automatically gating 620 clocks that have not been explicitly enabled by a Linux 621 device driver but are enabled in hardware at reset or 622 by the bootloader/firmware. Note that this does not 623 force such clocks to be always-on nor does it reserve 624 those clocks in any way. This parameter is useful for 625 debug and development, but should not be needed on a 626 platform with proper driver support. For more 627 information, see Documentation/driver-api/clk.rst. 628 629 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 630 [Deprecated] 631 Forces specified clocksource (if available) to be used 632 when calculating gettimeofday(). If specified 633 clocksource is not available, it defaults to PIT. 634 Format: { pit | tsc | cyclone | pmtmr } 635 636 clocksource= Override the default clocksource 637 Format: <string> 638 Override the default clocksource and use the clocksource 639 with the name specified. 640 Some clocksource names to choose from, depending on 641 the platform: 642 [all] jiffies (this is the base, fallback clocksource) 643 [ACPI] acpi_pm 644 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 645 pxa_timer,timer3,32k_counter,timer0_1 646 [X86-32] pit,hpet,tsc; 647 scx200_hrt on Geode; cyclone on IBM x440 648 [MIPS] MIPS 649 [PARISC] cr16 650 [S390] tod 651 [SH] SuperH 652 [SPARC64] tick 653 [X86-64] hpet,tsc 654 655 clocksource.arm_arch_timer.evtstrm= 656 [ARM,ARM64] 657 Format: <bool> 658 Enable/disable the eventstream feature of the ARM 659 architected timer so that code using WFE-based polling 660 loops can be debugged more effectively on production 661 systems. 662 663 clocksource.max_cswd_read_retries= [KNL] 664 Number of clocksource_watchdog() retries due to 665 external delays before the clock will be marked 666 unstable. Defaults to two retries, that is, 667 three attempts to read the clock under test. 668 669 clocksource.verify_n_cpus= [KNL] 670 Limit the number of CPUs checked for clocksources 671 marked with CLOCK_SOURCE_VERIFY_PERCPU that 672 are marked unstable due to excessive skew. 673 A negative value says to check all CPUs, while 674 zero says not to check any. Values larger than 675 nr_cpu_ids are silently truncated to nr_cpu_ids. 676 The actual CPUs are chosen randomly, with 677 no replacement if the same CPU is chosen twice. 678 679 clocksource-wdtest.holdoff= [KNL] 680 Set the time in seconds that the clocksource 681 watchdog test waits before commencing its tests. 682 Defaults to zero when built as a module and to 683 10 seconds when built into the kernel. 684 685 cma=nn[MG]@[start[MG][-end[MG]]] 686 [KNL,CMA] 687 Sets the size of kernel global memory area for 688 contiguous memory allocations and optionally the 689 placement constraint by the physical address range of 690 memory allocations. A value of 0 disables CMA 691 altogether. For more information, see 692 kernel/dma/contiguous.c 693 694 cma_pernuma=nn[MG] 695 [ARM64,KNL,CMA] 696 Sets the size of kernel per-numa memory area for 697 contiguous memory allocations. A value of 0 disables 698 per-numa CMA altogether. And If this option is not 699 specified, the default value is 0. 700 With per-numa CMA enabled, DMA users on node nid will 701 first try to allocate buffer from the pernuma area 702 which is located in node nid, if the allocation fails, 703 they will fallback to the global default memory area. 704 705 cmo_free_hint= [PPC] Format: { yes | no } 706 Specify whether pages are marked as being inactive 707 when they are freed. This is used in CMO environments 708 to determine OS memory pressure for page stealing by 709 a hypervisor. 710 Default: yes 711 712 coherent_pool=nn[KMG] [ARM,KNL] 713 Sets the size of memory pool for coherent, atomic dma 714 allocations, by default set to 256K. 715 716 com20020= [HW,NET] ARCnet - COM20020 chipset 717 Format: 718 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 719 720 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 721 Format: <io>[,<irq>] 722 723 com90xx= [HW,NET] 724 ARCnet - COM90xx chipset (memory-mapped buffers) 725 Format: <io>[,<irq>[,<memstart>]] 726 727 condev= [HW,S390] console device 728 conmode= 729 730 con3215_drop= [S390] 3215 console drop mode. 731 Format: y|n|Y|N|1|0 732 When set to true, drop data on the 3215 console when 733 the console buffer is full. In this case the 734 operator using a 3270 terminal emulator (for example 735 x3270) does not have to enter the clear key for the 736 console output to advance and the kernel to continue. 737 This leads to a much faster boot time when a 3270 738 terminal emulator is active. If no 3270 terminal 739 emulator is used, this parameter has no effect. 740 741 console= [KNL] Output console device and options. 742 743 tty<n> Use the virtual console device <n>. 744 745 ttyS<n>[,options] 746 ttyUSB0[,options] 747 Use the specified serial port. The options are of 748 the form "bbbbpnf", where "bbbb" is the baud rate, 749 "p" is parity ("n", "o", or "e"), "n" is number of 750 bits, and "f" is flow control ("r" for RTS or 751 omit it). Default is "9600n8". 752 753 See Documentation/admin-guide/serial-console.rst for more 754 information. See 755 Documentation/networking/netconsole.rst for an 756 alternative. 757 758 uart[8250],io,<addr>[,options] 759 uart[8250],mmio,<addr>[,options] 760 uart[8250],mmio16,<addr>[,options] 761 uart[8250],mmio32,<addr>[,options] 762 uart[8250],0x<addr>[,options] 763 Start an early, polled-mode console on the 8250/16550 764 UART at the specified I/O port or MMIO address, 765 switching to the matching ttyS device later. 766 MMIO inter-register address stride is either 8-bit 767 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 768 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 769 to be equivalent to 'mmio'. 'options' are specified in 770 the same format described for ttyS above; if unspecified, 771 the h/w is not re-initialized. 772 773 hvc<n> Use the hypervisor console device <n>. This is for 774 both Xen and PowerPC hypervisors. 775 776 { null | "" } 777 Use to disable console output, i.e., to have kernel 778 console messages discarded. 779 This must be the only console= parameter used on the 780 kernel command line. 781 782 If the device connected to the port is not a TTY but a braille 783 device, prepend "brl," before the device type, for instance 784 console=brl,ttyS0 785 For now, only VisioBraille is supported. 786 787 console_msg_format= 788 [KNL] Change console messages format 789 default 790 By default we print messages on consoles in 791 "[time stamp] text\n" format (time stamp may not be 792 printed, depending on CONFIG_PRINTK_TIME or 793 `printk_time' param). 794 syslog 795 Switch to syslog format: "<%u>[time stamp] text\n" 796 IOW, each message will have a facility and loglevel 797 prefix. The format is similar to one used by syslog() 798 syscall, or to executing "dmesg -S --raw" or to reading 799 from /proc/kmsg. 800 801 consoleblank= [KNL] The console blank (screen saver) timeout in 802 seconds. A value of 0 disables the blank timer. 803 Defaults to 0. 804 805 coredump_filter= 806 [KNL] Change the default value for 807 /proc/<pid>/coredump_filter. 808 See also Documentation/filesystems/proc.rst. 809 810 coresight_cpu_debug.enable 811 [ARM,ARM64] 812 Format: <bool> 813 Enable/disable the CPU sampling based debugging. 814 0: default value, disable debugging 815 1: enable debugging at boot time 816 817 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 818 Format: 819 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 820 821 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 822 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 823 Some features depend on CPU0. Known dependencies are: 824 1. Resume from suspend/hibernate depends on CPU0. 825 Suspend/hibernate will fail if CPU0 is offline and you 826 need to online CPU0 before suspend/hibernate. 827 2. PIC interrupts also depend on CPU0. CPU0 can't be 828 removed if a PIC interrupt is detected. 829 It's said poweroff/reboot may depend on CPU0 on some 830 machines although I haven't seen such issues so far 831 after CPU0 is offline on a few tested machines. 832 If the dependencies are under your control, you can 833 turn on cpu0_hotplug. 834 835 cpuidle.off=1 [CPU_IDLE] 836 disable the cpuidle sub-system 837 838 cpuidle.governor= 839 [CPU_IDLE] Name of the cpuidle governor to use. 840 841 cpufreq.off=1 [CPU_FREQ] 842 disable the cpufreq sub-system 843 844 cpufreq.default_governor= 845 [CPU_FREQ] Name of the default cpufreq governor or 846 policy to use. This governor must be registered in the 847 kernel before the cpufreq driver probes. 848 849 cpu_init_udelay=N 850 [X86] Delay for N microsec between assert and de-assert 851 of APIC INIT to start processors. This delay occurs 852 on every CPU online, such as boot, and resume from suspend. 853 Default: 10000 854 855 crash_kexec_post_notifiers 856 Run kdump after running panic-notifiers and dumping 857 kmsg. This only for the users who doubt kdump always 858 succeeds in any situation. 859 Note that this also increases risks of kdump failure, 860 because some panic notifiers can make the crashed 861 kernel more unstable. 862 863 crashkernel=size[KMG][@offset[KMG]] 864 [KNL] Using kexec, Linux can switch to a 'crash kernel' 865 upon panic. This parameter reserves the physical 866 memory region [offset, offset + size] for that kernel 867 image. If '@offset' is omitted, then a suitable offset 868 is selected automatically. 869 [KNL, X86-64, ARM64] Select a region under 4G first, and 870 fall back to reserve region above 4G when '@offset' 871 hasn't been specified. 872 See Documentation/admin-guide/kdump/kdump.rst for further details. 873 874 crashkernel=range1:size1[,range2:size2,...][@offset] 875 [KNL] Same as above, but depends on the memory 876 in the running system. The syntax of range is 877 start-[end] where start and end are both 878 a memory unit (amount[KMG]). See also 879 Documentation/admin-guide/kdump/kdump.rst for an example. 880 881 crashkernel=size[KMG],high 882 [KNL, X86-64, ARM64] range could be above 4G. Allow kernel 883 to allocate physical memory region from top, so could 884 be above 4G if system have more than 4G ram installed. 885 Otherwise memory region will be allocated below 4G, if 886 available. 887 It will be ignored if crashkernel=X is specified. 888 crashkernel=size[KMG],low 889 [KNL, X86-64, ARM64] range under 4G. When crashkernel=X,high 890 is passed, kernel could allocate physical memory region 891 above 4G, that cause second kernel crash on system 892 that require some amount of low memory, e.g. swiotlb 893 requires at least 64M+32K low memory, also enough extra 894 low memory is needed to make sure DMA buffers for 32-bit 895 devices won't run out. Kernel would try to allocate 896 default size of memory below 4G automatically. The default 897 size is platform dependent. 898 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 899 --> arm64: 128MiB 900 This one lets the user specify own low range under 4G 901 for second kernel instead. 902 0: to disable low allocation. 903 It will be ignored when crashkernel=X,high is not used 904 or memory reserved is below 4G. 905 906 cryptomgr.notests 907 [KNL] Disable crypto self-tests 908 909 cs89x0_dma= [HW,NET] 910 Format: <dma> 911 912 cs89x0_media= [HW,NET] 913 Format: { rj45 | aui | bnc } 914 915 csdlock_debug= [KNL] Enable debug add-ons of cross-CPU function call 916 handling. When switched on, additional debug data is 917 printed to the console in case a hanging CPU is 918 detected, and that CPU is pinged again in order to try 919 to resolve the hang situation. 920 0: disable csdlock debugging (default) 921 1: enable basic csdlock debugging (minor impact) 922 ext: enable extended csdlock debugging (more impact, 923 but more data) 924 925 dasd= [HW,NET] 926 See header of drivers/s390/block/dasd_devmap.c. 927 928 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 929 (one device per port) 930 Format: <port#>,<type> 931 See also Documentation/input/devices/joystick-parport.rst 932 933 debug [KNL] Enable kernel debugging (events log level). 934 935 debug_boot_weak_hash 936 [KNL] Enable printing [hashed] pointers early in the 937 boot sequence. If enabled, we use a weak hash instead 938 of siphash to hash pointers. Use this option if you are 939 seeing instances of '(___ptrval___)') and need to see a 940 value (hashed pointer) instead. Cryptographically 941 insecure, please do not use on production kernels. 942 943 debug_locks_verbose= 944 [KNL] verbose locking self-tests 945 Format: <int> 946 Print debugging info while doing the locking API 947 self-tests. 948 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 949 (no extra messages), setting it to -1 (all bits set) 950 will print _a_lot_ more information - normally only 951 useful to lockdep developers. 952 953 debug_objects [KNL] Enable object debugging 954 955 debug_guardpage_minorder= 956 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 957 parameter allows control of the order of pages that will 958 be intentionally kept free (and hence protected) by the 959 buddy allocator. Bigger value increase the probability 960 of catching random memory corruption, but reduce the 961 amount of memory for normal system use. The maximum 962 possible value is MAX_ORDER/2. Setting this parameter 963 to 1 or 2 should be enough to identify most random 964 memory corruption problems caused by bugs in kernel or 965 driver code when a CPU writes to (or reads from) a 966 random memory location. Note that there exists a class 967 of memory corruptions problems caused by buggy H/W or 968 F/W or by drivers badly programming DMA (basically when 969 memory is written at bus level and the CPU MMU is 970 bypassed) which are not detectable by 971 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 972 tracking down these problems. 973 974 debug_pagealloc= 975 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 976 enables the feature at boot time. By default, it is 977 disabled and the system will work mostly the same as a 978 kernel built without CONFIG_DEBUG_PAGEALLOC. 979 Note: to get most of debug_pagealloc error reports, it's 980 useful to also enable the page_owner functionality. 981 on: enable the feature 982 983 debugfs= [KNL] This parameter enables what is exposed to userspace 984 and debugfs internal clients. 985 Format: { on, no-mount, off } 986 on: All functions are enabled. 987 no-mount: 988 Filesystem is not registered but kernel clients can 989 access APIs and a crashkernel can be used to read 990 its content. There is nothing to mount. 991 off: Filesystem is not registered and clients 992 get a -EPERM as result when trying to register files 993 or directories within debugfs. 994 This is equivalent of the runtime functionality if 995 debugfs was not enabled in the kernel at all. 996 Default value is set in build-time with a kernel configuration. 997 998 debugpat [X86] Enable PAT debugging 999 1000 default_hugepagesz= 1001 [HW] The size of the default HugeTLB page. This is 1002 the size represented by the legacy /proc/ hugepages 1003 APIs. In addition, this is the default hugetlb size 1004 used for shmget(), mmap() and mounting hugetlbfs 1005 filesystems. If not specified, defaults to the 1006 architecture's default huge page size. Huge page 1007 sizes are architecture dependent. See also 1008 Documentation/admin-guide/mm/hugetlbpage.rst. 1009 Format: size[KMG] 1010 1011 deferred_probe_timeout= 1012 [KNL] Debugging option to set a timeout in seconds for 1013 deferred probe to give up waiting on dependencies to 1014 probe. Only specific dependencies (subsystems or 1015 drivers) that have opted in will be ignored. A timeout 1016 of 0 will timeout at the end of initcalls. If the time 1017 out hasn't expired, it'll be restarted by each 1018 successful driver registration. This option will also 1019 dump out devices still on the deferred probe list after 1020 retrying. 1021 1022 delayacct [KNL] Enable per-task delay accounting 1023 1024 dell_smm_hwmon.ignore_dmi= 1025 [HW] Continue probing hardware even if DMI data 1026 indicates that the driver is running on unsupported 1027 hardware. 1028 1029 dell_smm_hwmon.force= 1030 [HW] Activate driver even if SMM BIOS signature does 1031 not match list of supported models and enable otherwise 1032 blacklisted features. 1033 1034 dell_smm_hwmon.power_status= 1035 [HW] Report power status in /proc/i8k 1036 (disabled by default). 1037 1038 dell_smm_hwmon.restricted= 1039 [HW] Allow controlling fans only if SYS_ADMIN 1040 capability is set. 1041 1042 dell_smm_hwmon.fan_mult= 1043 [HW] Factor to multiply fan speed with. 1044 1045 dell_smm_hwmon.fan_max= 1046 [HW] Maximum configurable fan speed. 1047 1048 dfltcc= [HW,S390] 1049 Format: { on | off | def_only | inf_only | always } 1050 on: s390 zlib hardware support for compression on 1051 level 1 and decompression (default) 1052 off: No s390 zlib hardware support 1053 def_only: s390 zlib hardware support for deflate 1054 only (compression on level 1) 1055 inf_only: s390 zlib hardware support for inflate 1056 only (decompression) 1057 always: Same as 'on' but ignores the selected compression 1058 level always using hardware support (used for debugging) 1059 1060 dhash_entries= [KNL] 1061 Set number of hash buckets for dentry cache. 1062 1063 disable_1tb_segments [PPC] 1064 Disables the use of 1TB hash page table segments. This 1065 causes the kernel to fall back to 256MB segments which 1066 can be useful when debugging issues that require an SLB 1067 miss to occur. 1068 1069 disable= [IPV6] 1070 See Documentation/networking/ipv6.rst. 1071 1072 disable_radix [PPC] 1073 Disable RADIX MMU mode on POWER9 1074 1075 disable_tlbie [PPC] 1076 Disable TLBIE instruction. Currently does not work 1077 with KVM, with HASH MMU, or with coherent accelerators. 1078 1079 disable_cpu_apicid= [X86,APIC,SMP] 1080 Format: <int> 1081 The number of initial APIC ID for the 1082 corresponding CPU to be disabled at boot, 1083 mostly used for the kdump 2nd kernel to 1084 disable BSP to wake up multiple CPUs without 1085 causing system reset or hang due to sending 1086 INIT from AP to BSP. 1087 1088 disable_ddw [PPC/PSERIES] 1089 Disable Dynamic DMA Window support. Use this 1090 to workaround buggy firmware. 1091 1092 disable_ipv6= [IPV6] 1093 See Documentation/networking/ipv6.rst. 1094 1095 disable_mtrr_cleanup [X86] 1096 The kernel tries to adjust MTRR layout from continuous 1097 to discrete, to make X server driver able to add WB 1098 entry later. This parameter disables that. 1099 1100 disable_mtrr_trim [X86, Intel and AMD only] 1101 By default the kernel will trim any uncacheable 1102 memory out of your available memory pool based on 1103 MTRR settings. This parameter disables that behavior, 1104 possibly causing your machine to run very slowly. 1105 1106 disable_timer_pin_1 [X86] 1107 Disable PIN 1 of APIC timer 1108 Can be useful to work around chipset bugs. 1109 1110 dis_ucode_ldr [X86] Disable the microcode loader. 1111 1112 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1113 this option disables the debugging code at boot. 1114 1115 dma_debug_entries=<number> 1116 This option allows to tune the number of preallocated 1117 entries for DMA-API debugging code. One entry is 1118 required per DMA-API allocation. Use this if the 1119 DMA-API debugging code disables itself because the 1120 architectural default is too low. 1121 1122 dma_debug_driver=<driver_name> 1123 With this option the DMA-API debugging driver 1124 filter feature can be enabled at boot time. Just 1125 pass the driver to filter for as the parameter. 1126 The filter can be disabled or changed to another 1127 driver later using sysfs. 1128 1129 driver_async_probe= [KNL] 1130 List of driver names to be probed asynchronously. * 1131 matches with all driver names. If * is specified, the 1132 rest of the listed driver names are those that will NOT 1133 match the *. 1134 Format: <driver_name1>,<driver_name2>... 1135 1136 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1137 Broken monitors, graphic adapters, KVMs and EDIDless 1138 panels may send no or incorrect EDID data sets. 1139 This parameter allows to specify an EDID data sets 1140 in the /lib/firmware directory that are used instead. 1141 Generic built-in EDID data sets are used, if one of 1142 edid/1024x768.bin, edid/1280x1024.bin, 1143 edid/1680x1050.bin, or edid/1920x1080.bin is given 1144 and no file with the same name exists. Details and 1145 instructions how to build your own EDID data are 1146 available in Documentation/admin-guide/edid.rst. An EDID 1147 data set will only be used for a particular connector, 1148 if its name and a colon are prepended to the EDID 1149 name. Each connector may use a unique EDID data 1150 set by separating the files with a comma. An EDID 1151 data set with no connector name will be used for 1152 any connectors not explicitly specified. 1153 1154 dscc4.setup= [NET] 1155 1156 dt_cpu_ftrs= [PPC] 1157 Format: {"off" | "known"} 1158 Control how the dt_cpu_ftrs device-tree binding is 1159 used for CPU feature discovery and setup (if it 1160 exists). 1161 off: Do not use it, fall back to legacy cpu table. 1162 known: Do not pass through unknown features to guests 1163 or userspace, only those that the kernel is aware of. 1164 1165 dump_apple_properties [X86] 1166 Dump name and content of EFI device properties on 1167 x86 Macs. Useful for driver authors to determine 1168 what data is available or for reverse-engineering. 1169 1170 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1171 <module>.dyndbg[="val"] 1172 Enable debug messages at boot time. See 1173 Documentation/admin-guide/dynamic-debug-howto.rst 1174 for details. 1175 1176 early_ioremap_debug [KNL] 1177 Enable debug messages in early_ioremap support. This 1178 is useful for tracking down temporary early mappings 1179 which are not unmapped. 1180 1181 earlycon= [KNL] Output early console device and options. 1182 1183 When used with no options, the early console is 1184 determined by stdout-path property in device tree's 1185 chosen node or the ACPI SPCR table if supported by 1186 the platform. 1187 1188 cdns,<addr>[,options] 1189 Start an early, polled-mode console on a Cadence 1190 (xuartps) serial port at the specified address. Only 1191 supported option is baud rate. If baud rate is not 1192 specified, the serial port must already be setup and 1193 configured. 1194 1195 uart[8250],io,<addr>[,options[,uartclk]] 1196 uart[8250],mmio,<addr>[,options[,uartclk]] 1197 uart[8250],mmio32,<addr>[,options[,uartclk]] 1198 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1199 uart[8250],0x<addr>[,options] 1200 Start an early, polled-mode console on the 8250/16550 1201 UART at the specified I/O port or MMIO address. 1202 MMIO inter-register address stride is either 8-bit 1203 (mmio) or 32-bit (mmio32 or mmio32be). 1204 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1205 to be equivalent to 'mmio'. 'options' are specified 1206 in the same format described for "console=ttyS<n>"; if 1207 unspecified, the h/w is not initialized. 'uartclk' is 1208 the uart clock frequency; if unspecified, it is set 1209 to 'BASE_BAUD' * 16. 1210 1211 pl011,<addr> 1212 pl011,mmio32,<addr> 1213 Start an early, polled-mode console on a pl011 serial 1214 port at the specified address. The pl011 serial port 1215 must already be setup and configured. Options are not 1216 yet supported. If 'mmio32' is specified, then only 1217 the driver will use only 32-bit accessors to read/write 1218 the device registers. 1219 1220 liteuart,<addr> 1221 Start an early console on a litex serial port at the 1222 specified address. The serial port must already be 1223 setup and configured. Options are not yet supported. 1224 1225 meson,<addr> 1226 Start an early, polled-mode console on a meson serial 1227 port at the specified address. The serial port must 1228 already be setup and configured. Options are not yet 1229 supported. 1230 1231 msm_serial,<addr> 1232 Start an early, polled-mode console on an msm serial 1233 port at the specified address. The serial port 1234 must already be setup and configured. Options are not 1235 yet supported. 1236 1237 msm_serial_dm,<addr> 1238 Start an early, polled-mode console on an msm serial 1239 dm port at the specified address. The serial port 1240 must already be setup and configured. Options are not 1241 yet supported. 1242 1243 owl,<addr> 1244 Start an early, polled-mode console on a serial port 1245 of an Actions Semi SoC, such as S500 or S900, at the 1246 specified address. The serial port must already be 1247 setup and configured. Options are not yet supported. 1248 1249 rda,<addr> 1250 Start an early, polled-mode console on a serial port 1251 of an RDA Micro SoC, such as RDA8810PL, at the 1252 specified address. The serial port must already be 1253 setup and configured. Options are not yet supported. 1254 1255 sbi 1256 Use RISC-V SBI (Supervisor Binary Interface) for early 1257 console. 1258 1259 smh Use ARM semihosting calls for early console. 1260 1261 s3c2410,<addr> 1262 s3c2412,<addr> 1263 s3c2440,<addr> 1264 s3c6400,<addr> 1265 s5pv210,<addr> 1266 exynos4210,<addr> 1267 Use early console provided by serial driver available 1268 on Samsung SoCs, requires selecting proper type and 1269 a correct base address of the selected UART port. The 1270 serial port must already be setup and configured. 1271 Options are not yet supported. 1272 1273 lantiq,<addr> 1274 Start an early, polled-mode console on a lantiq serial 1275 (lqasc) port at the specified address. The serial port 1276 must already be setup and configured. Options are not 1277 yet supported. 1278 1279 lpuart,<addr> 1280 lpuart32,<addr> 1281 Use early console provided by Freescale LP UART driver 1282 found on Freescale Vybrid and QorIQ LS1021A processors. 1283 A valid base address must be provided, and the serial 1284 port must already be setup and configured. 1285 1286 ec_imx21,<addr> 1287 ec_imx6q,<addr> 1288 Start an early, polled-mode, output-only console on the 1289 Freescale i.MX UART at the specified address. The UART 1290 must already be setup and configured. 1291 1292 ar3700_uart,<addr> 1293 Start an early, polled-mode console on the 1294 Armada 3700 serial port at the specified 1295 address. The serial port must already be setup 1296 and configured. Options are not yet supported. 1297 1298 qcom_geni,<addr> 1299 Start an early, polled-mode console on a Qualcomm 1300 Generic Interface (GENI) based serial port at the 1301 specified address. The serial port must already be 1302 setup and configured. Options are not yet supported. 1303 1304 efifb,[options] 1305 Start an early, unaccelerated console on the EFI 1306 memory mapped framebuffer (if available). On cache 1307 coherent non-x86 systems that use system memory for 1308 the framebuffer, pass the 'ram' option so that it is 1309 mapped with the correct attributes. 1310 1311 linflex,<addr> 1312 Use early console provided by Freescale LINFlexD UART 1313 serial driver for NXP S32V234 SoCs. A valid base 1314 address must be provided, and the serial port must 1315 already be setup and configured. 1316 1317 earlyprintk= [X86,SH,ARM,M68k,S390] 1318 earlyprintk=vga 1319 earlyprintk=sclp 1320 earlyprintk=xen 1321 earlyprintk=serial[,ttySn[,baudrate]] 1322 earlyprintk=serial[,0x...[,baudrate]] 1323 earlyprintk=ttySn[,baudrate] 1324 earlyprintk=dbgp[debugController#] 1325 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1326 earlyprintk=xdbc[xhciController#] 1327 1328 earlyprintk is useful when the kernel crashes before 1329 the normal console is initialized. It is not enabled by 1330 default because it has some cosmetic problems. 1331 1332 Append ",keep" to not disable it when the real console 1333 takes over. 1334 1335 Only one of vga, serial, or usb debug port can 1336 be used at a time. 1337 1338 Currently only ttyS0 and ttyS1 may be specified by 1339 name. Other I/O ports may be explicitly specified 1340 on some architectures (x86 and arm at least) by 1341 replacing ttySn with an I/O port address, like this: 1342 earlyprintk=serial,0x1008,115200 1343 You can find the port for a given device in 1344 /proc/tty/driver/serial: 1345 2: uart:ST16650V2 port:00001008 irq:18 ... 1346 1347 Interaction with the standard serial driver is not 1348 very good. 1349 1350 The VGA output is eventually overwritten by 1351 the real console. 1352 1353 The xen option can only be used in Xen domains. 1354 1355 The sclp output can only be used on s390. 1356 1357 The optional "force" to "pciserial" enables use of a 1358 PCI device even when its classcode is not of the 1359 UART class. 1360 1361 edac_report= [HW,EDAC] Control how to report EDAC event 1362 Format: {"on" | "off" | "force"} 1363 on: enable EDAC to report H/W event. May be overridden 1364 by other higher priority error reporting module. 1365 off: disable H/W event reporting through EDAC. 1366 force: enforce the use of EDAC to report H/W event. 1367 default: on. 1368 1369 edd= [EDD] 1370 Format: {"off" | "on" | "skip[mbr]"} 1371 1372 efi= [EFI] 1373 Format: { "debug", "disable_early_pci_dma", 1374 "nochunk", "noruntime", "nosoftreserve", 1375 "novamap", "no_disable_early_pci_dma" } 1376 debug: enable misc debug output. 1377 disable_early_pci_dma: disable the busmaster bit on all 1378 PCI bridges while in the EFI boot stub. 1379 nochunk: disable reading files in "chunks" in the EFI 1380 boot stub, as chunking can cause problems with some 1381 firmware implementations. 1382 noruntime : disable EFI runtime services support 1383 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1384 attribute may cause the kernel to reserve the 1385 memory range for a memory mapping driver to 1386 claim. Specify efi=nosoftreserve to disable this 1387 reservation and treat the memory by its base type 1388 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1389 novamap: do not call SetVirtualAddressMap(). 1390 no_disable_early_pci_dma: Leave the busmaster bit set 1391 on all PCI bridges while in the EFI boot stub 1392 1393 efi_no_storage_paranoia [EFI; X86] 1394 Using this parameter you can use more than 50% of 1395 your efi variable storage. Use this parameter only if 1396 you are really sure that your UEFI does sane gc and 1397 fulfills the spec otherwise your board may brick. 1398 1399 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1400 Add arbitrary attribute to specific memory range by 1401 updating original EFI memory map. 1402 Region of memory which aa attribute is added to is 1403 from ss to ss+nn. 1404 1405 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1406 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1407 attribute is added to range 0x100000000-0x180000000 and 1408 0x10a0000000-0x1120000000. 1409 1410 If efi_fake_mem=8G@9G:0x40000 is specified, the 1411 EFI_MEMORY_SP(0x40000) attribute is added to 1412 range 0x240000000-0x43fffffff. 1413 1414 Using this parameter you can do debugging of EFI memmap 1415 related features. For example, you can do debugging of 1416 Address Range Mirroring feature even if your box 1417 doesn't support it, or mark specific memory as 1418 "soft reserved". 1419 1420 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1421 that is to be dynamically loaded by Linux. If there are 1422 multiple variables with the same name but with different 1423 vendor GUIDs, all of them will be loaded. See 1424 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1425 1426 1427 eisa_irq_edge= [PARISC,HW] 1428 See header of drivers/parisc/eisa.c. 1429 1430 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1431 Format: ekgdboc=kbd 1432 1433 This is designed to be used in conjunction with 1434 the boot argument: earlyprintk=vga 1435 1436 This parameter works in place of the kgdboc parameter 1437 but can only be used if the backing tty is available 1438 very early in the boot process. For early debugging 1439 via a serial port see kgdboc_earlycon instead. 1440 1441 elanfreq= [X86-32] 1442 See comment before function elanfreq_setup() in 1443 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1444 1445 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1446 Specifies physical address of start of kernel core 1447 image elf header and optionally the size. Generally 1448 kexec loader will pass this option to capture kernel. 1449 See Documentation/admin-guide/kdump/kdump.rst for details. 1450 1451 enable_mtrr_cleanup [X86] 1452 The kernel tries to adjust MTRR layout from continuous 1453 to discrete, to make X server driver able to add WB 1454 entry later. This parameter enables that. 1455 1456 enable_timer_pin_1 [X86] 1457 Enable PIN 1 of APIC timer 1458 Can be useful to work around chipset bugs 1459 (in particular on some ATI chipsets). 1460 The kernel tries to set a reasonable default. 1461 1462 enforcing= [SELINUX] Set initial enforcing status. 1463 Format: {"0" | "1"} 1464 See security/selinux/Kconfig help text. 1465 0 -- permissive (log only, no denials). 1466 1 -- enforcing (deny and log). 1467 Default value is 0. 1468 Value can be changed at runtime via 1469 /sys/fs/selinux/enforce. 1470 1471 erst_disable [ACPI] 1472 Disable Error Record Serialization Table (ERST) 1473 support. 1474 1475 ether= [HW,NET] Ethernet cards parameters 1476 This option is obsoleted by the "netdev=" option, which 1477 has equivalent usage. See its documentation for details. 1478 1479 evm= [EVM] 1480 Format: { "fix" } 1481 Permit 'security.evm' to be updated regardless of 1482 current integrity status. 1483 1484 early_page_ext [KNL] Enforces page_ext initialization to earlier 1485 stages so cover more early boot allocations. 1486 Please note that as side effect some optimizations 1487 might be disabled to achieve that (e.g. parallelized 1488 memory initialization is disabled) so the boot process 1489 might take longer, especially on systems with a lot of 1490 memory. Available with CONFIG_PAGE_EXTENSION=y. 1491 1492 failslab= 1493 fail_usercopy= 1494 fail_page_alloc= 1495 fail_make_request=[KNL] 1496 General fault injection mechanism. 1497 Format: <interval>,<probability>,<space>,<times> 1498 See also Documentation/fault-injection/. 1499 1500 fb_tunnels= [NET] 1501 Format: { initns | none } 1502 See Documentation/admin-guide/sysctl/net.rst for 1503 fb_tunnels_only_for_init_ns 1504 1505 floppy= [HW] 1506 See Documentation/admin-guide/blockdev/floppy.rst. 1507 1508 force_pal_cache_flush 1509 [IA-64] Avoid check_sal_cache_flush which may hang on 1510 buggy SAL_CACHE_FLUSH implementations. Using this 1511 parameter will force ia64_sal_cache_flush to call 1512 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1513 1514 forcepae [X86-32] 1515 Forcefully enable Physical Address Extension (PAE). 1516 Many Pentium M systems disable PAE but may have a 1517 functionally usable PAE implementation. 1518 Warning: use of this parameter will taint the kernel 1519 and may cause unknown problems. 1520 1521 ftrace=[tracer] 1522 [FTRACE] will set and start the specified tracer 1523 as early as possible in order to facilitate early 1524 boot debugging. 1525 1526 ftrace_boot_snapshot 1527 [FTRACE] On boot up, a snapshot will be taken of the 1528 ftrace ring buffer that can be read at: 1529 /sys/kernel/tracing/snapshot. 1530 This is useful if you need tracing information from kernel 1531 boot up that is likely to be overridden by user space 1532 start up functionality. 1533 1534 Optionally, the snapshot can also be defined for a tracing 1535 instance that was created by the trace_instance= command 1536 line parameter. 1537 1538 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1539 1540 The above will cause the "foo" tracing instance to trigger 1541 a snapshot at the end of boot up. 1542 1543 ftrace_dump_on_oops[=orig_cpu] 1544 [FTRACE] will dump the trace buffers on oops. 1545 If no parameter is passed, ftrace will dump 1546 buffers of all CPUs, but if you pass orig_cpu, it will 1547 dump only the buffer of the CPU that triggered the 1548 oops. 1549 1550 ftrace_filter=[function-list] 1551 [FTRACE] Limit the functions traced by the function 1552 tracer at boot up. function-list is a comma-separated 1553 list of functions. This list can be changed at run 1554 time by the set_ftrace_filter file in the debugfs 1555 tracing directory. 1556 1557 ftrace_notrace=[function-list] 1558 [FTRACE] Do not trace the functions specified in 1559 function-list. This list can be changed at run time 1560 by the set_ftrace_notrace file in the debugfs 1561 tracing directory. 1562 1563 ftrace_graph_filter=[function-list] 1564 [FTRACE] Limit the top level callers functions traced 1565 by the function graph tracer at boot up. 1566 function-list is a comma-separated list of functions 1567 that can be changed at run time by the 1568 set_graph_function file in the debugfs tracing directory. 1569 1570 ftrace_graph_notrace=[function-list] 1571 [FTRACE] Do not trace from the functions specified in 1572 function-list. This list is a comma-separated list of 1573 functions that can be changed at run time by the 1574 set_graph_notrace file in the debugfs tracing directory. 1575 1576 ftrace_graph_max_depth=<uint> 1577 [FTRACE] Used with the function graph tracer. This is 1578 the max depth it will trace into a function. This value 1579 can be changed at run time by the max_graph_depth file 1580 in the tracefs tracing directory. default: 0 (no limit) 1581 1582 fw_devlink= [KNL] Create device links between consumer and supplier 1583 devices by scanning the firmware to infer the 1584 consumer/supplier relationships. This feature is 1585 especially useful when drivers are loaded as modules as 1586 it ensures proper ordering of tasks like device probing 1587 (suppliers first, then consumers), supplier boot state 1588 clean up (only after all consumers have probed), 1589 suspend/resume & runtime PM (consumers first, then 1590 suppliers). 1591 Format: { off | permissive | on | rpm } 1592 off -- Don't create device links from firmware info. 1593 permissive -- Create device links from firmware info 1594 but use it only for ordering boot state clean 1595 up (sync_state() calls). 1596 on -- Create device links from firmware info and use it 1597 to enforce probe and suspend/resume ordering. 1598 rpm -- Like "on", but also use to order runtime PM. 1599 1600 fw_devlink.strict=<bool> 1601 [KNL] Treat all inferred dependencies as mandatory 1602 dependencies. This only applies for fw_devlink=on|rpm. 1603 Format: <bool> 1604 1605 fw_devlink.sync_state = 1606 [KNL] When all devices that could probe have finished 1607 probing, this parameter controls what to do with 1608 devices that haven't yet received their sync_state() 1609 calls. 1610 Format: { strict | timeout } 1611 strict -- Default. Continue waiting on consumers to 1612 probe successfully. 1613 timeout -- Give up waiting on consumers and call 1614 sync_state() on any devices that haven't yet 1615 received their sync_state() calls after 1616 deferred_probe_timeout has expired or by 1617 late_initcall() if !CONFIG_MODULES. 1618 1619 gamecon.map[2|3]= 1620 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1621 support via parallel port (up to 5 devices per port) 1622 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1623 See also Documentation/input/devices/joystick-parport.rst 1624 1625 gamma= [HW,DRM] 1626 1627 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART 1628 Format: off | on 1629 default: on 1630 1631 gcov_persist= [GCOV] When non-zero (default), profiling data for 1632 kernel modules is saved and remains accessible via 1633 debugfs, even when the module is unloaded/reloaded. 1634 When zero, profiling data is discarded and associated 1635 debugfs files are removed at module unload time. 1636 1637 goldfish [X86] Enable the goldfish android emulator platform. 1638 Don't use this when you are not running on the 1639 android emulator 1640 1641 gpio-mockup.gpio_mockup_ranges 1642 [HW] Sets the ranges of gpiochip of for this device. 1643 Format: <start1>,<end1>,<start2>,<end2>... 1644 gpio-mockup.gpio_mockup_named_lines 1645 [HW] Let the driver know GPIO lines should be named. 1646 1647 gpt [EFI] Forces disk with valid GPT signature but 1648 invalid Protective MBR to be treated as GPT. If the 1649 primary GPT is corrupted, it enables the backup/alternate 1650 GPT to be used instead. 1651 1652 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1653 the "Enable 0" bit of the configuration register. 1654 Format: 0 | 1 1655 Default: 0 1656 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1657 the "Enable 0" bit of the configuration register. 1658 Format: 0 | 1 1659 Default: 0 1660 grcan.select= [HW] Select which physical interface to use. 1661 Format: 0 | 1 1662 Default: 0 1663 grcan.txsize= [HW] Sets the size of the tx buffer. 1664 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1665 Default: 1024 1666 grcan.rxsize= [HW] Sets the size of the rx buffer. 1667 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1668 Default: 1024 1669 1670 hardened_usercopy= 1671 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1672 hardening is enabled for this boot. Hardened 1673 usercopy checking is used to protect the kernel 1674 from reading or writing beyond known memory 1675 allocation boundaries as a proactive defense 1676 against bounds-checking flaws in the kernel's 1677 copy_to_user()/copy_from_user() interface. 1678 on Perform hardened usercopy checks (default). 1679 off Disable hardened usercopy checks. 1680 1681 hardlockup_all_cpu_backtrace= 1682 [KNL] Should the hard-lockup detector generate 1683 backtraces on all cpus. 1684 Format: 0 | 1 1685 1686 hashdist= [KNL,NUMA] Large hashes allocated during boot 1687 are distributed across NUMA nodes. Defaults on 1688 for 64-bit NUMA, off otherwise. 1689 Format: 0 | 1 (for off | on) 1690 1691 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1692 1693 hd= [EIDE] (E)IDE hard drive subsystem geometry 1694 Format: <cyl>,<head>,<sect> 1695 1696 hest_disable [ACPI] 1697 Disable Hardware Error Source Table (HEST) support; 1698 corresponding firmware-first mode error processing 1699 logic will be disabled. 1700 1701 hibernate= [HIBERNATION] 1702 noresume Don't check if there's a hibernation image 1703 present during boot. 1704 nocompress Don't compress/decompress hibernation images. 1705 no Disable hibernation and resume. 1706 protect_image Turn on image protection during restoration 1707 (that will set all pages holding image data 1708 during restoration read-only). 1709 1710 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1711 size of <nn>. This works even on boxes that have no 1712 highmem otherwise. This also works to reduce highmem 1713 size on bigger boxes. 1714 1715 highres= [KNL] Enable/disable high resolution timer mode. 1716 Valid parameters: "on", "off" 1717 Default: "on" 1718 1719 hlt [BUGS=ARM,SH] 1720 1721 hostname= [KNL] Set the hostname (aka UTS nodename). 1722 Format: <string> 1723 This allows setting the system's hostname during early 1724 startup. This sets the name returned by gethostname. 1725 Using this parameter to set the hostname makes it 1726 possible to ensure the hostname is correctly set before 1727 any userspace processes run, avoiding the possibility 1728 that a process may call gethostname before the hostname 1729 has been explicitly set, resulting in the calling 1730 process getting an incorrect result. The string must 1731 not exceed the maximum allowed hostname length (usually 1732 64 characters) and will be truncated otherwise. 1733 1734 hpet= [X86-32,HPET] option to control HPET usage 1735 Format: { enable (default) | disable | force | 1736 verbose } 1737 disable: disable HPET and use PIT instead 1738 force: allow force enabled of undocumented chips (ICH4, 1739 VIA, nVidia) 1740 verbose: show contents of HPET registers during setup 1741 1742 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1743 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1744 1745 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1746 If this follows hugepagesz (below), it specifies 1747 the number of pages of hugepagesz to be allocated. 1748 If this is the first HugeTLB parameter on the command 1749 line, it specifies the number of pages to allocate for 1750 the default huge page size. If using node format, the 1751 number of pages to allocate per-node can be specified. 1752 See also Documentation/admin-guide/mm/hugetlbpage.rst. 1753 Format: <integer> or (node format) 1754 <node>:<integer>[,<node>:<integer>] 1755 1756 hugepagesz= 1757 [HW] The size of the HugeTLB pages. This is used in 1758 conjunction with hugepages (above) to allocate huge 1759 pages of a specific size at boot. The pair 1760 hugepagesz=X hugepages=Y can be specified once for 1761 each supported huge page size. Huge page sizes are 1762 architecture dependent. See also 1763 Documentation/admin-guide/mm/hugetlbpage.rst. 1764 Format: size[KMG] 1765 1766 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation 1767 of gigantic hugepages. Or using node format, the size 1768 of a CMA area per node can be specified. 1769 Format: nn[KMGTPE] or (node format) 1770 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 1771 1772 Reserve a CMA area of given size and allocate gigantic 1773 hugepages using the CMA allocator. If enabled, the 1774 boot-time allocation of gigantic hugepages is skipped. 1775 1776 hugetlb_free_vmemmap= 1777 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 1778 enabled. 1779 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 1780 Allows heavy hugetlb users to free up some more 1781 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 1782 Format: { on | off (default) } 1783 1784 on: enable HVO 1785 off: disable HVO 1786 1787 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 1788 the default is on. 1789 1790 Note that the vmemmap pages may be allocated from the added 1791 memory block itself when memory_hotplug.memmap_on_memory is 1792 enabled, those vmemmap pages cannot be optimized even if this 1793 feature is enabled. Other vmemmap pages not allocated from 1794 the added memory block itself do not be affected. 1795 1796 hung_task_panic= 1797 [KNL] Should the hung task detector generate panics. 1798 Format: 0 | 1 1799 1800 A value of 1 instructs the kernel to panic when a 1801 hung task is detected. The default value is controlled 1802 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1803 option. The value selected by this boot parameter can 1804 be changed later by the kernel.hung_task_panic sysctl. 1805 1806 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1807 terminal devices. Valid values: 0..8 1808 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1809 If specified, z/VM IUCV HVC accepts connections 1810 from listed z/VM user IDs only. 1811 1812 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1813 which allow the hypervisor to 'idle' the 1814 guest on lock contention. 1815 1816 i2c_bus= [HW] Override the default board specific I2C bus speed 1817 or register an additional I2C bus that is not 1818 registered from board initialization code. 1819 Format: 1820 <bus_id>,<clkrate> 1821 1822 i8042.debug [HW] Toggle i8042 debug mode 1823 i8042.unmask_kbd_data 1824 [HW] Enable printing of interrupt data from the KBD port 1825 (disabled by default, and as a pre-condition 1826 requires that i8042.debug=1 be enabled) 1827 i8042.direct [HW] Put keyboard port into non-translated mode 1828 i8042.dumbkbd [HW] Pretend that controller can only read data from 1829 keyboard and cannot control its state 1830 (Don't attempt to blink the leds) 1831 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1832 i8042.nokbd [HW] Don't check/create keyboard port 1833 i8042.noloop [HW] Disable the AUX Loopback command while probing 1834 for the AUX port 1835 i8042.nomux [HW] Don't check presence of an active multiplexing 1836 controller 1837 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1838 controllers 1839 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1840 i8042.reset [HW] Reset the controller during init, cleanup and 1841 suspend-to-ram transitions, only during s2r 1842 transitions, or never reset 1843 Format: { 1 | Y | y | 0 | N | n } 1844 1, Y, y: always reset controller 1845 0, N, n: don't ever reset controller 1846 Default: only on s2r transitions on x86; most other 1847 architectures force reset to be always executed 1848 i8042.unlock [HW] Unlock (ignore) the keylock 1849 i8042.kbdreset [HW] Reset device connected to KBD port 1850 i8042.probe_defer 1851 [HW] Allow deferred probing upon i8042 probe errors 1852 1853 i810= [HW,DRM] 1854 1855 i915.invert_brightness= 1856 [DRM] Invert the sense of the variable that is used to 1857 set the brightness of the panel backlight. Normally a 1858 brightness value of 0 indicates backlight switched off, 1859 and the maximum of the brightness value sets the backlight 1860 to maximum brightness. If this parameter is set to 0 1861 (default) and the machine requires it, or this parameter 1862 is set to 1, a brightness value of 0 sets the backlight 1863 to maximum brightness, and the maximum of the brightness 1864 value switches the backlight off. 1865 -1 -- never invert brightness 1866 0 -- machine default 1867 1 -- force brightness inversion 1868 1869 icn= [HW,ISDN] 1870 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1871 1872 1873 idle= [X86] 1874 Format: idle=poll, idle=halt, idle=nomwait 1875 Poll forces a polling idle loop that can slightly 1876 improve the performance of waking up a idle CPU, but 1877 will use a lot of power and make the system run hot. 1878 Not recommended. 1879 idle=halt: Halt is forced to be used for CPU idle. 1880 In such case C2/C3 won't be used again. 1881 idle=nomwait: Disable mwait for CPU C-states 1882 1883 idxd.sva= [HW] 1884 Format: <bool> 1885 Allow force disabling of Shared Virtual Memory (SVA) 1886 support for the idxd driver. By default it is set to 1887 true (1). 1888 1889 idxd.tc_override= [HW] 1890 Format: <bool> 1891 Allow override of default traffic class configuration 1892 for the device. By default it is set to false (0). 1893 1894 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1895 Format: { strict | legacy | 2008 | relaxed } 1896 Default: strict 1897 1898 Choose which programs will be accepted for execution 1899 based on the IEEE 754 NaN encoding(s) supported by 1900 the FPU and the NaN encoding requested with the value 1901 of an ELF file header flag individually set by each 1902 binary. Hardware implementations are permitted to 1903 support either or both of the legacy and the 2008 NaN 1904 encoding mode. 1905 1906 Available settings are as follows: 1907 strict accept binaries that request a NaN encoding 1908 supported by the FPU 1909 legacy only accept legacy-NaN binaries, if supported 1910 by the FPU 1911 2008 only accept 2008-NaN binaries, if supported 1912 by the FPU 1913 relaxed accept any binaries regardless of whether 1914 supported by the FPU 1915 1916 The FPU emulator is always able to support both NaN 1917 encodings, so if no FPU hardware is present or it has 1918 been disabled with 'nofpu', then the settings of 1919 'legacy' and '2008' strap the emulator accordingly, 1920 'relaxed' straps the emulator for both legacy-NaN and 1921 2008-NaN, whereas 'strict' enables legacy-NaN only on 1922 legacy processors and both NaN encodings on MIPS32 or 1923 MIPS64 CPUs. 1924 1925 The setting for ABS.fmt/NEG.fmt instruction execution 1926 mode generally follows that for the NaN encoding, 1927 except where unsupported by hardware. 1928 1929 ignore_loglevel [KNL] 1930 Ignore loglevel setting - this will print /all/ 1931 kernel messages to the console. Useful for debugging. 1932 We also add it as printk module parameter, so users 1933 could change it dynamically, usually by 1934 /sys/module/printk/parameters/ignore_loglevel. 1935 1936 ignore_rlimit_data 1937 Ignore RLIMIT_DATA setting for data mappings, 1938 print warning at first misuse. Can be changed via 1939 /sys/module/kernel/parameters/ignore_rlimit_data. 1940 1941 ihash_entries= [KNL] 1942 Set number of hash buckets for inode cache. 1943 1944 ima_appraise= [IMA] appraise integrity measurements 1945 Format: { "off" | "enforce" | "fix" | "log" } 1946 default: "enforce" 1947 1948 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1949 The builtin appraise policy appraises all files 1950 owned by uid=0. 1951 1952 ima_canonical_fmt [IMA] 1953 Use the canonical format for the binary runtime 1954 measurements, instead of host native format. 1955 1956 ima_hash= [IMA] 1957 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1958 | sha512 | ... } 1959 default: "sha1" 1960 1961 The list of supported hash algorithms is defined 1962 in crypto/hash_info.h. 1963 1964 ima_policy= [IMA] 1965 The builtin policies to load during IMA setup. 1966 Format: "tcb | appraise_tcb | secure_boot | 1967 fail_securely | critical_data" 1968 1969 The "tcb" policy measures all programs exec'd, files 1970 mmap'd for exec, and all files opened with the read 1971 mode bit set by either the effective uid (euid=0) or 1972 uid=0. 1973 1974 The "appraise_tcb" policy appraises the integrity of 1975 all files owned by root. 1976 1977 The "secure_boot" policy appraises the integrity 1978 of files (eg. kexec kernel image, kernel modules, 1979 firmware, policy, etc) based on file signatures. 1980 1981 The "fail_securely" policy forces file signature 1982 verification failure also on privileged mounted 1983 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1984 flag. 1985 1986 The "critical_data" policy measures kernel integrity 1987 critical data. 1988 1989 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1990 Load a policy which meets the needs of the Trusted 1991 Computing Base. This means IMA will measure all 1992 programs exec'd, files mmap'd for exec, and all files 1993 opened for read by uid=0. 1994 1995 ima_template= [IMA] 1996 Select one of defined IMA measurements template formats. 1997 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 1998 "ima-sigv2" } 1999 Default: "ima-ng" 2000 2001 ima_template_fmt= 2002 [IMA] Define a custom template format. 2003 Format: { "field1|...|fieldN" } 2004 2005 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2006 Format: <min_file_size> 2007 Set the minimal file size for using asynchronous hash. 2008 If left unspecified, ahash usage is disabled. 2009 2010 ahash performance varies for different data sizes on 2011 different crypto accelerators. This option can be used 2012 to achieve the best performance for a particular HW. 2013 2014 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2015 Format: <bufsize> 2016 Set hashing buffer size. Default: 4k. 2017 2018 ahash performance varies for different chunk sizes on 2019 different crypto accelerators. This option can be used 2020 to achieve best performance for particular HW. 2021 2022 init= [KNL] 2023 Format: <full_path> 2024 Run specified binary instead of /sbin/init as init 2025 process. 2026 2027 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2028 for working out where the kernel is dying during 2029 startup. 2030 2031 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2032 initcall functions. Useful for debugging built-in 2033 modules and initcalls. 2034 2035 initramfs_async= [KNL] 2036 Format: <bool> 2037 Default: 1 2038 This parameter controls whether the initramfs 2039 image is unpacked asynchronously, concurrently 2040 with devices being probed and 2041 initialized. This should normally just work, 2042 but as a debugging aid, one can get the 2043 historical behaviour of the initramfs 2044 unpacking being completed before device_ and 2045 late_ initcalls. 2046 2047 initrd= [BOOT] Specify the location of the initial ramdisk 2048 2049 initrdmem= [KNL] Specify a physical address and size from which to 2050 load the initrd. If an initrd is compiled in or 2051 specified in the bootparams, it takes priority over this 2052 setting. 2053 Format: ss[KMG],nn[KMG] 2054 Default is 0, 0 2055 2056 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 2057 zeroes. 2058 Format: 0 | 1 2059 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2060 2061 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 2062 Format: 0 | 1 2063 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2064 2065 init_pkru= [X86] Specify the default memory protection keys rights 2066 register contents for all processes. 0x55555554 by 2067 default (disallow access to all but pkey 0). Can 2068 override in debugfs after boot. 2069 2070 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2071 Format: <irq> 2072 2073 int_pln_enable [X86] Enable power limit notification interrupt 2074 2075 integrity_audit=[IMA] 2076 Format: { "0" | "1" } 2077 0 -- basic integrity auditing messages. (Default) 2078 1 -- additional integrity auditing messages. 2079 2080 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2081 on 2082 Enable intel iommu driver. 2083 off 2084 Disable intel iommu driver. 2085 igfx_off [Default Off] 2086 By default, gfx is mapped as normal device. If a gfx 2087 device has a dedicated DMAR unit, the DMAR unit is 2088 bypassed by not enabling DMAR with this option. In 2089 this case, gfx device will use physical address for 2090 DMA. 2091 strict [Default Off] 2092 Deprecated, equivalent to iommu.strict=1. 2093 sp_off [Default Off] 2094 By default, super page will be supported if Intel IOMMU 2095 has the capability. With this option, super page will 2096 not be supported. 2097 sm_on 2098 Enable the Intel IOMMU scalable mode if the hardware 2099 advertises that it has support for the scalable mode 2100 translation. 2101 sm_off 2102 Disallow use of the Intel IOMMU scalable mode. 2103 tboot_noforce [Default Off] 2104 Do not force the Intel IOMMU enabled under tboot. 2105 By default, tboot will force Intel IOMMU on, which 2106 could harm performance of some high-throughput 2107 devices like 40GBit network cards, even if identity 2108 mapping is enabled. 2109 Note that using this option lowers the security 2110 provided by tboot because it makes the system 2111 vulnerable to DMA attacks. 2112 2113 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2114 0 disables intel_idle and fall back on acpi_idle. 2115 1 to 9 specify maximum depth of C-state. 2116 2117 intel_pstate= [X86] 2118 disable 2119 Do not enable intel_pstate as the default 2120 scaling driver for the supported processors 2121 passive 2122 Use intel_pstate as a scaling driver, but configure it 2123 to work with generic cpufreq governors (instead of 2124 enabling its internal governor). This mode cannot be 2125 used along with the hardware-managed P-states (HWP) 2126 feature. 2127 force 2128 Enable intel_pstate on systems that prohibit it by default 2129 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2130 instead of acpi-cpufreq may disable platform features, such 2131 as thermal controls and power capping, that rely on ACPI 2132 P-States information being indicated to OSPM and therefore 2133 should be used with caution. This option does not work with 2134 processors that aren't supported by the intel_pstate driver 2135 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2136 no_hwp 2137 Do not enable hardware P state control (HWP) 2138 if available. 2139 hwp_only 2140 Only load intel_pstate on systems which support 2141 hardware P state control (HWP) if available. 2142 support_acpi_ppc 2143 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2144 Description Table, specifies preferred power management 2145 profile as "Enterprise Server" or "Performance Server", 2146 then this feature is turned on by default. 2147 per_cpu_perf_limits 2148 Allow per-logical-CPU P-State performance control limits using 2149 cpufreq sysfs interface 2150 2151 intremap= [X86-64, Intel-IOMMU] 2152 on enable Interrupt Remapping (default) 2153 off disable Interrupt Remapping 2154 nosid disable Source ID checking 2155 no_x2apic_optout 2156 BIOS x2APIC opt-out request will be ignored 2157 nopost disable Interrupt Posting 2158 2159 iomem= Disable strict checking of access to MMIO memory 2160 strict regions from userspace. 2161 relaxed 2162 2163 iommu= [X86] 2164 off 2165 force 2166 noforce 2167 biomerge 2168 panic 2169 nopanic 2170 merge 2171 nomerge 2172 soft 2173 pt [X86] 2174 nopt [X86] 2175 nobypass [PPC/POWERNV] 2176 Disable IOMMU bypass, using IOMMU for PCI devices. 2177 2178 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices. 2179 Format: { "0" | "1" } 2180 0 - Try to allocate a 32-bit DMA address first, before 2181 falling back to the full range if needed. 2182 1 - Allocate directly from the full usable range, 2183 forcing Dual Address Cycle for PCI cards supporting 2184 greater than 32-bit addressing. 2185 2186 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour 2187 Format: { "0" | "1" } 2188 0 - Lazy mode. 2189 Request that DMA unmap operations use deferred 2190 invalidation of hardware TLBs, for increased 2191 throughput at the cost of reduced device isolation. 2192 Will fall back to strict mode if not supported by 2193 the relevant IOMMU driver. 2194 1 - Strict mode. 2195 DMA unmap operations invalidate IOMMU hardware TLBs 2196 synchronously. 2197 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2198 Note: on x86, strict mode specified via one of the 2199 legacy driver-specific options takes precedence. 2200 2201 iommu.passthrough= 2202 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 2203 Format: { "0" | "1" } 2204 0 - Use IOMMU translation for DMA. 2205 1 - Bypass the IOMMU for DMA. 2206 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2207 2208 io7= [HW] IO7 for Marvel-based Alpha systems 2209 See comment before marvel_specify_io7 in 2210 arch/alpha/kernel/core_marvel.c. 2211 2212 io_delay= [X86] I/O delay method 2213 0x80 2214 Standard port 0x80 based delay 2215 0xed 2216 Alternate port 0xed based delay (needed on some systems) 2217 udelay 2218 Simple two microseconds delay 2219 none 2220 No delay 2221 2222 ip= [IP_PNP] 2223 See Documentation/admin-guide/nfs/nfsroot.rst. 2224 2225 ipcmni_extend [KNL] Extend the maximum number of unique System V 2226 IPC identifiers from 32,768 to 16,777,216. 2227 2228 irqaffinity= [SMP] Set the default irq affinity mask 2229 The argument is a cpu list, as described above. 2230 2231 irqchip.gicv2_force_probe= 2232 [ARM, ARM64] 2233 Format: <bool> 2234 Force the kernel to look for the second 4kB page 2235 of a GICv2 controller even if the memory range 2236 exposed by the device tree is too small. 2237 2238 irqchip.gicv3_nolpi= 2239 [ARM, ARM64] 2240 Force the kernel to ignore the availability of 2241 LPIs (and by consequence ITSs). Intended for system 2242 that use the kernel as a bootloader, and thus want 2243 to let secondary kernels in charge of setting up 2244 LPIs. 2245 2246 irqchip.gicv3_pseudo_nmi= [ARM64] 2247 Enables support for pseudo-NMIs in the kernel. This 2248 requires the kernel to be built with 2249 CONFIG_ARM64_PSEUDO_NMI. 2250 2251 irqfixup [HW] 2252 When an interrupt is not handled search all handlers 2253 for it. Intended to get systems with badly broken 2254 firmware running. 2255 2256 irqpoll [HW] 2257 When an interrupt is not handled search all handlers 2258 for it. Also check all handlers each timer 2259 interrupt. Intended to get systems with badly broken 2260 firmware running. 2261 2262 isapnp= [ISAPNP] 2263 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2264 2265 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2266 [Deprecated - use cpusets instead] 2267 Format: [flag-list,]<cpu-list> 2268 2269 Specify one or more CPUs to isolate from disturbances 2270 specified in the flag list (default: domain): 2271 2272 nohz 2273 Disable the tick when a single task runs. 2274 2275 A residual 1Hz tick is offloaded to workqueues, which you 2276 need to affine to housekeeping through the global 2277 workqueue's affinity configured via the 2278 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2279 by using the 'domain' flag described below. 2280 2281 NOTE: by default the global workqueue runs on all CPUs, 2282 so to protect individual CPUs the 'cpumask' file has to 2283 be configured manually after bootup. 2284 2285 domain 2286 Isolate from the general SMP balancing and scheduling 2287 algorithms. Note that performing domain isolation this way 2288 is irreversible: it's not possible to bring back a CPU to 2289 the domains once isolated through isolcpus. It's strongly 2290 advised to use cpusets instead to disable scheduler load 2291 balancing through the "cpuset.sched_load_balance" file. 2292 It offers a much more flexible interface where CPUs can 2293 move in and out of an isolated set anytime. 2294 2295 You can move a process onto or off an "isolated" CPU via 2296 the CPU affinity syscalls or cpuset. 2297 <cpu number> begins at 0 and the maximum value is 2298 "number of CPUs in system - 1". 2299 2300 managed_irq 2301 2302 Isolate from being targeted by managed interrupts 2303 which have an interrupt mask containing isolated 2304 CPUs. The affinity of managed interrupts is 2305 handled by the kernel and cannot be changed via 2306 the /proc/irq/* interfaces. 2307 2308 This isolation is best effort and only effective 2309 if the automatically assigned interrupt mask of a 2310 device queue contains isolated and housekeeping 2311 CPUs. If housekeeping CPUs are online then such 2312 interrupts are directed to the housekeeping CPU 2313 so that IO submitted on the housekeeping CPU 2314 cannot disturb the isolated CPU. 2315 2316 If a queue's affinity mask contains only isolated 2317 CPUs then this parameter has no effect on the 2318 interrupt routing decision, though interrupts are 2319 only delivered when tasks running on those 2320 isolated CPUs submit IO. IO submitted on 2321 housekeeping CPUs has no influence on those 2322 queues. 2323 2324 The format of <cpu-list> is described above. 2325 2326 iucv= [HW,NET] 2327 2328 ivrs_ioapic [HW,X86-64] 2329 Provide an override to the IOAPIC-ID<->DEVICE-ID 2330 mapping provided in the IVRS ACPI table. 2331 By default, PCI segment is 0, and can be omitted. 2332 2333 For example, to map IOAPIC-ID decimal 10 to 2334 PCI segment 0x1 and PCI device 00:14.0, 2335 write the parameter as: 2336 ivrs_ioapic=10@0001:00:14.0 2337 2338 Deprecated formats: 2339 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2340 write the parameter as: 2341 ivrs_ioapic[10]=00:14.0 2342 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2343 PCI device 00:14.0 write the parameter as: 2344 ivrs_ioapic[10]=0001:00:14.0 2345 2346 ivrs_hpet [HW,X86-64] 2347 Provide an override to the HPET-ID<->DEVICE-ID 2348 mapping provided in the IVRS ACPI table. 2349 By default, PCI segment is 0, and can be omitted. 2350 2351 For example, to map HPET-ID decimal 10 to 2352 PCI segment 0x1 and PCI device 00:14.0, 2353 write the parameter as: 2354 ivrs_hpet=10@0001:00:14.0 2355 2356 Deprecated formats: 2357 * To map HPET-ID decimal 0 to PCI device 00:14.0 2358 write the parameter as: 2359 ivrs_hpet[0]=00:14.0 2360 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2361 PCI device 00:14.0 write the parameter as: 2362 ivrs_ioapic[10]=0001:00:14.0 2363 2364 ivrs_acpihid [HW,X86-64] 2365 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2366 mapping provided in the IVRS ACPI table. 2367 By default, PCI segment is 0, and can be omitted. 2368 2369 For example, to map UART-HID:UID AMD0020:0 to 2370 PCI segment 0x1 and PCI device ID 00:14.5, 2371 write the parameter as: 2372 ivrs_acpihid=AMD0020:0@0001:00:14.5 2373 2374 Deprecated formats: 2375 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2376 PCI device ID 00:14.5, write the parameter as: 2377 ivrs_acpihid[00:14.5]=AMD0020:0 2378 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2379 PCI device ID 00:14.5, write the parameter as: 2380 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2381 2382 js= [HW,JOY] Analog joystick 2383 See Documentation/input/joydev/joystick.rst. 2384 2385 kasan_multi_shot 2386 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2387 report on every invalid memory access. Without this 2388 parameter KASAN will print report only for the first 2389 invalid access. 2390 2391 keep_bootcon [KNL] 2392 Do not unregister boot console at start. This is only 2393 useful for debugging when something happens in the window 2394 between unregistering the boot console and initializing 2395 the real console. 2396 2397 keepinitrd [HW,ARM] 2398 2399 kernelcore= [KNL,X86,IA-64,PPC] 2400 Format: nn[KMGTPE] | nn% | "mirror" 2401 This parameter specifies the amount of memory usable by 2402 the kernel for non-movable allocations. The requested 2403 amount is spread evenly throughout all nodes in the 2404 system as ZONE_NORMAL. The remaining memory is used for 2405 movable memory in its own zone, ZONE_MOVABLE. In the 2406 event, a node is too small to have both ZONE_NORMAL and 2407 ZONE_MOVABLE, kernelcore memory will take priority and 2408 other nodes will have a larger ZONE_MOVABLE. 2409 2410 ZONE_MOVABLE is used for the allocation of pages that 2411 may be reclaimed or moved by the page migration 2412 subsystem. Note that allocations like PTEs-from-HighMem 2413 still use the HighMem zone if it exists, and the Normal 2414 zone if it does not. 2415 2416 It is possible to specify the exact amount of memory in 2417 the form of "nn[KMGTPE]", a percentage of total system 2418 memory in the form of "nn%", or "mirror". If "mirror" 2419 option is specified, mirrored (reliable) memory is used 2420 for non-movable allocations and remaining memory is used 2421 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2422 are exclusive, so you cannot specify multiple forms. 2423 2424 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2425 Format: <Controller#>[,poll interval] 2426 The controller # is the number of the ehci usb debug 2427 port as it is probed via PCI. The poll interval is 2428 optional and is the number seconds in between 2429 each poll cycle to the debug port in case you need 2430 the functionality for interrupting the kernel with 2431 gdb or control-c on the dbgp connection. When 2432 not using this parameter you use sysrq-g to break into 2433 the kernel debugger. 2434 2435 kgdboc= [KGDB,HW] kgdb over consoles. 2436 Requires a tty driver that supports console polling, 2437 or a supported polling keyboard driver (non-usb). 2438 Serial only format: <serial_device>[,baud] 2439 keyboard only format: kbd 2440 keyboard and serial format: kbd,<serial_device>[,baud] 2441 Optional Kernel mode setting: 2442 kms, kbd format: kms,kbd 2443 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2444 2445 kgdboc_earlycon= [KGDB,HW] 2446 If the boot console provides the ability to read 2447 characters and can work in polling mode, you can use 2448 this parameter to tell kgdb to use it as a backend 2449 until the normal console is registered. Intended to 2450 be used together with the kgdboc parameter which 2451 specifies the normal console to transition to. 2452 2453 The name of the early console should be specified 2454 as the value of this parameter. Note that the name of 2455 the early console might be different than the tty 2456 name passed to kgdboc. It's OK to leave the value 2457 blank and the first boot console that implements 2458 read() will be picked. 2459 2460 kgdbwait [KGDB] Stop kernel execution and enter the 2461 kernel debugger at the earliest opportunity. 2462 2463 kmac= [MIPS] Korina ethernet MAC address. 2464 Configure the RouterBoard 532 series on-chip 2465 Ethernet adapter MAC address. 2466 2467 kmemleak= [KNL] Boot-time kmemleak enable/disable 2468 Valid arguments: on, off 2469 Default: on 2470 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2471 the default is off. 2472 2473 kprobe_event=[probe-list] 2474 [FTRACE] Add kprobe events and enable at boot time. 2475 The probe-list is a semicolon delimited list of probe 2476 definitions. Each definition is same as kprobe_events 2477 interface, but the parameters are comma delimited. 2478 For example, to add a kprobe event on vfs_read with 2479 arg1 and arg2, add to the command line; 2480 2481 kprobe_event=p,vfs_read,$arg1,$arg2 2482 2483 See also Documentation/trace/kprobetrace.rst "Kernel 2484 Boot Parameter" section. 2485 2486 kpti= [ARM64] Control page table isolation of user 2487 and kernel address spaces. 2488 Default: enabled on cores which need mitigation. 2489 0: force disabled 2490 1: force enabled 2491 2492 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2493 CONFIG_KUNIT to be set to be fully enabled. The 2494 default value can be overridden via 2495 KUNIT_DEFAULT_ENABLED. 2496 Default is 1 (enabled) 2497 2498 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2499 Default is 0 (don't ignore, but inject #GP) 2500 2501 kvm.eager_page_split= 2502 [KVM,X86] Controls whether or not KVM will try to 2503 proactively split all huge pages during dirty logging. 2504 Eager page splitting reduces interruptions to vCPU 2505 execution by eliminating the write-protection faults 2506 and MMU lock contention that would otherwise be 2507 required to split huge pages lazily. 2508 2509 VM workloads that rarely perform writes or that write 2510 only to a small region of VM memory may benefit from 2511 disabling eager page splitting to allow huge pages to 2512 still be used for reads. 2513 2514 The behavior of eager page splitting depends on whether 2515 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 2516 disabled, all huge pages in a memslot will be eagerly 2517 split when dirty logging is enabled on that memslot. If 2518 enabled, eager page splitting will be performed during 2519 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 2520 cleared. 2521 2522 Eager page splitting is only supported when kvm.tdp_mmu=Y. 2523 2524 Default is Y (on). 2525 2526 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2527 Default is false (don't support). 2528 2529 kvm.nx_huge_pages= 2530 [KVM] Controls the software workaround for the 2531 X86_BUG_ITLB_MULTIHIT bug. 2532 force : Always deploy workaround. 2533 off : Never deploy workaround. 2534 auto : Deploy workaround based on the presence of 2535 X86_BUG_ITLB_MULTIHIT. 2536 2537 Default is 'auto'. 2538 2539 If the software workaround is enabled for the host, 2540 guests do need not to enable it for nested guests. 2541 2542 kvm.nx_huge_pages_recovery_ratio= 2543 [KVM] Controls how many 4KiB pages are periodically zapped 2544 back to huge pages. 0 disables the recovery, otherwise if 2545 the value is N KVM will zap 1/Nth of the 4KiB pages every 2546 period (see below). The default is 60. 2547 2548 kvm.nx_huge_pages_recovery_period_ms= 2549 [KVM] Controls the time period at which KVM zaps 4KiB pages 2550 back to huge pages. If the value is a non-zero N, KVM will 2551 zap a portion (see ratio above) of the pages every N msecs. 2552 If the value is 0 (the default), KVM will pick a period based 2553 on the ratio, such that a page is zapped after 1 hour on average. 2554 2555 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2556 Default is 1 (enabled) 2557 2558 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2559 for all guests. 2560 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2561 2562 kvm-arm.mode= 2563 [KVM,ARM] Select one of KVM/arm64's modes of operation. 2564 2565 none: Forcefully disable KVM. 2566 2567 nvhe: Standard nVHE-based mode, without support for 2568 protected guests. 2569 2570 protected: nVHE-based mode with support for guests whose 2571 state is kept private from the host. 2572 2573 nested: VHE-based mode with support for nested 2574 virtualization. Requires at least ARMv8.3 2575 hardware. 2576 2577 Defaults to VHE/nVHE based on hardware support. Setting 2578 mode to "protected" will disable kexec and hibernation 2579 for the host. "nested" is experimental and should be 2580 used with extreme caution. 2581 2582 kvm-arm.vgic_v3_group0_trap= 2583 [KVM,ARM] Trap guest accesses to GICv3 group-0 2584 system registers 2585 2586 kvm-arm.vgic_v3_group1_trap= 2587 [KVM,ARM] Trap guest accesses to GICv3 group-1 2588 system registers 2589 2590 kvm-arm.vgic_v3_common_trap= 2591 [KVM,ARM] Trap guest accesses to GICv3 common 2592 system registers 2593 2594 kvm-arm.vgic_v4_enable= 2595 [KVM,ARM] Allow use of GICv4 for direct injection of 2596 LPIs. 2597 2598 kvm_cma_resv_ratio=n [PPC] 2599 Reserves given percentage from system memory area for 2600 contiguous memory allocation for KVM hash pagetable 2601 allocation. 2602 By default it reserves 5% of total system memory. 2603 Format: <integer> 2604 Default: 5 2605 2606 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2607 (virtualized MMU) support on capable Intel chips. 2608 Default is 1 (enabled) 2609 2610 kvm-intel.emulate_invalid_guest_state= 2611 [KVM,Intel] Disable emulation of invalid guest state. 2612 Ignored if kvm-intel.enable_unrestricted_guest=1, as 2613 guest state is never invalid for unrestricted guests. 2614 This param doesn't apply to nested guests (L2), as KVM 2615 never emulates invalid L2 guest state. 2616 Default is 1 (enabled) 2617 2618 kvm-intel.flexpriority= 2619 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2620 Default is 1 (enabled) 2621 2622 kvm-intel.nested= 2623 [KVM,Intel] Enable VMX nesting (nVMX). 2624 Default is 0 (disabled) 2625 2626 kvm-intel.unrestricted_guest= 2627 [KVM,Intel] Disable unrestricted guest feature 2628 (virtualized real and unpaged mode) on capable 2629 Intel chips. Default is 1 (enabled) 2630 2631 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2632 CVE-2018-3620. 2633 2634 Valid arguments: never, cond, always 2635 2636 always: L1D cache flush on every VMENTER. 2637 cond: Flush L1D on VMENTER only when the code between 2638 VMEXIT and VMENTER can leak host memory. 2639 never: Disables the mitigation 2640 2641 Default is cond (do L1 cache flush in specific instances) 2642 2643 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2644 feature (tagged TLBs) on capable Intel chips. 2645 Default is 1 (enabled) 2646 2647 l1d_flush= [X86,INTEL] 2648 Control mitigation for L1D based snooping vulnerability. 2649 2650 Certain CPUs are vulnerable to an exploit against CPU 2651 internal buffers which can forward information to a 2652 disclosure gadget under certain conditions. 2653 2654 In vulnerable processors, the speculatively 2655 forwarded data can be used in a cache side channel 2656 attack, to access data to which the attacker does 2657 not have direct access. 2658 2659 This parameter controls the mitigation. The 2660 options are: 2661 2662 on - enable the interface for the mitigation 2663 2664 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2665 affected CPUs 2666 2667 The kernel PTE inversion protection is unconditionally 2668 enabled and cannot be disabled. 2669 2670 full 2671 Provides all available mitigations for the 2672 L1TF vulnerability. Disables SMT and 2673 enables all mitigations in the 2674 hypervisors, i.e. unconditional L1D flush. 2675 2676 SMT control and L1D flush control via the 2677 sysfs interface is still possible after 2678 boot. Hypervisors will issue a warning 2679 when the first VM is started in a 2680 potentially insecure configuration, 2681 i.e. SMT enabled or L1D flush disabled. 2682 2683 full,force 2684 Same as 'full', but disables SMT and L1D 2685 flush runtime control. Implies the 2686 'nosmt=force' command line option. 2687 (i.e. sysfs control of SMT is disabled.) 2688 2689 flush 2690 Leaves SMT enabled and enables the default 2691 hypervisor mitigation, i.e. conditional 2692 L1D flush. 2693 2694 SMT control and L1D flush control via the 2695 sysfs interface is still possible after 2696 boot. Hypervisors will issue a warning 2697 when the first VM is started in a 2698 potentially insecure configuration, 2699 i.e. SMT enabled or L1D flush disabled. 2700 2701 flush,nosmt 2702 2703 Disables SMT and enables the default 2704 hypervisor mitigation. 2705 2706 SMT control and L1D flush control via the 2707 sysfs interface is still possible after 2708 boot. Hypervisors will issue a warning 2709 when the first VM is started in a 2710 potentially insecure configuration, 2711 i.e. SMT enabled or L1D flush disabled. 2712 2713 flush,nowarn 2714 Same as 'flush', but hypervisors will not 2715 warn when a VM is started in a potentially 2716 insecure configuration. 2717 2718 off 2719 Disables hypervisor mitigations and doesn't 2720 emit any warnings. 2721 It also drops the swap size and available 2722 RAM limit restriction on both hypervisor and 2723 bare metal. 2724 2725 Default is 'flush'. 2726 2727 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2728 2729 l2cr= [PPC] 2730 2731 l3cr= [PPC] 2732 2733 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2734 disabled it. 2735 2736 lapic= [X86,APIC] Do not use TSC deadline 2737 value for LAPIC timer one-shot implementation. Default 2738 back to the programmable timer unit in the LAPIC. 2739 Format: notscdeadline 2740 2741 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2742 in C2 power state. 2743 2744 libata.dma= [LIBATA] DMA control 2745 libata.dma=0 Disable all PATA and SATA DMA 2746 libata.dma=1 PATA and SATA Disk DMA only 2747 libata.dma=2 ATAPI (CDROM) DMA only 2748 libata.dma=4 Compact Flash DMA only 2749 Combinations also work, so libata.dma=3 enables DMA 2750 for disks and CDROMs, but not CFs. 2751 2752 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2753 libata.ignore_hpa=0 keep BIOS limits (default) 2754 libata.ignore_hpa=1 ignore limits, using full disk 2755 2756 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2757 when set. 2758 Format: <int> 2759 2760 libata.force= [LIBATA] Force configurations. The format is a comma- 2761 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 2762 PORT and DEVICE are decimal numbers matching port, link 2763 or device. Basically, it matches the ATA ID string 2764 printed on console by libata. If the whole ID part is 2765 omitted, the last PORT and DEVICE values are used. If 2766 ID hasn't been specified yet, the configuration applies 2767 to all ports, links and devices. 2768 2769 If only DEVICE is omitted, the parameter applies to 2770 the port and all links and devices behind it. DEVICE 2771 number of 0 either selects the first device or the 2772 first fan-out link behind PMP device. It does not 2773 select the host link. DEVICE number of 15 selects the 2774 host link and device attached to it. 2775 2776 The VAL specifies the configuration to force. As long 2777 as there is no ambiguity, shortcut notation is allowed. 2778 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2779 The following configurations can be forced. 2780 2781 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2782 Any ID with matching PORT is used. 2783 2784 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2785 2786 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2787 udma[/][16,25,33,44,66,100,133] notation is also 2788 allowed. 2789 2790 * nohrst, nosrst, norst: suppress hard, soft and both 2791 resets. 2792 2793 * rstonce: only attempt one reset during hot-unplug 2794 link recovery. 2795 2796 * [no]dbdelay: Enable or disable the extra 200ms delay 2797 before debouncing a link PHY and device presence 2798 detection. 2799 2800 * [no]ncq: Turn on or off NCQ. 2801 2802 * [no]ncqtrim: Enable or disable queued DSM TRIM. 2803 2804 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 2805 2806 * [no]trim: Enable or disable (unqueued) TRIM. 2807 2808 * trim_zero: Indicate that TRIM command zeroes data. 2809 2810 * max_trim_128m: Set 128M maximum trim size limit. 2811 2812 * [no]dma: Turn on or off DMA transfers. 2813 2814 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 2815 2816 * atapi_mod16_dma: Enable the use of ATAPI DMA for 2817 commands that are not a multiple of 16 bytes. 2818 2819 * [no]dmalog: Enable or disable the use of the 2820 READ LOG DMA EXT command to access logs. 2821 2822 * [no]iddevlog: Enable or disable access to the 2823 identify device data log. 2824 2825 * [no]logdir: Enable or disable access to the general 2826 purpose log directory. 2827 2828 * max_sec_128: Set transfer size limit to 128 sectors. 2829 2830 * max_sec_1024: Set or clear transfer size limit to 2831 1024 sectors. 2832 2833 * max_sec_lba48: Set or clear transfer size limit to 2834 65535 sectors. 2835 2836 * [no]lpm: Enable or disable link power management. 2837 2838 * [no]setxfer: Indicate if transfer speed mode setting 2839 should be skipped. 2840 2841 * [no]fua: Disable or enable FUA (Force Unit Access) 2842 support for devices supporting this feature. 2843 2844 * dump_id: Dump IDENTIFY data. 2845 2846 * disable: Disable this device. 2847 2848 If there are multiple matching configurations changing 2849 the same attribute, the last one is used. 2850 2851 load_ramdisk= [RAM] [Deprecated] 2852 2853 lockd.nlm_grace_period=P [NFS] Assign grace period. 2854 Format: <integer> 2855 2856 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2857 Format: <integer> 2858 2859 lockd.nlm_timeout=T [NFS] Assign timeout value. 2860 Format: <integer> 2861 2862 lockd.nlm_udpport=M [NFS] Assign UDP port. 2863 Format: <integer> 2864 2865 lockdown= [SECURITY] 2866 { integrity | confidentiality } 2867 Enable the kernel lockdown feature. If set to 2868 integrity, kernel features that allow userland to 2869 modify the running kernel are disabled. If set to 2870 confidentiality, kernel features that allow userland 2871 to extract confidential information from the kernel 2872 are also disabled. 2873 2874 locktorture.nreaders_stress= [KNL] 2875 Set the number of locking read-acquisition kthreads. 2876 Defaults to being automatically set based on the 2877 number of online CPUs. 2878 2879 locktorture.nwriters_stress= [KNL] 2880 Set the number of locking write-acquisition kthreads. 2881 2882 locktorture.onoff_holdoff= [KNL] 2883 Set time (s) after boot for CPU-hotplug testing. 2884 2885 locktorture.onoff_interval= [KNL] 2886 Set time (s) between CPU-hotplug operations, or 2887 zero to disable CPU-hotplug testing. 2888 2889 locktorture.shuffle_interval= [KNL] 2890 Set task-shuffle interval (jiffies). Shuffling 2891 tasks allows some CPUs to go into dyntick-idle 2892 mode during the locktorture test. 2893 2894 locktorture.shutdown_secs= [KNL] 2895 Set time (s) after boot system shutdown. This 2896 is useful for hands-off automated testing. 2897 2898 locktorture.stat_interval= [KNL] 2899 Time (s) between statistics printk()s. 2900 2901 locktorture.stutter= [KNL] 2902 Time (s) to stutter testing, for example, 2903 specifying five seconds causes the test to run for 2904 five seconds, wait for five seconds, and so on. 2905 This tests the locking primitive's ability to 2906 transition abruptly to and from idle. 2907 2908 locktorture.torture_type= [KNL] 2909 Specify the locking implementation to test. 2910 2911 locktorture.verbose= [KNL] 2912 Enable additional printk() statements. 2913 2914 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2915 Format: <irq> 2916 2917 loglevel= All Kernel Messages with a loglevel smaller than the 2918 console loglevel will be printed to the console. It can 2919 also be changed with klogd or other programs. The 2920 loglevels are defined as follows: 2921 2922 0 (KERN_EMERG) system is unusable 2923 1 (KERN_ALERT) action must be taken immediately 2924 2 (KERN_CRIT) critical conditions 2925 3 (KERN_ERR) error conditions 2926 4 (KERN_WARNING) warning conditions 2927 5 (KERN_NOTICE) normal but significant condition 2928 6 (KERN_INFO) informational 2929 7 (KERN_DEBUG) debug-level messages 2930 2931 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2932 in bytes. n must be a power of two and greater 2933 than the minimal size. The minimal size is defined 2934 by LOG_BUF_SHIFT kernel config parameter. There is 2935 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2936 that allows to increase the default size depending on 2937 the number of CPUs. See init/Kconfig for more details. 2938 2939 logo.nologo [FB] Disables display of the built-in Linux logo. 2940 This may be used to provide more screen space for 2941 kernel log messages and is useful when debugging 2942 kernel boot problems. 2943 2944 lp=0 [LP] Specify parallel ports to use, e.g, 2945 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2946 lp=reset first parallel port). 'lp=0' disables the 2947 lp=auto printer driver. 'lp=reset' (which can be 2948 specified in addition to the ports) causes 2949 attached printers to be reset. Using 2950 lp=port1,port2,... specifies the parallel ports 2951 to associate lp devices with, starting with 2952 lp0. A port specification may be 'none' to skip 2953 that lp device, or a parport name such as 2954 'parport0'. Specifying 'lp=auto' instead of a 2955 port specification list means that device IDs 2956 from each port should be examined, to see if 2957 an IEEE 1284-compliant printer is attached; if 2958 so, the driver will manage that printer. 2959 See also header of drivers/char/lp.c. 2960 2961 lpj=n [KNL] 2962 Sets loops_per_jiffy to given constant, thus avoiding 2963 time-consuming boot-time autodetection (up to 250 ms per 2964 CPU). 0 enables autodetection (default). To determine 2965 the correct value for your kernel, boot with normal 2966 autodetection and see what value is printed. Note that 2967 on SMP systems the preset will be applied to all CPUs, 2968 which is likely to cause problems if your CPUs need 2969 significantly divergent settings. An incorrect value 2970 will cause delays in the kernel to be wrong, leading to 2971 unpredictable I/O errors and other breakage. Although 2972 unlikely, in the extreme case this might damage your 2973 hardware. 2974 2975 ltpc= [NET] 2976 Format: <io>,<irq>,<dma> 2977 2978 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2979 2980 lsm=lsm1,...,lsmN 2981 [SECURITY] Choose order of LSM initialization. This 2982 overrides CONFIG_LSM, and the "security=" parameter. 2983 2984 machvec= [IA-64] Force the use of a particular machine-vector 2985 (machvec) in a generic kernel. 2986 Example: machvec=hpzx1 2987 2988 machtype= [Loongson] Share the same kernel image file between 2989 different yeeloong laptops. 2990 Example: machtype=lemote-yeeloong-2f-7inch 2991 2992 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater 2993 than or equal to this physical address is ignored. 2994 2995 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2996 will bring up during bootup. maxcpus=n : n >= 0 limits 2997 the kernel to bring up 'n' processors. Surely after 2998 bootup you can bring up the other plugged cpu by executing 2999 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3000 only takes effect during system bootup. 3001 While n=0 is a special case, it is equivalent to "nosmp", 3002 which also disables the IO APIC. 3003 3004 max_loop= [LOOP] The number of loop block devices that get 3005 (loop.max_loop) unconditionally pre-created at init time. The default 3006 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3007 of statically allocating a predefined number, loop 3008 devices can be requested on-demand with the 3009 /dev/loop-control interface. 3010 3011 mce [X86-32] Machine Check Exception 3012 3013 mce=option [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 3014 3015 md= [HW] RAID subsystems devices and level 3016 See Documentation/admin-guide/md.rst. 3017 3018 mdacon= [MDA] 3019 Format: <first>,<last> 3020 Specifies range of consoles to be captured by the MDA. 3021 3022 mds= [X86,INTEL] 3023 Control mitigation for the Micro-architectural Data 3024 Sampling (MDS) vulnerability. 3025 3026 Certain CPUs are vulnerable to an exploit against CPU 3027 internal buffers which can forward information to a 3028 disclosure gadget under certain conditions. 3029 3030 In vulnerable processors, the speculatively 3031 forwarded data can be used in a cache side channel 3032 attack, to access data to which the attacker does 3033 not have direct access. 3034 3035 This parameter controls the MDS mitigation. The 3036 options are: 3037 3038 full - Enable MDS mitigation on vulnerable CPUs 3039 full,nosmt - Enable MDS mitigation and disable 3040 SMT on vulnerable CPUs 3041 off - Unconditionally disable MDS mitigation 3042 3043 On TAA-affected machines, mds=off can be prevented by 3044 an active TAA mitigation as both vulnerabilities are 3045 mitigated with the same mechanism so in order to disable 3046 this mitigation, you need to specify tsx_async_abort=off 3047 too. 3048 3049 Not specifying this option is equivalent to 3050 mds=full. 3051 3052 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3053 3054 mem=nn[KMG] [HEXAGON] Set the memory size. 3055 Must be specified, otherwise memory size will be 0. 3056 3057 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 3058 Amount of memory to be used in cases as follows: 3059 3060 1 for test; 3061 2 when the kernel is not able to see the whole system memory; 3062 3 memory that lies after 'mem=' boundary is excluded from 3063 the hypervisor, then assigned to KVM guests. 3064 4 to limit the memory available for kdump kernel. 3065 3066 [ARC,MICROBLAZE] - the limit applies only to low memory, 3067 high memory is not affected. 3068 3069 [ARM64] - only limits memory covered by the linear 3070 mapping. The NOMAP regions are not affected. 3071 3072 [X86] Work as limiting max address. Use together 3073 with memmap= to avoid physical address space collisions. 3074 Without memmap= PCI devices could be placed at addresses 3075 belonging to unused RAM. 3076 3077 Note that this only takes effects during boot time since 3078 in above case 3, memory may need be hot added after boot 3079 if system memory of hypervisor is not sufficient. 3080 3081 mem=nn[KMG]@ss[KMG] 3082 [ARM,MIPS] - override the memory layout reported by 3083 firmware. 3084 Define a memory region of size nn[KMG] starting at 3085 ss[KMG]. 3086 Multiple different regions can be specified with 3087 multiple mem= parameters on the command line. 3088 3089 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3090 memory. 3091 3092 memblock=debug [KNL] Enable memblock debug messages. 3093 3094 memchunk=nn[KMG] 3095 [KNL,SH] Allow user to override the default size for 3096 per-device physically contiguous DMA buffers. 3097 3098 memhp_default_state=online/offline 3099 [KNL] Set the initial state for the memory hotplug 3100 onlining policy. If not specified, the default value is 3101 set according to the 3102 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 3103 option. 3104 See Documentation/admin-guide/mm/memory-hotplug.rst. 3105 3106 memmap=exactmap [KNL,X86] Enable setting of an exact 3107 E820 memory map, as specified by the user. 3108 Such memmap=exactmap lines can be constructed based on 3109 BIOS output or other requirements. See the memmap=nn@ss 3110 option description. 3111 3112 memmap=nn[KMG]@ss[KMG] 3113 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory. 3114 Region of memory to be used is from ss to ss+nn. 3115 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3116 which limits max address to nn[KMG]. 3117 Multiple different regions can be specified, 3118 comma delimited. 3119 Example: 3120 memmap=100M@2G,100M#3G,1G!1024G 3121 3122 memmap=nn[KMG]#ss[KMG] 3123 [KNL,ACPI] Mark specific memory as ACPI data. 3124 Region of memory to be marked is from ss to ss+nn. 3125 3126 memmap=nn[KMG]$ss[KMG] 3127 [KNL,ACPI] Mark specific memory as reserved. 3128 Region of memory to be reserved is from ss to ss+nn. 3129 Example: Exclude memory from 0x18690000-0x1869ffff 3130 memmap=64K$0x18690000 3131 or 3132 memmap=0x10000$0x18690000 3133 Some bootloaders may need an escape character before '$', 3134 like Grub2, otherwise '$' and the following number 3135 will be eaten. 3136 3137 memmap=nn[KMG]!ss[KMG] 3138 [KNL,X86] Mark specific memory as protected. 3139 Region of memory to be used, from ss to ss+nn. 3140 The memory region may be marked as e820 type 12 (0xc) 3141 and is NVDIMM or ADR memory. 3142 3143 memmap=<size>%<offset>-<oldtype>+<newtype> 3144 [KNL,ACPI] Convert memory within the specified region 3145 from <oldtype> to <newtype>. If "-<oldtype>" is left 3146 out, the whole region will be marked as <newtype>, 3147 even if previously unavailable. If "+<newtype>" is left 3148 out, matching memory will be removed. Types are 3149 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3150 3 = ACPI, 12 = PRAM. 3151 3152 memory_corruption_check=0/1 [X86] 3153 Some BIOSes seem to corrupt the first 64k of 3154 memory when doing things like suspend/resume. 3155 Setting this option will scan the memory 3156 looking for corruption. Enabling this will 3157 both detect corruption and prevent the kernel 3158 from using the memory being corrupted. 3159 However, its intended as a diagnostic tool; if 3160 repeatable BIOS-originated corruption always 3161 affects the same memory, you can use memmap= 3162 to prevent the kernel from using that memory. 3163 3164 memory_corruption_check_size=size [X86] 3165 By default it checks for corruption in the low 3166 64k, making this memory unavailable for normal 3167 use. Use this parameter to scan for 3168 corruption in more or less memory. 3169 3170 memory_corruption_check_period=seconds [X86] 3171 By default it checks for corruption every 60 3172 seconds. Use this parameter to check at some 3173 other rate. 0 disables periodic checking. 3174 3175 memory_hotplug.memmap_on_memory 3176 [KNL,X86,ARM] Boolean flag to enable this feature. 3177 Format: {on | off (default)} 3178 When enabled, runtime hotplugged memory will 3179 allocate its internal metadata (struct pages, 3180 those vmemmap pages cannot be optimized even 3181 if hugetlb_free_vmemmap is enabled) from the 3182 hotadded memory which will allow to hotadd a 3183 lot of memory without requiring additional 3184 memory to do so. 3185 This feature is disabled by default because it 3186 has some implication on large (e.g. GB) 3187 allocations in some configurations (e.g. small 3188 memory blocks). 3189 The state of the flag can be read in 3190 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3191 Note that even when enabled, there are a few cases where 3192 the feature is not effective. 3193 3194 memtest= [KNL,X86,ARM,M68K,PPC,RISCV] Enable memtest 3195 Format: <integer> 3196 default : 0 <disable> 3197 Specifies the number of memtest passes to be 3198 performed. Each pass selects another test 3199 pattern from a given set of patterns. Memtest 3200 fills the memory with this pattern, validates 3201 memory contents and reserves bad memory 3202 regions that are detected. 3203 3204 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3205 Valid arguments: on, off 3206 Default (depends on kernel configuration option): 3207 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 3208 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 3209 mem_encrypt=on: Activate SME 3210 mem_encrypt=off: Do not activate SME 3211 3212 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3213 for details on when memory encryption can be activated. 3214 3215 mem_sleep_default= [SUSPEND] Default system suspend mode: 3216 s2idle - Suspend-To-Idle 3217 shallow - Power-On Suspend or equivalent (if supported) 3218 deep - Suspend-To-RAM or equivalent (if supported) 3219 See Documentation/admin-guide/pm/sleep-states.rst. 3220 3221 mfgpt_irq= [IA-32] Specify the IRQ to use for the 3222 Multi-Function General Purpose Timers on AMD Geode 3223 platforms. 3224 3225 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3226 the BIOS has incorrectly applied a workaround. TinyBIOS 3227 version 0.98 is known to be affected, 0.99 fixes the 3228 problem by letting the user disable the workaround. 3229 3230 mga= [HW,DRM] 3231 3232 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this 3233 physical address is ignored. 3234 3235 mini2440= [ARM,HW,KNL] 3236 Format:[0..2][b][c][t] 3237 Default: "0tb" 3238 MINI2440 configuration specification: 3239 0 - The attached screen is the 3.5" TFT 3240 1 - The attached screen is the 7" TFT 3241 2 - The VGA Shield is attached (1024x768) 3242 Leaving out the screen size parameter will not load 3243 the TFT driver, and the framebuffer will be left 3244 unconfigured. 3245 b - Enable backlight. The TFT backlight pin will be 3246 linked to the kernel VESA blanking code and a GPIO 3247 LED. This parameter is not necessary when using the 3248 VGA shield. 3249 c - Enable the s3c camera interface. 3250 t - Reserved for enabling touchscreen support. The 3251 touchscreen support is not enabled in the mainstream 3252 kernel as of 2.6.30, a preliminary port can be found 3253 in the "bleeding edge" mini2440 support kernel at 3254 https://repo.or.cz/w/linux-2.6/mini2440.git 3255 3256 mitigations= 3257 [X86,PPC,S390,ARM64] Control optional mitigations for 3258 CPU vulnerabilities. This is a set of curated, 3259 arch-independent options, each of which is an 3260 aggregation of existing arch-specific options. 3261 3262 off 3263 Disable all optional CPU mitigations. This 3264 improves system performance, but it may also 3265 expose users to several CPU vulnerabilities. 3266 Equivalent to: nopti [X86,PPC] 3267 if nokaslr then kpti=0 [ARM64] 3268 nospectre_v1 [X86,PPC] 3269 nobp=0 [S390] 3270 nospectre_v2 [X86,PPC,S390,ARM64] 3271 spectre_v2_user=off [X86] 3272 spec_store_bypass_disable=off [X86,PPC] 3273 ssbd=force-off [ARM64] 3274 nospectre_bhb [ARM64] 3275 l1tf=off [X86] 3276 mds=off [X86] 3277 tsx_async_abort=off [X86] 3278 kvm.nx_huge_pages=off [X86] 3279 srbds=off [X86,INTEL] 3280 no_entry_flush [PPC] 3281 no_uaccess_flush [PPC] 3282 mmio_stale_data=off [X86] 3283 retbleed=off [X86] 3284 3285 Exceptions: 3286 This does not have any effect on 3287 kvm.nx_huge_pages when 3288 kvm.nx_huge_pages=force. 3289 3290 auto (default) 3291 Mitigate all CPU vulnerabilities, but leave SMT 3292 enabled, even if it's vulnerable. This is for 3293 users who don't want to be surprised by SMT 3294 getting disabled across kernel upgrades, or who 3295 have other ways of avoiding SMT-based attacks. 3296 Equivalent to: (default behavior) 3297 3298 auto,nosmt 3299 Mitigate all CPU vulnerabilities, disabling SMT 3300 if needed. This is for users who always want to 3301 be fully mitigated, even if it means losing SMT. 3302 Equivalent to: l1tf=flush,nosmt [X86] 3303 mds=full,nosmt [X86] 3304 tsx_async_abort=full,nosmt [X86] 3305 mmio_stale_data=full,nosmt [X86] 3306 retbleed=auto,nosmt [X86] 3307 3308 mminit_loglevel= 3309 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 3310 parameter allows control of the logging verbosity for 3311 the additional memory initialisation checks. A value 3312 of 0 disables mminit logging and a level of 4 will 3313 log everything. Information is printed at KERN_DEBUG 3314 so loglevel=8 may also need to be specified. 3315 3316 mmio_stale_data= 3317 [X86,INTEL] Control mitigation for the Processor 3318 MMIO Stale Data vulnerabilities. 3319 3320 Processor MMIO Stale Data is a class of 3321 vulnerabilities that may expose data after an MMIO 3322 operation. Exposed data could originate or end in 3323 the same CPU buffers as affected by MDS and TAA. 3324 Therefore, similar to MDS and TAA, the mitigation 3325 is to clear the affected CPU buffers. 3326 3327 This parameter controls the mitigation. The 3328 options are: 3329 3330 full - Enable mitigation on vulnerable CPUs 3331 3332 full,nosmt - Enable mitigation and disable SMT on 3333 vulnerable CPUs. 3334 3335 off - Unconditionally disable mitigation 3336 3337 On MDS or TAA affected machines, 3338 mmio_stale_data=off can be prevented by an active 3339 MDS or TAA mitigation as these vulnerabilities are 3340 mitigated with the same mechanism so in order to 3341 disable this mitigation, you need to specify 3342 mds=off and tsx_async_abort=off too. 3343 3344 Not specifying this option is equivalent to 3345 mmio_stale_data=full. 3346 3347 For details see: 3348 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3349 3350 <module>.async_probe[=<bool>] [KNL] 3351 If no <bool> value is specified or if the value 3352 specified is not a valid <bool>, enable asynchronous 3353 probe on this module. Otherwise, enable/disable 3354 asynchronous probe on this module as indicated by the 3355 <bool> value. See also: module.async_probe 3356 3357 module.async_probe=<bool> 3358 [KNL] When set to true, modules will use async probing 3359 by default. To enable/disable async probing for a 3360 specific module, use the module specific control that 3361 is documented under <module>.async_probe. When both 3362 module.async_probe and <module>.async_probe are 3363 specified, <module>.async_probe takes precedence for 3364 the specific module. 3365 3366 module.enable_dups_trace 3367 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 3368 this means that duplicate request_module() calls will 3369 trigger a WARN_ON() instead of a pr_warn(). Note that 3370 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 3371 will always be issued and this option does nothing. 3372 module.sig_enforce 3373 [KNL] When CONFIG_MODULE_SIG is set, this means that 3374 modules without (valid) signatures will fail to load. 3375 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3376 is always true, so this option does nothing. 3377 3378 module_blacklist= [KNL] Do not load a comma-separated list of 3379 modules. Useful for debugging problem modules. 3380 3381 mousedev.tap_time= 3382 [MOUSE] Maximum time between finger touching and 3383 leaving touchpad surface for touch to be considered 3384 a tap and be reported as a left button click (for 3385 touchpads working in absolute mode only). 3386 Format: <msecs> 3387 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 3388 reporting absolute coordinates, such as tablets 3389 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 3390 reporting absolute coordinates, such as tablets 3391 3392 movablecore= [KNL,X86,IA-64,PPC] 3393 Format: nn[KMGTPE] | nn% 3394 This parameter is the complement to kernelcore=, it 3395 specifies the amount of memory used for migratable 3396 allocations. If both kernelcore and movablecore is 3397 specified, then kernelcore will be at *least* the 3398 specified value but may be more. If movablecore on its 3399 own is specified, the administrator must be careful 3400 that the amount of memory usable for all allocations 3401 is not too small. 3402 3403 movable_node [KNL] Boot-time switch to make hotplugable memory 3404 NUMA nodes to be movable. This means that the memory 3405 of such nodes will be usable only for movable 3406 allocations which rules out almost all kernel 3407 allocations. Use with caution! 3408 3409 MTD_Partition= [MTD] 3410 Format: <name>,<region-number>,<size>,<offset> 3411 3412 MTD_Region= [MTD] Format: 3413 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 3414 3415 mtdparts= [MTD] 3416 See drivers/mtd/parsers/cmdlinepart.c 3417 3418 mtdset= [ARM] 3419 ARM/S3C2412 JIVE boot control 3420 3421 See arch/arm/mach-s3c/mach-jive.c 3422 3423 mtouchusb.raw_coordinates= 3424 [HW] Make the MicroTouch USB driver use raw coordinates 3425 ('y', default) or cooked coordinates ('n') 3426 3427 mtrr_chunk_size=nn[KMG] [X86] 3428 used for mtrr cleanup. It is largest continuous chunk 3429 that could hold holes aka. UC entries. 3430 3431 mtrr_gran_size=nn[KMG] [X86] 3432 Used for mtrr cleanup. It is granularity of mtrr block. 3433 Default is 1. 3434 Large value could prevent small alignment from 3435 using up MTRRs. 3436 3437 mtrr_spare_reg_nr=n [X86] 3438 Format: <integer> 3439 Range: 0,7 : spare reg number 3440 Default : 1 3441 Used for mtrr cleanup. It is spare mtrr entries number. 3442 Set to 2 or more if your graphical card needs more. 3443 3444 multitce=off [PPC] This parameter disables the use of the pSeries 3445 firmware feature for updating multiple TCE entries 3446 at a time. 3447 3448 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 3449 3450 netdev= [NET] Network devices parameters 3451 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 3452 Note that mem_start is often overloaded to mean 3453 something different and driver-specific. 3454 This usage is only documented in each driver source 3455 file if at all. 3456 3457 netpoll.carrier_timeout= 3458 [NET] Specifies amount of time (in seconds) that 3459 netpoll should wait for a carrier. By default netpoll 3460 waits 4 seconds. 3461 3462 nf_conntrack.acct= 3463 [NETFILTER] Enable connection tracking flow accounting 3464 0 to disable accounting 3465 1 to enable accounting 3466 Default value is 0. 3467 3468 nfs.cache_getent= 3469 [NFS] sets the pathname to the program which is used 3470 to update the NFS client cache entries. 3471 3472 nfs.cache_getent_timeout= 3473 [NFS] sets the timeout after which an attempt to 3474 update a cache entry is deemed to have failed. 3475 3476 nfs.callback_nr_threads= 3477 [NFSv4] set the total number of threads that the 3478 NFS client will assign to service NFSv4 callback 3479 requests. 3480 3481 nfs.callback_tcpport= 3482 [NFS] set the TCP port on which the NFSv4 callback 3483 channel should listen. 3484 3485 nfs.enable_ino64= 3486 [NFS] enable 64-bit inode numbers. 3487 If zero, the NFS client will fake up a 32-bit inode 3488 number for the readdir() and stat() syscalls instead 3489 of returning the full 64-bit number. 3490 The default is to return 64-bit inode numbers. 3491 3492 nfs.idmap_cache_timeout= 3493 [NFS] set the maximum lifetime for idmapper cache 3494 entries. 3495 3496 nfs.max_session_cb_slots= 3497 [NFSv4.1] Sets the maximum number of session 3498 slots the client will assign to the callback 3499 channel. This determines the maximum number of 3500 callbacks the client will process in parallel for 3501 a particular server. 3502 3503 nfs.max_session_slots= 3504 [NFSv4.1] Sets the maximum number of session slots 3505 the client will attempt to negotiate with the server. 3506 This limits the number of simultaneous RPC requests 3507 that the client can send to the NFSv4.1 server. 3508 Note that there is little point in setting this 3509 value higher than the max_tcp_slot_table_limit. 3510 3511 nfs.nfs4_disable_idmapping= 3512 [NFSv4] When set to the default of '1', this option 3513 ensures that both the RPC level authentication 3514 scheme and the NFS level operations agree to use 3515 numeric uids/gids if the mount is using the 3516 'sec=sys' security flavour. In effect it is 3517 disabling idmapping, which can make migration from 3518 legacy NFSv2/v3 systems to NFSv4 easier. 3519 Servers that do not support this mode of operation 3520 will be autodetected by the client, and it will fall 3521 back to using the idmapper. 3522 To turn off this behaviour, set the value to '0'. 3523 3524 nfs.nfs4_unique_id= 3525 [NFS4] Specify an additional fixed unique ident- 3526 ification string that NFSv4 clients can insert into 3527 their nfs_client_id4 string. This is typically a 3528 UUID that is generated at system install time. 3529 3530 nfs.recover_lost_locks= 3531 [NFSv4] Attempt to recover locks that were lost due 3532 to a lease timeout on the server. Please note that 3533 doing this risks data corruption, since there are 3534 no guarantees that the file will remain unchanged 3535 after the locks are lost. 3536 If you want to enable the kernel legacy behaviour of 3537 attempting to recover these locks, then set this 3538 parameter to '1'. 3539 The default parameter value of '0' causes the kernel 3540 not to attempt recovery of lost locks. 3541 3542 nfs.send_implementation_id= 3543 [NFSv4.1] Send client implementation identification 3544 information in exchange_id requests. 3545 If zero, no implementation identification information 3546 will be sent. 3547 The default is to send the implementation identification 3548 information. 3549 3550 nfs4.layoutstats_timer= 3551 [NFSv4.2] Change the rate at which the kernel sends 3552 layoutstats to the pNFS metadata server. 3553 3554 Setting this to value to 0 causes the kernel to use 3555 whatever value is the default set by the layout 3556 driver. A non-zero value sets the minimum interval 3557 in seconds between layoutstats transmissions. 3558 3559 nfsd.inter_copy_offload_enable= 3560 [NFSv4.2] When set to 1, the server will support 3561 server-to-server copies for which this server is 3562 the destination of the copy. 3563 3564 nfsd.nfs4_disable_idmapping= 3565 [NFSv4] When set to the default of '1', the NFSv4 3566 server will return only numeric uids and gids to 3567 clients using auth_sys, and will accept numeric uids 3568 and gids from such clients. This is intended to ease 3569 migration from NFSv2/v3. 3570 3571 nfsd.nfsd4_ssc_umount_timeout= 3572 [NFSv4.2] When used as the destination of a 3573 server-to-server copy, knfsd temporarily mounts 3574 the source server. It caches the mount in case 3575 it will be needed again, and discards it if not 3576 used for the number of milliseconds specified by 3577 this parameter. 3578 3579 nfsaddrs= [NFS] Deprecated. Use ip= instead. 3580 See Documentation/admin-guide/nfs/nfsroot.rst. 3581 3582 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 3583 See Documentation/admin-guide/nfs/nfsroot.rst. 3584 3585 nfsrootdebug [NFS] enable nfsroot debugging messages. 3586 See Documentation/admin-guide/nfs/nfsroot.rst. 3587 3588 nmi_backtrace.backtrace_idle [KNL] 3589 Dump stacks even of idle CPUs in response to an 3590 NMI stack-backtrace request. 3591 3592 nmi_debug= [KNL,SH] Specify one or more actions to take 3593 when a NMI is triggered. 3594 Format: [state][,regs][,debounce][,die] 3595 3596 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3597 Format: [panic,][nopanic,][num] 3598 Valid num: 0 or 1 3599 0 - turn hardlockup detector in nmi_watchdog off 3600 1 - turn hardlockup detector in nmi_watchdog on 3601 When panic is specified, panic when an NMI watchdog 3602 timeout occurs (or 'nopanic' to not panic on an NMI 3603 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3604 To disable both hard and soft lockup detectors, 3605 please see 'nowatchdog'. 3606 This is useful when you use a panic=... timeout and 3607 need the box quickly up again. 3608 3609 These settings can be accessed at runtime via 3610 the nmi_watchdog and hardlockup_panic sysctls. 3611 3612 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3613 emulation library even if a 387 maths coprocessor 3614 is present. 3615 3616 no5lvl [X86-64] Disable 5-level paging mode. Forces 3617 kernel to use 4-level paging instead. 3618 3619 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3620 caches in the slab allocator. Saves per-node memory, 3621 but will impact performance. 3622 3623 noalign [KNL,ARM] 3624 3625 noaltinstr [S390] Disables alternative instructions patching 3626 (CPU alternatives feature). 3627 3628 noapic [SMP,APIC] Tells the kernel to not make use of any 3629 IOAPICs that may be present in the system. 3630 3631 noautogroup Disable scheduler automatic task group creation. 3632 3633 nocache [ARM] 3634 3635 no_console_suspend 3636 [HW] Never suspend the console 3637 Disable suspending of consoles during suspend and 3638 hibernate operations. Once disabled, debugging 3639 messages can reach various consoles while the rest 3640 of the system is being put to sleep (ie, while 3641 debugging driver suspend/resume hooks). This may 3642 not work reliably with all consoles, but is known 3643 to work with serial and VGA consoles. 3644 To facilitate more flexible debugging, we also add 3645 console_suspend, a printk module parameter to control 3646 it. Users could use console_suspend (usually 3647 /sys/module/printk/parameters/console_suspend) to 3648 turn on/off it dynamically. 3649 3650 no_debug_objects 3651 [KNL] Disable object debugging 3652 3653 nodsp [SH] Disable hardware DSP at boot time. 3654 3655 noefi Disable EFI runtime services support. 3656 3657 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel. 3658 3659 noexec [IA-64] 3660 3661 noexec32 [X86-64] 3662 This affects only 32-bit executables. 3663 noexec32=on: enable non-executable mappings (default) 3664 read doesn't imply executable mappings 3665 noexec32=off: disable non-executable mappings 3666 read implies executable mappings 3667 3668 no_file_caps Tells the kernel not to honor file capabilities. The 3669 only way then for a file to be executed with privilege 3670 is to be setuid root or executed by root. 3671 3672 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3673 3674 nofsgsbase [X86] Disables FSGSBASE instructions. 3675 3676 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3677 register save and restore. The kernel will only save 3678 legacy floating-point registers on task switch. 3679 3680 nohalt [IA-64] Tells the kernel not to use the power saving 3681 function PAL_HALT_LIGHT when idle. This increases 3682 power-consumption. On the positive side, it reduces 3683 interrupt wake-up latency, which may improve performance 3684 in certain environments such as networked servers or 3685 real-time systems. 3686 3687 no_hash_pointers 3688 Force pointers printed to the console or buffers to be 3689 unhashed. By default, when a pointer is printed via %p 3690 format string, that pointer is "hashed", i.e. obscured 3691 by hashing the pointer value. This is a security feature 3692 that hides actual kernel addresses from unprivileged 3693 users, but it also makes debugging the kernel more 3694 difficult since unequal pointers can no longer be 3695 compared. However, if this command-line option is 3696 specified, then all normal pointers will have their true 3697 value printed. This option should only be specified when 3698 debugging the kernel. Please do not use on production 3699 kernels. 3700 3701 nohibernate [HIBERNATION] Disable hibernation and resume. 3702 3703 nohlt [ARM,ARM64,MICROBLAZE,SH] Forces the kernel to busy wait 3704 in do_idle() and not use the arch_cpu_idle() 3705 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 3706 to be effective. This is useful on platforms where the 3707 sleep(SH) or wfi(ARM,ARM64) instructions do not work 3708 correctly or when doing power measurements to evaluate 3709 the impact of the sleep instructions. This is also 3710 useful when using JTAG debugger. 3711 3712 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings. 3713 3714 nohugevmalloc [KNL,X86,PPC,ARM64] Disable kernel huge vmalloc mappings. 3715 3716 nohz= [KNL] Boottime enable/disable dynamic ticks 3717 Valid arguments: on, off 3718 Default: on 3719 3720 nohz_full= [KNL,BOOT,SMP,ISOL] 3721 The argument is a cpu list, as described above. 3722 In kernels built with CONFIG_NO_HZ_FULL=y, set 3723 the specified list of CPUs whose tick will be stopped 3724 whenever possible. The boot CPU will be forced outside 3725 the range to maintain the timekeeping. Any CPUs 3726 in this list will have their RCU callbacks offloaded, 3727 just as if they had also been called out in the 3728 rcu_nocbs= boot parameter. 3729 3730 Note that this argument takes precedence over 3731 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 3732 3733 noinitrd [RAM] Tells the kernel not to load any configured 3734 initial RAM disk. 3735 3736 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3737 remapping. 3738 [Deprecated - use intremap=off] 3739 3740 nointroute [IA-64] 3741 3742 noinvpcid [X86] Disable the INVPCID cpu feature. 3743 3744 noiotrap [SH] Disables trapped I/O port accesses. 3745 3746 noirqdebug [X86-32] Disables the code which attempts to detect and 3747 disable unhandled interrupt sources. 3748 3749 noisapnp [ISAPNP] Disables ISA PnP code. 3750 3751 nojitter [IA-64] Disables jitter checking for ITC timers. 3752 3753 nokaslr [KNL] 3754 When CONFIG_RANDOMIZE_BASE is set, this disables 3755 kernel and module base offset ASLR (Address Space 3756 Layout Randomization). 3757 3758 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3759 fault handling. 3760 3761 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3762 3763 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3764 3765 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3766 3767 nomca [IA-64] Disable machine check abort handling 3768 3769 nomce [X86-32] Disable Machine Check Exception 3770 3771 nomfgpt [X86-32] Disable Multi-Function General Purpose 3772 Timer usage (for AMD Geode machines). 3773 3774 nomodeset Disable kernel modesetting. Most systems' firmware 3775 sets up a display mode and provides framebuffer memory 3776 for output. With nomodeset, DRM and fbdev drivers will 3777 not load if they could possibly displace the pre- 3778 initialized output. Only the system framebuffer will 3779 be available for use. The respective drivers will not 3780 perform display-mode changes or accelerated rendering. 3781 3782 Useful as error fallback, or for testing and debugging. 3783 3784 nomodule Disable module load 3785 3786 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3787 shutdown the other cpus. Instead use the REBOOT_VECTOR 3788 irq. 3789 3790 nopat [X86] Disable PAT (page attribute table extension of 3791 pagetables) support. 3792 3793 nopcid [X86-64] Disable the PCID cpu feature. 3794 3795 nopku [X86] Disable Memory Protection Keys CPU feature found 3796 in some Intel CPUs. 3797 3798 nopti [X86-64] 3799 Equivalent to pti=off 3800 3801 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 3802 Disables the PV optimizations forcing the guest to run 3803 as generic guest with no PV drivers. Currently support 3804 XEN HVM, KVM, HYPER_V and VMWARE guest. 3805 3806 nopvspin [X86,XEN,KVM] 3807 Disables the qspinlock slow path using PV optimizations 3808 which allow the hypervisor to 'idle' the guest on lock 3809 contention. 3810 3811 norandmaps Don't use address space randomization. Equivalent to 3812 echo 0 > /proc/sys/kernel/randomize_va_space 3813 3814 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3815 with UP alternatives 3816 3817 noresume [SWSUSP] Disables resume and restores original swap 3818 space. 3819 3820 nosbagart [IA-64] 3821 3822 no-scroll [VGA] Disables scrollback. 3823 This is required for the Braillex ib80-piezo Braille 3824 reader made by F.H. Papenmeier (Germany). 3825 3826 nosgx [X86-64,SGX] Disables Intel SGX kernel support. 3827 3828 nosmap [PPC] 3829 Disable SMAP (Supervisor Mode Access Prevention) 3830 even if it is supported by processor. 3831 3832 nosmep [PPC64s] 3833 Disable SMEP (Supervisor Mode Execution Prevention) 3834 even if it is supported by processor. 3835 3836 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3837 and disable the IO APIC. legacy for "maxcpus=0". 3838 3839 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3840 Equivalent to smt=1. 3841 3842 [KNL,X86] Disable symmetric multithreading (SMT). 3843 nosmt=force: Force disable SMT, cannot be undone 3844 via the sysfs control file. 3845 3846 nosoftlockup [KNL] Disable the soft-lockup detector. 3847 3848 nospec_store_bypass_disable 3849 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3850 3851 nospectre_bhb [ARM64] Disable all mitigations for Spectre-BHB (branch 3852 history injection) vulnerability. System may allow data leaks 3853 with this option. 3854 3855 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3856 (bounds check bypass). With this option data leaks are 3857 possible in the system. 3858 3859 nospectre_v2 [X86,PPC_E500,ARM64] Disable all mitigations for 3860 the Spectre variant 2 (indirect branch prediction) 3861 vulnerability. System may allow data leaks with this 3862 option. 3863 3864 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES] Disable paravirtualized 3865 steal time accounting. steal time is computed, but 3866 won't influence scheduler behaviour 3867 3868 nosync [HW,M68K] Disables sync negotiation for all devices. 3869 3870 no_timer_check [X86,APIC] Disables the code which tests for 3871 broken timer IRQ sources. 3872 3873 no_uaccess_flush 3874 [PPC] Don't flush the L1-D cache after accessing user data. 3875 3876 novmcoredd [KNL,KDUMP] 3877 Disable device dump. Device dump allows drivers to 3878 append dump data to vmcore so you can collect driver 3879 specified debug info. Drivers can append the data 3880 without any limit and this data is stored in memory, 3881 so this may cause significant memory stress. Disabling 3882 device dump can help save memory but the driver debug 3883 data will be no longer available. This parameter 3884 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3885 is set. 3886 3887 no-vmw-sched-clock 3888 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3889 clock and use the default one. 3890 3891 nowatchdog [KNL] Disable both lockup detectors, i.e. 3892 soft-lockup and NMI watchdog (hard-lockup). 3893 3894 nowb [ARM] 3895 3896 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3897 3898 NOTE: this parameter will be ignored on systems with the 3899 LEGACY_XAPIC_DISABLED bit set in the 3900 IA32_XAPIC_DISABLE_STATUS MSR. 3901 3902 noxsave [BUGS=X86] Disables x86 extended register state save 3903 and restore using xsave. The kernel will fallback to 3904 enabling legacy floating-point and sse state. 3905 3906 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3907 register states. The kernel will fall back to use 3908 xsave to save the states. By using this parameter, 3909 performance of saving the states is degraded because 3910 xsave doesn't support modified optimization while 3911 xsaveopt supports it on xsaveopt enabled systems. 3912 3913 noxsaves [X86] Disables xsaves and xrstors used in saving and 3914 restoring x86 extended register state in compacted 3915 form of xsave area. The kernel will fall back to use 3916 xsaveopt and xrstor to save and restore the states 3917 in standard form of xsave area. By using this 3918 parameter, xsave area per process might occupy more 3919 memory on xsaves enabled systems. 3920 3921 nps_mtm_hs_ctr= [KNL,ARC] 3922 This parameter sets the maximum duration, in 3923 cycles, each HW thread of the CTOP can run 3924 without interruptions, before HW switches it. 3925 The actual maximum duration is 16 times this 3926 parameter's value. 3927 Format: integer between 1 and 255 3928 Default: 255 3929 3930 nptcg= [IA-64] Override max number of concurrent global TLB 3931 purges which is reported from either PAL_VM_SUMMARY or 3932 SAL PALO. 3933 3934 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3935 could support. nr_cpus=n : n >= 1 limits the kernel to 3936 support 'n' processors. It could be larger than the 3937 number of already plugged CPU during bootup, later in 3938 runtime you can physically add extra cpu until it reaches 3939 n. So during boot up some boot time memory for per-cpu 3940 variables need be pre-allocated for later physical cpu 3941 hot plugging. 3942 3943 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3944 3945 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only 3946 set up a single NUMA node spanning all memory. 3947 3948 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 3949 NUMA balancing. 3950 Allowed values are enable and disable 3951 3952 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3953 'node', 'default' can be specified 3954 This can be set from sysctl after boot. 3955 See Documentation/admin-guide/sysctl/vm.rst for details. 3956 3957 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3958 See Documentation/core-api/debugging-via-ohci1394.rst for more 3959 info. 3960 3961 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3962 Rather than timing out after 20 ms if an EC 3963 command is not properly ACKed, override the length 3964 of the timeout. We have interrupts disabled while 3965 waiting for the ACK, so if this is set too high 3966 interrupts *may* be lost! 3967 3968 omap_mux= [OMAP] Override bootloader pin multiplexing. 3969 Format: <mux_mode0.mode_name=value>... 3970 For example, to override I2C bus2: 3971 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3972 3973 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 3974 3975 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 3976 3977 boundary - index of last SLC block on Flex-OneNAND. 3978 The remaining blocks are configured as MLC blocks. 3979 lock - Configure if Flex-OneNAND boundary should be locked. 3980 Once locked, the boundary cannot be changed. 3981 1 indicates lock status, 0 indicates unlock status. 3982 3983 oops=panic Always panic on oopses. Default is to just kill the 3984 process, but there is a small probability of 3985 deadlocking the machine. 3986 This will also cause panics on machine check exceptions. 3987 Useful together with panic=30 to trigger a reboot. 3988 3989 page_alloc.shuffle= 3990 [KNL] Boolean flag to control whether the page allocator 3991 should randomize its free lists. The randomization may 3992 be automatically enabled if the kernel detects it is 3993 running on a platform with a direct-mapped memory-side 3994 cache, and this parameter can be used to 3995 override/disable that behavior. The state of the flag 3996 can be read from sysfs at: 3997 /sys/module/page_alloc/parameters/shuffle. 3998 3999 page_owner= [KNL] Boot-time page_owner enabling option. 4000 Storage of the information about who allocated 4001 each page is disabled in default. With this switch, 4002 we can turn it on. 4003 on: enable the feature 4004 4005 page_poison= [KNL] Boot-time parameter changing the state of 4006 poisoning on the buddy allocator, available with 4007 CONFIG_PAGE_POISONING=y. 4008 off: turn off poisoning (default) 4009 on: turn on poisoning 4010 4011 page_reporting.page_reporting_order= 4012 [KNL] Minimal page reporting order 4013 Format: <integer> 4014 Adjust the minimal page reporting order. The page 4015 reporting is disabled when it exceeds MAX_ORDER. 4016 4017 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4018 timeout > 0: seconds before rebooting 4019 timeout = 0: wait forever 4020 timeout < 0: reboot immediately 4021 Format: <timeout> 4022 4023 panic_print= Bitmask for printing system info when panic happens. 4024 User can chose combination of the following bits: 4025 bit 0: print all tasks info 4026 bit 1: print system memory info 4027 bit 2: print timer info 4028 bit 3: print locks info if CONFIG_LOCKDEP is on 4029 bit 4: print ftrace buffer 4030 bit 5: print all printk messages in buffer 4031 bit 6: print all CPUs backtrace (if available in the arch) 4032 *Be aware* that this option may print a _lot_ of lines, 4033 so there are risks of losing older messages in the log. 4034 Use this option carefully, maybe worth to setup a 4035 bigger log buffer with "log_buf_len" along with this. 4036 4037 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 4038 Format: <hex>[,nousertaint] 4039 Hexadecimal bitmask representing the set of TAINT flags 4040 that will cause the kernel to panic when add_taint() is 4041 called with any of the flags in this set. 4042 The optional switch "nousertaint" can be utilized to 4043 prevent userspace forced crashes by writing to sysctl 4044 /proc/sys/kernel/tainted any flagset matching with the 4045 bitmask set on panic_on_taint. 4046 See Documentation/admin-guide/tainted-kernels.rst for 4047 extra details on the taint flags that users can pick 4048 to compose the bitmask to assign to panic_on_taint. 4049 4050 panic_on_warn panic() instead of WARN(). Useful to cause kdump 4051 on a WARN(). 4052 4053 parkbd.port= [HW] Parallel port number the keyboard adapter is 4054 connected to, default is 0. 4055 Format: <parport#> 4056 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4057 0 for XT, 1 for AT (default is AT). 4058 Format: <mode> 4059 4060 parport= [HW,PPT] Specify parallel ports. 0 disables. 4061 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4062 Use 'auto' to force the driver to use any 4063 IRQ/DMA settings detected (the default is to 4064 ignore detected IRQ/DMA settings because of 4065 possible conflicts). You can specify the base 4066 address, IRQ, and DMA settings; IRQ and DMA 4067 should be numbers, or 'auto' (for using detected 4068 settings on that particular port), or 'nofifo' 4069 (to avoid using a FIFO even if it is detected). 4070 Parallel ports are assigned in the order they 4071 are specified on the command line, starting 4072 with parport0. 4073 4074 parport_init_mode= [HW,PPT] 4075 Configure VIA parallel port to operate in 4076 a specific mode. This is necessary on Pegasos 4077 computer where firmware has no options for setting 4078 up parallel port mode and sets it to spp. 4079 Currently this function knows 686a and 8231 chips. 4080 Format: [spp|ps2|epp|ecp|ecpepp] 4081 4082 pata_legacy.all= [HW,LIBATA] 4083 Format: <int> 4084 Set to non-zero to probe primary and secondary ISA 4085 port ranges on PCI systems where no PCI PATA device 4086 has been found at either range. Disabled by default. 4087 4088 pata_legacy.autospeed= [HW,LIBATA] 4089 Format: <int> 4090 Set to non-zero if a chip is present that snoops speed 4091 changes. Disabled by default. 4092 4093 pata_legacy.ht6560a= [HW,LIBATA] 4094 Format: <int> 4095 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4096 the secondary channel, or both channels respectively. 4097 Disabled by default. 4098 4099 pata_legacy.ht6560b= [HW,LIBATA] 4100 Format: <int> 4101 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4102 the secondary channel, or both channels respectively. 4103 Disabled by default. 4104 4105 pata_legacy.iordy_mask= [HW,LIBATA] 4106 Format: <int> 4107 IORDY enable mask. Set individual bits to allow IORDY 4108 for the respective channel. Bit 0 is for the first 4109 legacy channel handled by this driver, bit 1 is for 4110 the second channel, and so on. The sequence will often 4111 correspond to the primary legacy channel, the secondary 4112 legacy channel, and so on, but the handling of a PCI 4113 bus and the use of other driver options may interfere 4114 with the sequence. By default IORDY is allowed across 4115 all channels. 4116 4117 pata_legacy.opti82c46x= [HW,LIBATA] 4118 Format: <int> 4119 Set to 1, 2, or 3 for Opti 82c611A on the primary 4120 channel, the secondary channel, or both channels 4121 respectively. Disabled by default. 4122 4123 pata_legacy.opti82c611a= [HW,LIBATA] 4124 Format: <int> 4125 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4126 channel, the secondary channel, or both channels 4127 respectively. Disabled by default. 4128 4129 pata_legacy.pio_mask= [HW,LIBATA] 4130 Format: <int> 4131 PIO mode mask for autospeed devices. Set individual 4132 bits to allow the use of the respective PIO modes. 4133 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4134 All modes allowed by default. 4135 4136 pata_legacy.probe_all= [HW,LIBATA] 4137 Format: <int> 4138 Set to non-zero to probe tertiary and further ISA 4139 port ranges on PCI systems. Disabled by default. 4140 4141 pata_legacy.probe_mask= [HW,LIBATA] 4142 Format: <int> 4143 Probe mask for legacy ISA PATA ports. Depending on 4144 platform configuration and the use of other driver 4145 options up to 6 legacy ports are supported: 0x1f0, 4146 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4147 of individual ports can be disabled by setting the 4148 corresponding bits in the mask to 1. Bit 0 is for 4149 the first port in the list above (0x1f0), and so on. 4150 By default all supported ports are probed. 4151 4152 pata_legacy.qdi= [HW,LIBATA] 4153 Format: <int> 4154 Set to non-zero to probe QDI controllers. By default 4155 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4156 4157 pata_legacy.winbond= [HW,LIBATA] 4158 Format: <int> 4159 Set to non-zero to probe Winbond controllers. Use 4160 the standard I/O port (0x130) if 1, otherwise the 4161 value given is the I/O port to use (typically 0x1b0). 4162 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4163 0 otherwise. 4164 4165 pata_platform.pio_mask= [HW,LIBATA] 4166 Format: <int> 4167 Supported PIO mode mask. Set individual bits to allow 4168 the use of the respective PIO modes. Bit 0 is for 4169 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4170 allowed by default. 4171 4172 pause_on_oops= 4173 Halt all CPUs after the first oops has been printed for 4174 the specified number of seconds. This is to be used if 4175 your oopses keep scrolling off the screen. 4176 4177 pcbit= [HW,ISDN] 4178 4179 pci=option[,option...] [PCI] various PCI subsystem options. 4180 4181 Some options herein operate on a specific device 4182 or a set of devices (<pci_dev>). These are 4183 specified in one of the following formats: 4184 4185 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4186 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4187 4188 Note: the first format specifies a PCI 4189 bus/device/function address which may change 4190 if new hardware is inserted, if motherboard 4191 firmware changes, or due to changes caused 4192 by other kernel parameters. If the 4193 domain is left unspecified, it is 4194 taken to be zero. Optionally, a path 4195 to a device through multiple device/function 4196 addresses can be specified after the base 4197 address (this is more robust against 4198 renumbering issues). The second format 4199 selects devices using IDs from the 4200 configuration space which may match multiple 4201 devices in the system. 4202 4203 earlydump dump PCI config space before the kernel 4204 changes anything 4205 off [X86] don't probe for the PCI bus 4206 bios [X86-32] force use of PCI BIOS, don't access 4207 the hardware directly. Use this if your machine 4208 has a non-standard PCI host bridge. 4209 nobios [X86-32] disallow use of PCI BIOS, only direct 4210 hardware access methods are allowed. Use this 4211 if you experience crashes upon bootup and you 4212 suspect they are caused by the BIOS. 4213 conf1 [X86] Force use of PCI Configuration Access 4214 Mechanism 1 (config address in IO port 0xCF8, 4215 data in IO port 0xCFC, both 32-bit). 4216 conf2 [X86] Force use of PCI Configuration Access 4217 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4218 the function, IO port 0xCFA, also 8-bit, sets 4219 bus number. The config space is then accessed 4220 through ports 0xC000-0xCFFF). 4221 See http://wiki.osdev.org/PCI for more info 4222 on the configuration access mechanisms. 4223 noaer [PCIE] If the PCIEAER kernel config parameter is 4224 enabled, this kernel boot option can be used to 4225 disable the use of PCIE advanced error reporting. 4226 nodomains [PCI] Disable support for multiple PCI 4227 root domains (aka PCI segments, in ACPI-speak). 4228 nommconf [X86] Disable use of MMCONFIG for PCI 4229 Configuration 4230 check_enable_amd_mmconf [X86] check for and enable 4231 properly configured MMIO access to PCI 4232 config space on AMD family 10h CPU 4233 nomsi [MSI] If the PCI_MSI kernel config parameter is 4234 enabled, this kernel boot option can be used to 4235 disable the use of MSI interrupts system-wide. 4236 noioapicquirk [APIC] Disable all boot interrupt quirks. 4237 Safety option to keep boot IRQs enabled. This 4238 should never be necessary. 4239 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4240 primary IO-APIC for bridges that cannot disable 4241 boot IRQs. This fixes a source of spurious IRQs 4242 when the system masks IRQs. 4243 noioapicreroute [APIC] Disable workaround that uses the 4244 boot IRQ equivalent of an IRQ that connects to 4245 a chipset where boot IRQs cannot be disabled. 4246 The opposite of ioapicreroute. 4247 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4248 routing table. These calls are known to be buggy 4249 on several machines and they hang the machine 4250 when used, but on other computers it's the only 4251 way to get the interrupt routing table. Try 4252 this option if the kernel is unable to allocate 4253 IRQs or discover secondary PCI buses on your 4254 motherboard. 4255 rom [X86] Assign address space to expansion ROMs. 4256 Use with caution as certain devices share 4257 address decoders between ROMs and other 4258 resources. 4259 norom [X86] Do not assign address space to 4260 expansion ROMs that do not already have 4261 BIOS assigned address ranges. 4262 nobar [X86] Do not assign address space to the 4263 BARs that weren't assigned by the BIOS. 4264 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4265 assigned automatically to PCI devices. You can 4266 make the kernel exclude IRQs of your ISA cards 4267 this way. 4268 pirqaddr=0xAAAAA [X86] Specify the physical address 4269 of the PIRQ table (normally generated 4270 by the BIOS) if it is outside the 4271 F0000h-100000h range. 4272 lastbus=N [X86] Scan all buses thru bus #N. Can be 4273 useful if the kernel is unable to find your 4274 secondary buses and you want to tell it 4275 explicitly which ones they are. 4276 assign-busses [X86] Always assign all PCI bus 4277 numbers ourselves, overriding 4278 whatever the firmware may have done. 4279 usepirqmask [X86] Honor the possible IRQ mask stored 4280 in the BIOS $PIR table. This is needed on 4281 some systems with broken BIOSes, notably 4282 some HP Pavilion N5400 and Omnibook XE3 4283 notebooks. This will have no effect if ACPI 4284 IRQ routing is enabled. 4285 noacpi [X86] Do not use ACPI for IRQ routing 4286 or for PCI scanning. 4287 use_crs [X86] Use PCI host bridge window information 4288 from ACPI. On BIOSes from 2008 or later, this 4289 is enabled by default. If you need to use this, 4290 please report a bug. 4291 nocrs [X86] Ignore PCI host bridge windows from ACPI. 4292 If you need to use this, please report a bug. 4293 use_e820 [X86] Use E820 reservations to exclude parts of 4294 PCI host bridge windows. This is a workaround 4295 for BIOS defects in host bridge _CRS methods. 4296 If you need to use this, please report a bug to 4297 <linux-pci@vger.kernel.org>. 4298 no_e820 [X86] Ignore E820 reservations for PCI host 4299 bridge windows. This is the default on modern 4300 hardware. If you need to use this, please report 4301 a bug to <linux-pci@vger.kernel.org>. 4302 routeirq Do IRQ routing for all PCI devices. 4303 This is normally done in pci_enable_device(), 4304 so this option is a temporary workaround 4305 for broken drivers that don't call it. 4306 skip_isa_align [X86] do not align io start addr, so can 4307 handle more pci cards 4308 noearly [X86] Don't do any early type 1 scanning. 4309 This might help on some broken boards which 4310 machine check when some devices' config space 4311 is read. But various workarounds are disabled 4312 and some IOMMU drivers will not work. 4313 bfsort Sort PCI devices into breadth-first order. 4314 This sorting is done to get a device 4315 order compatible with older (<= 2.4) kernels. 4316 nobfsort Don't sort PCI devices into breadth-first order. 4317 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 4318 tuning and use the BIOS-configured MPS defaults. 4319 pcie_bus_safe Set every device's MPS to the largest value 4320 supported by all devices below the root complex. 4321 pcie_bus_perf Set device MPS to the largest allowable MPS 4322 based on its parent bus. Also set MRRS (Max 4323 Read Request Size) to the largest supported 4324 value (no larger than the MPS that the device 4325 or bus can support) for best performance. 4326 pcie_bus_peer2peer Set every device's MPS to 128B, which 4327 every device is guaranteed to support. This 4328 configuration allows peer-to-peer DMA between 4329 any pair of devices, possibly at the cost of 4330 reduced performance. This also guarantees 4331 that hot-added devices will work. 4332 cbiosize=nn[KMG] The fixed amount of bus space which is 4333 reserved for the CardBus bridge's IO window. 4334 The default value is 256 bytes. 4335 cbmemsize=nn[KMG] The fixed amount of bus space which is 4336 reserved for the CardBus bridge's memory 4337 window. The default value is 64 megabytes. 4338 resource_alignment= 4339 Format: 4340 [<order of align>@]<pci_dev>[; ...] 4341 Specifies alignment and device to reassign 4342 aligned memory resources. How to 4343 specify the device is described above. 4344 If <order of align> is not specified, 4345 PAGE_SIZE is used as alignment. 4346 A PCI-PCI bridge can be specified if resource 4347 windows need to be expanded. 4348 To specify the alignment for several 4349 instances of a device, the PCI vendor, 4350 device, subvendor, and subdevice may be 4351 specified, e.g., 12@pci:8086:9c22:103c:198f 4352 for 4096-byte alignment. 4353 ecrc= Enable/disable PCIe ECRC (transaction layer 4354 end-to-end CRC checking). Only effective if 4355 OS has native AER control (either granted by 4356 ACPI _OSC or forced via "pcie_ports=native") 4357 bios: Use BIOS/firmware settings. This is the 4358 the default. 4359 off: Turn ECRC off 4360 on: Turn ECRC on. 4361 hpiosize=nn[KMG] The fixed amount of bus space which is 4362 reserved for hotplug bridge's IO window. 4363 Default size is 256 bytes. 4364 hpmmiosize=nn[KMG] The fixed amount of bus space which is 4365 reserved for hotplug bridge's MMIO window. 4366 Default size is 2 megabytes. 4367 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 4368 reserved for hotplug bridge's MMIO_PREF window. 4369 Default size is 2 megabytes. 4370 hpmemsize=nn[KMG] The fixed amount of bus space which is 4371 reserved for hotplug bridge's MMIO and 4372 MMIO_PREF window. 4373 Default size is 2 megabytes. 4374 hpbussize=nn The minimum amount of additional bus numbers 4375 reserved for buses below a hotplug bridge. 4376 Default is 1. 4377 realloc= Enable/disable reallocating PCI bridge resources 4378 if allocations done by BIOS are too small to 4379 accommodate resources required by all child 4380 devices. 4381 off: Turn realloc off 4382 on: Turn realloc on 4383 realloc same as realloc=on 4384 noari do not use PCIe ARI. 4385 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 4386 do not use PCIe ATS (and IOMMU device IOTLB). 4387 pcie_scan_all Scan all possible PCIe devices. Otherwise we 4388 only look for one device below a PCIe downstream 4389 port. 4390 big_root_window Try to add a big 64bit memory window to the PCIe 4391 root complex on AMD CPUs. Some GFX hardware 4392 can resize a BAR to allow access to all VRAM. 4393 Adding the window is slightly risky (it may 4394 conflict with unreported devices), so this 4395 taints the kernel. 4396 disable_acs_redir=<pci_dev>[; ...] 4397 Specify one or more PCI devices (in the format 4398 specified above) separated by semicolons. 4399 Each device specified will have the PCI ACS 4400 redirect capabilities forced off which will 4401 allow P2P traffic between devices through 4402 bridges without forcing it upstream. Note: 4403 this removes isolation between devices and 4404 may put more devices in an IOMMU group. 4405 force_floating [S390] Force usage of floating interrupts. 4406 nomio [S390] Do not use MIO instructions. 4407 norid [S390] ignore the RID field and force use of 4408 one PCI domain per PCI function 4409 4410 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 4411 Management. 4412 off Disable ASPM. 4413 force Enable ASPM even on devices that claim not to support it. 4414 WARNING: Forcing ASPM on may cause system lockups. 4415 4416 pcie_ports= [PCIE] PCIe port services handling: 4417 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 4418 even if the platform doesn't give the OS permission to 4419 use them. This may cause conflicts if the platform 4420 also tries to use these services. 4421 dpc-native Use native PCIe service for DPC only. May 4422 cause conflicts if firmware uses AER or DPC. 4423 compat Disable native PCIe services (PME, AER, DPC, PCIe 4424 hotplug). 4425 4426 pcie_port_pm= [PCIE] PCIe port power management handling: 4427 off Disable power management of all PCIe ports 4428 force Forcibly enable power management of all PCIe ports 4429 4430 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 4431 nomsi Do not use MSI for native PCIe PME signaling (this makes 4432 all PCIe root ports use INTx for all services). 4433 4434 pcmv= [HW,PCMCIA] BadgePAD 4 4435 4436 pd_ignore_unused 4437 [PM] 4438 Keep all power-domains already enabled by bootloader on, 4439 even if no driver has claimed them. This is useful 4440 for debug and development, but should not be 4441 needed on a platform with proper driver support. 4442 4443 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 4444 boot time. 4445 Format: { 0 | 1 } 4446 See arch/parisc/kernel/pdc_chassis.c 4447 4448 percpu_alloc= Select which percpu first chunk allocator to use. 4449 Currently supported values are "embed" and "page". 4450 Archs may support subset or none of the selections. 4451 See comments in mm/percpu.c for details on each 4452 allocator. This parameter is primarily for debugging 4453 and performance comparison. 4454 4455 pirq= [SMP,APIC] Manual mp-table setup 4456 See Documentation/arch/x86/i386/IO-APIC.rst. 4457 4458 plip= [PPT,NET] Parallel port network link 4459 Format: { parport<nr> | timid | 0 } 4460 See also Documentation/admin-guide/parport.rst. 4461 4462 pmtmr= [X86] Manual setup of pmtmr I/O Port. 4463 Override pmtimer IOPort with a hex value. 4464 e.g. pmtmr=0x508 4465 4466 pmu_override= [PPC] Override the PMU. 4467 This option takes over the PMU facility, so it is no 4468 longer usable by perf. Setting this option starts the 4469 PMU counters by setting MMCR0 to 0 (the FC bit is 4470 cleared). If a number is given, then MMCR1 is set to 4471 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 4472 remains 0. 4473 4474 pm_debug_messages [SUSPEND,KNL] 4475 Enable suspend/resume debug messages during boot up. 4476 4477 pnp.debug=1 [PNP] 4478 Enable PNP debug messages (depends on the 4479 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 4480 via /sys/module/pnp/parameters/debug. We always show 4481 current resource usage; turning this on also shows 4482 possible settings and some assignment information. 4483 4484 pnpacpi= [ACPI] 4485 { off } 4486 4487 pnpbios= [ISAPNP] 4488 { on | off | curr | res | no-curr | no-res } 4489 4490 pnp_reserve_irq= 4491 [ISAPNP] Exclude IRQs for the autoconfiguration 4492 4493 pnp_reserve_dma= 4494 [ISAPNP] Exclude DMAs for the autoconfiguration 4495 4496 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 4497 Ranges are in pairs (I/O port base and size). 4498 4499 pnp_reserve_mem= 4500 [ISAPNP] Exclude memory regions for the 4501 autoconfiguration. 4502 Ranges are in pairs (memory base and size). 4503 4504 ports= [IP_VS_FTP] IPVS ftp helper module 4505 Default is 21. 4506 Up to 8 (IP_VS_APP_MAX_PORTS) ports 4507 may be specified. 4508 Format: <port>,<port>.... 4509 4510 powersave=off [PPC] This option disables power saving features. 4511 It specifically disables cpuidle and sets the 4512 platform machine description specific power_save 4513 function to NULL. On Idle the CPU just reduces 4514 execution priority. 4515 4516 ppc_strict_facility_enable 4517 [PPC] This option catches any kernel floating point, 4518 Altivec, VSX and SPE outside of regions specifically 4519 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 4520 There is some performance impact when enabling this. 4521 4522 ppc_tm= [PPC] 4523 Format: {"off"} 4524 Disable Hardware Transactional Memory 4525 4526 preempt= [KNL] 4527 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 4528 none - Limited to cond_resched() calls 4529 voluntary - Limited to cond_resched() and might_sleep() calls 4530 full - Any section that isn't explicitly preempt disabled 4531 can be preempted anytime. 4532 4533 print-fatal-signals= 4534 [KNL] debug: print fatal signals 4535 4536 If enabled, warn about various signal handling 4537 related application anomalies: too many signals, 4538 too many POSIX.1 timers, fatal signals causing a 4539 coredump - etc. 4540 4541 If you hit the warning due to signal overflow, 4542 you might want to try "ulimit -i unlimited". 4543 4544 default: off. 4545 4546 printk.always_kmsg_dump= 4547 Trigger kmsg_dump for cases other than kernel oops or 4548 panics 4549 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4550 default: disabled 4551 4552 printk.console_no_auto_verbose= 4553 Disable console loglevel raise on oops, panic 4554 or lockdep-detected issues (only if lock debug is on). 4555 With an exception to setups with low baudrate on 4556 serial console, keeping this 0 is a good choice 4557 in order to provide more debug information. 4558 Format: <bool> 4559 default: 0 (auto_verbose is enabled) 4560 4561 printk.devkmsg={on,off,ratelimit} 4562 Control writing to /dev/kmsg. 4563 on - unlimited logging to /dev/kmsg from userspace 4564 off - logging to /dev/kmsg disabled 4565 ratelimit - ratelimit the logging 4566 Default: ratelimit 4567 4568 printk.time= Show timing data prefixed to each printk message line 4569 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4570 4571 processor.max_cstate= [HW,ACPI] 4572 Limit processor to maximum C-state 4573 max_cstate=9 overrides any DMI blacklist limit. 4574 4575 processor.nocst [HW,ACPI] 4576 Ignore the _CST method to determine C-states, 4577 instead using the legacy FADT method 4578 4579 profile= [KNL] Enable kernel profiling via /proc/profile 4580 Format: [<profiletype>,]<number> 4581 Param: <profiletype>: "schedule", "sleep", or "kvm" 4582 [defaults to kernel profiling] 4583 Param: "schedule" - profile schedule points. 4584 Param: "sleep" - profile D-state sleeping (millisecs). 4585 Requires CONFIG_SCHEDSTATS 4586 Param: "kvm" - profile VM exits. 4587 Param: <number> - step/bucket size as a power of 2 for 4588 statistical time based profiling. 4589 4590 prompt_ramdisk= [RAM] [Deprecated] 4591 4592 prot_virt= [S390] enable hosting protected virtual machines 4593 isolated from the hypervisor (if hardware supports 4594 that). 4595 Format: <bool> 4596 4597 psi= [KNL] Enable or disable pressure stall information 4598 tracking. 4599 Format: <bool> 4600 4601 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 4602 probe for; one of (bare|imps|exps|lifebook|any). 4603 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 4604 per second. 4605 psmouse.resetafter= [HW,MOUSE] 4606 Try to reset the device after so many bad packets 4607 (0 = never). 4608 psmouse.resolution= 4609 [HW,MOUSE] Set desired mouse resolution, in dpi. 4610 psmouse.smartscroll= 4611 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 4612 0 = disabled, 1 = enabled (default). 4613 4614 pstore.backend= Specify the name of the pstore backend to use 4615 4616 pti= [X86-64] Control Page Table Isolation of user and 4617 kernel address spaces. Disabling this feature 4618 removes hardening, but improves performance of 4619 system calls and interrupts. 4620 4621 on - unconditionally enable 4622 off - unconditionally disable 4623 auto - kernel detects whether your CPU model is 4624 vulnerable to issues that PTI mitigates 4625 4626 Not specifying this option is equivalent to pti=auto. 4627 4628 pty.legacy_count= 4629 [KNL] Number of legacy pty's. Overwrites compiled-in 4630 default number. 4631 4632 quiet [KNL] Disable most log messages 4633 4634 r128= [HW,DRM] 4635 4636 radix_hcall_invalidate=on [PPC/PSERIES] 4637 Disable RADIX GTSE feature and use hcall for TLB 4638 invalidate. 4639 4640 raid= [HW,RAID] 4641 See Documentation/admin-guide/md.rst. 4642 4643 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 4644 See Documentation/admin-guide/blockdev/ramdisk.rst. 4645 4646 ramdisk_start= [RAM] RAM disk image start address 4647 4648 random.trust_cpu=off 4649 [KNL] Disable trusting the use of the CPU's 4650 random number generator (if available) to 4651 initialize the kernel's RNG. 4652 4653 random.trust_bootloader=off 4654 [KNL] Disable trusting the use of the a seed 4655 passed by the bootloader (if available) to 4656 initialize the kernel's RNG. 4657 4658 randomize_kstack_offset= 4659 [KNL] Enable or disable kernel stack offset 4660 randomization, which provides roughly 5 bits of 4661 entropy, frustrating memory corruption attacks 4662 that depend on stack address determinism or 4663 cross-syscall address exposures. This is only 4664 available on architectures that have defined 4665 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 4666 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4667 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 4668 4669 ras=option[,option,...] [KNL] RAS-specific options 4670 4671 cec_disable [X86] 4672 Disable the Correctable Errors Collector, 4673 see CONFIG_RAS_CEC help text. 4674 4675 rcu_nocbs[=cpu-list] 4676 [KNL] The optional argument is a cpu list, 4677 as described above. 4678 4679 In kernels built with CONFIG_RCU_NOCB_CPU=y, 4680 enable the no-callback CPU mode, which prevents 4681 such CPUs' callbacks from being invoked in 4682 softirq context. Invocation of such CPUs' RCU 4683 callbacks will instead be offloaded to "rcuox/N" 4684 kthreads created for that purpose, where "x" is 4685 "p" for RCU-preempt, "s" for RCU-sched, and "g" 4686 for the kthreads that mediate grace periods; and 4687 "N" is the CPU number. This reduces OS jitter on 4688 the offloaded CPUs, which can be useful for HPC 4689 and real-time workloads. It can also improve 4690 energy efficiency for asymmetric multiprocessors. 4691 4692 If a cpulist is passed as an argument, the specified 4693 list of CPUs is set to no-callback mode from boot. 4694 4695 Otherwise, if the '=' sign and the cpulist 4696 arguments are omitted, no CPU will be set to 4697 no-callback mode from boot but the mode may be 4698 toggled at runtime via cpusets. 4699 4700 Note that this argument takes precedence over 4701 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4702 4703 rcu_nocb_poll [KNL] 4704 Rather than requiring that offloaded CPUs 4705 (specified by rcu_nocbs= above) explicitly 4706 awaken the corresponding "rcuoN" kthreads, 4707 make these kthreads poll for callbacks. 4708 This improves the real-time response for the 4709 offloaded CPUs by relieving them of the need to 4710 wake up the corresponding kthread, but degrades 4711 energy efficiency by requiring that the kthreads 4712 periodically wake up to do the polling. 4713 4714 rcutree.blimit= [KNL] 4715 Set maximum number of finished RCU callbacks to 4716 process in one batch. 4717 4718 rcutree.dump_tree= [KNL] 4719 Dump the structure of the rcu_node combining tree 4720 out at early boot. This is used for diagnostic 4721 purposes, to verify correct tree setup. 4722 4723 rcutree.gp_cleanup_delay= [KNL] 4724 Set the number of jiffies to delay each step of 4725 RCU grace-period cleanup. 4726 4727 rcutree.gp_init_delay= [KNL] 4728 Set the number of jiffies to delay each step of 4729 RCU grace-period initialization. 4730 4731 rcutree.gp_preinit_delay= [KNL] 4732 Set the number of jiffies to delay each step of 4733 RCU grace-period pre-initialization, that is, 4734 the propagation of recent CPU-hotplug changes up 4735 the rcu_node combining tree. 4736 4737 rcutree.use_softirq= [KNL] 4738 If set to zero, move all RCU_SOFTIRQ processing to 4739 per-CPU rcuc kthreads. Defaults to a non-zero 4740 value, meaning that RCU_SOFTIRQ is used by default. 4741 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4742 4743 But note that CONFIG_PREEMPT_RT=y kernels disable 4744 this kernel boot parameter, forcibly setting it 4745 to zero. 4746 4747 rcutree.rcu_fanout_exact= [KNL] 4748 Disable autobalancing of the rcu_node combining 4749 tree. This is used by rcutorture, and might 4750 possibly be useful for architectures having high 4751 cache-to-cache transfer latencies. 4752 4753 rcutree.rcu_fanout_leaf= [KNL] 4754 Change the number of CPUs assigned to each 4755 leaf rcu_node structure. Useful for very 4756 large systems, which will choose the value 64, 4757 and for NUMA systems with large remote-access 4758 latencies, which will choose a value aligned 4759 with the appropriate hardware boundaries. 4760 4761 rcutree.rcu_min_cached_objs= [KNL] 4762 Minimum number of objects which are cached and 4763 maintained per one CPU. Object size is equal 4764 to PAGE_SIZE. The cache allows to reduce the 4765 pressure to page allocator, also it makes the 4766 whole algorithm to behave better in low memory 4767 condition. 4768 4769 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 4770 Set the page-cache refill delay (in milliseconds) 4771 in response to low-memory conditions. The range 4772 of permitted values is in the range 0:100000. 4773 4774 rcutree.jiffies_till_first_fqs= [KNL] 4775 Set delay from grace-period initialization to 4776 first attempt to force quiescent states. 4777 Units are jiffies, minimum value is zero, 4778 and maximum value is HZ. 4779 4780 rcutree.jiffies_till_next_fqs= [KNL] 4781 Set delay between subsequent attempts to force 4782 quiescent states. Units are jiffies, minimum 4783 value is one, and maximum value is HZ. 4784 4785 rcutree.jiffies_till_sched_qs= [KNL] 4786 Set required age in jiffies for a 4787 given grace period before RCU starts 4788 soliciting quiescent-state help from 4789 rcu_note_context_switch() and cond_resched(). 4790 If not specified, the kernel will calculate 4791 a value based on the most recent settings 4792 of rcutree.jiffies_till_first_fqs 4793 and rcutree.jiffies_till_next_fqs. 4794 This calculated value may be viewed in 4795 rcutree.jiffies_to_sched_qs. Any attempt to set 4796 rcutree.jiffies_to_sched_qs will be cheerfully 4797 overwritten. 4798 4799 rcutree.kthread_prio= [KNL,BOOT] 4800 Set the SCHED_FIFO priority of the RCU per-CPU 4801 kthreads (rcuc/N). This value is also used for 4802 the priority of the RCU boost threads (rcub/N) 4803 and for the RCU grace-period kthreads (rcu_bh, 4804 rcu_preempt, and rcu_sched). If RCU_BOOST is 4805 set, valid values are 1-99 and the default is 1 4806 (the least-favored priority). Otherwise, when 4807 RCU_BOOST is not set, valid values are 0-99 and 4808 the default is zero (non-realtime operation). 4809 When RCU_NOCB_CPU is set, also adjust the 4810 priority of NOCB callback kthreads. 4811 4812 rcutree.rcu_divisor= [KNL] 4813 Set the shift-right count to use to compute 4814 the callback-invocation batch limit bl from 4815 the number of callbacks queued on this CPU. 4816 The result will be bounded below by the value of 4817 the rcutree.blimit kernel parameter. Every bl 4818 callbacks, the softirq handler will exit in 4819 order to allow the CPU to do other work. 4820 4821 Please note that this callback-invocation batch 4822 limit applies only to non-offloaded callback 4823 invocation. Offloaded callbacks are instead 4824 invoked in the context of an rcuoc kthread, which 4825 scheduler will preempt as it does any other task. 4826 4827 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 4828 On callback-offloaded (rcu_nocbs) CPUs, 4829 RCU reduces the lock contention that would 4830 otherwise be caused by callback floods through 4831 use of the ->nocb_bypass list. However, in the 4832 common non-flooded case, RCU queues directly to 4833 the main ->cblist in order to avoid the extra 4834 overhead of the ->nocb_bypass list and its lock. 4835 But if there are too many callbacks queued during 4836 a single jiffy, RCU pre-queues the callbacks into 4837 the ->nocb_bypass queue. The definition of "too 4838 many" is supplied by this kernel boot parameter. 4839 4840 rcutree.rcu_nocb_gp_stride= [KNL] 4841 Set the number of NOCB callback kthreads in 4842 each group, which defaults to the square root 4843 of the number of CPUs. Larger numbers reduce 4844 the wakeup overhead on the global grace-period 4845 kthread, but increases that same overhead on 4846 each group's NOCB grace-period kthread. 4847 4848 rcutree.qhimark= [KNL] 4849 Set threshold of queued RCU callbacks beyond which 4850 batch limiting is disabled. 4851 4852 rcutree.qlowmark= [KNL] 4853 Set threshold of queued RCU callbacks below which 4854 batch limiting is re-enabled. 4855 4856 rcutree.qovld= [KNL] 4857 Set threshold of queued RCU callbacks beyond which 4858 RCU's force-quiescent-state scan will aggressively 4859 enlist help from cond_resched() and sched IPIs to 4860 help CPUs more quickly reach quiescent states. 4861 Set to less than zero to make this be set based 4862 on rcutree.qhimark at boot time and to zero to 4863 disable more aggressive help enlistment. 4864 4865 rcutree.rcu_kick_kthreads= [KNL] 4866 Cause the grace-period kthread to get an extra 4867 wake_up() if it sleeps three times longer than 4868 it should at force-quiescent-state time. 4869 This wake_up() will be accompanied by a 4870 WARN_ONCE() splat and an ftrace_dump(). 4871 4872 rcutree.rcu_unlock_delay= [KNL] 4873 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 4874 this specifies an rcu_read_unlock()-time delay 4875 in microseconds. This defaults to zero. 4876 Larger delays increase the probability of 4877 catching RCU pointer leaks, that is, buggy use 4878 of RCU-protected pointers after the relevant 4879 rcu_read_unlock() has completed. 4880 4881 rcutree.sysrq_rcu= [KNL] 4882 Commandeer a sysrq key to dump out Tree RCU's 4883 rcu_node tree with an eye towards determining 4884 why a new grace period has not yet started. 4885 4886 rcuscale.gp_async= [KNL] 4887 Measure performance of asynchronous 4888 grace-period primitives such as call_rcu(). 4889 4890 rcuscale.gp_async_max= [KNL] 4891 Specify the maximum number of outstanding 4892 callbacks per writer thread. When a writer 4893 thread exceeds this limit, it invokes the 4894 corresponding flavor of rcu_barrier() to allow 4895 previously posted callbacks to drain. 4896 4897 rcuscale.gp_exp= [KNL] 4898 Measure performance of expedited synchronous 4899 grace-period primitives. 4900 4901 rcuscale.holdoff= [KNL] 4902 Set test-start holdoff period. The purpose of 4903 this parameter is to delay the start of the 4904 test until boot completes in order to avoid 4905 interference. 4906 4907 rcuscale.kfree_rcu_test= [KNL] 4908 Set to measure performance of kfree_rcu() flooding. 4909 4910 rcuscale.kfree_rcu_test_double= [KNL] 4911 Test the double-argument variant of kfree_rcu(). 4912 If this parameter has the same value as 4913 rcuscale.kfree_rcu_test_single, both the single- 4914 and double-argument variants are tested. 4915 4916 rcuscale.kfree_rcu_test_single= [KNL] 4917 Test the single-argument variant of kfree_rcu(). 4918 If this parameter has the same value as 4919 rcuscale.kfree_rcu_test_double, both the single- 4920 and double-argument variants are tested. 4921 4922 rcuscale.kfree_nthreads= [KNL] 4923 The number of threads running loops of kfree_rcu(). 4924 4925 rcuscale.kfree_alloc_num= [KNL] 4926 Number of allocations and frees done in an iteration. 4927 4928 rcuscale.kfree_loops= [KNL] 4929 Number of loops doing rcuscale.kfree_alloc_num number 4930 of allocations and frees. 4931 4932 rcuscale.nreaders= [KNL] 4933 Set number of RCU readers. The value -1 selects 4934 N, where N is the number of CPUs. A value 4935 "n" less than -1 selects N-n+1, where N is again 4936 the number of CPUs. For example, -2 selects N 4937 (the number of CPUs), -3 selects N+1, and so on. 4938 A value of "n" less than or equal to -N selects 4939 a single reader. 4940 4941 rcuscale.nwriters= [KNL] 4942 Set number of RCU writers. The values operate 4943 the same as for rcuscale.nreaders. 4944 N, where N is the number of CPUs 4945 4946 rcuscale.perf_type= [KNL] 4947 Specify the RCU implementation to test. 4948 4949 rcuscale.shutdown= [KNL] 4950 Shut the system down after performance tests 4951 complete. This is useful for hands-off automated 4952 testing. 4953 4954 rcuscale.verbose= [KNL] 4955 Enable additional printk() statements. 4956 4957 rcuscale.writer_holdoff= [KNL] 4958 Write-side holdoff between grace periods, 4959 in microseconds. The default of zero says 4960 no holdoff. 4961 4962 rcutorture.fqs_duration= [KNL] 4963 Set duration of force_quiescent_state bursts 4964 in microseconds. 4965 4966 rcutorture.fqs_holdoff= [KNL] 4967 Set holdoff time within force_quiescent_state bursts 4968 in microseconds. 4969 4970 rcutorture.fqs_stutter= [KNL] 4971 Set wait time between force_quiescent_state bursts 4972 in seconds. 4973 4974 rcutorture.fwd_progress= [KNL] 4975 Specifies the number of kthreads to be used 4976 for RCU grace-period forward-progress testing 4977 for the types of RCU supporting this notion. 4978 Defaults to 1 kthread, values less than zero or 4979 greater than the number of CPUs cause the number 4980 of CPUs to be used. 4981 4982 rcutorture.fwd_progress_div= [KNL] 4983 Specify the fraction of a CPU-stall-warning 4984 period to do tight-loop forward-progress testing. 4985 4986 rcutorture.fwd_progress_holdoff= [KNL] 4987 Number of seconds to wait between successive 4988 forward-progress tests. 4989 4990 rcutorture.fwd_progress_need_resched= [KNL] 4991 Enclose cond_resched() calls within checks for 4992 need_resched() during tight-loop forward-progress 4993 testing. 4994 4995 rcutorture.gp_cond= [KNL] 4996 Use conditional/asynchronous update-side 4997 primitives, if available. 4998 4999 rcutorture.gp_exp= [KNL] 5000 Use expedited update-side primitives, if available. 5001 5002 rcutorture.gp_normal= [KNL] 5003 Use normal (non-expedited) asynchronous 5004 update-side primitives, if available. 5005 5006 rcutorture.gp_sync= [KNL] 5007 Use normal (non-expedited) synchronous 5008 update-side primitives, if available. If all 5009 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5010 rcutorture.gp_normal=, and rcutorture.gp_sync= 5011 are zero, rcutorture acts as if is interpreted 5012 they are all non-zero. 5013 5014 rcutorture.irqreader= [KNL] 5015 Run RCU readers from irq handlers, or, more 5016 accurately, from a timer handler. Not all RCU 5017 flavors take kindly to this sort of thing. 5018 5019 rcutorture.leakpointer= [KNL] 5020 Leak an RCU-protected pointer out of the reader. 5021 This can of course result in splats, and is 5022 intended to test the ability of things like 5023 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5024 such leaks. 5025 5026 rcutorture.n_barrier_cbs= [KNL] 5027 Set callbacks/threads for rcu_barrier() testing. 5028 5029 rcutorture.nfakewriters= [KNL] 5030 Set number of concurrent RCU writers. These just 5031 stress RCU, they don't participate in the actual 5032 test, hence the "fake". 5033 5034 rcutorture.nocbs_nthreads= [KNL] 5035 Set number of RCU callback-offload togglers. 5036 Zero (the default) disables toggling. 5037 5038 rcutorture.nocbs_toggle= [KNL] 5039 Set the delay in milliseconds between successive 5040 callback-offload toggling attempts. 5041 5042 rcutorture.nreaders= [KNL] 5043 Set number of RCU readers. The value -1 selects 5044 N-1, where N is the number of CPUs. A value 5045 "n" less than -1 selects N-n-2, where N is again 5046 the number of CPUs. For example, -2 selects N 5047 (the number of CPUs), -3 selects N+1, and so on. 5048 5049 rcutorture.object_debug= [KNL] 5050 Enable debug-object double-call_rcu() testing. 5051 5052 rcutorture.onoff_holdoff= [KNL] 5053 Set time (s) after boot for CPU-hotplug testing. 5054 5055 rcutorture.onoff_interval= [KNL] 5056 Set time (jiffies) between CPU-hotplug operations, 5057 or zero to disable CPU-hotplug testing. 5058 5059 rcutorture.read_exit= [KNL] 5060 Set the number of read-then-exit kthreads used 5061 to test the interaction of RCU updaters and 5062 task-exit processing. 5063 5064 rcutorture.read_exit_burst= [KNL] 5065 The number of times in a given read-then-exit 5066 episode that a set of read-then-exit kthreads 5067 is spawned. 5068 5069 rcutorture.read_exit_delay= [KNL] 5070 The delay, in seconds, between successive 5071 read-then-exit testing episodes. 5072 5073 rcutorture.shuffle_interval= [KNL] 5074 Set task-shuffle interval (s). Shuffling tasks 5075 allows some CPUs to go into dyntick-idle mode 5076 during the rcutorture test. 5077 5078 rcutorture.shutdown_secs= [KNL] 5079 Set time (s) after boot system shutdown. This 5080 is useful for hands-off automated testing. 5081 5082 rcutorture.stall_cpu= [KNL] 5083 Duration of CPU stall (s) to test RCU CPU stall 5084 warnings, zero to disable. 5085 5086 rcutorture.stall_cpu_block= [KNL] 5087 Sleep while stalling if set. This will result 5088 in warnings from preemptible RCU in addition 5089 to any other stall-related activity. 5090 5091 rcutorture.stall_cpu_holdoff= [KNL] 5092 Time to wait (s) after boot before inducing stall. 5093 5094 rcutorture.stall_cpu_irqsoff= [KNL] 5095 Disable interrupts while stalling if set. 5096 5097 rcutorture.stall_gp_kthread= [KNL] 5098 Duration (s) of forced sleep within RCU 5099 grace-period kthread to test RCU CPU stall 5100 warnings, zero to disable. If both stall_cpu 5101 and stall_gp_kthread are specified, the 5102 kthread is starved first, then the CPU. 5103 5104 rcutorture.stat_interval= [KNL] 5105 Time (s) between statistics printk()s. 5106 5107 rcutorture.stutter= [KNL] 5108 Time (s) to stutter testing, for example, specifying 5109 five seconds causes the test to run for five seconds, 5110 wait for five seconds, and so on. This tests RCU's 5111 ability to transition abruptly to and from idle. 5112 5113 rcutorture.test_boost= [KNL] 5114 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 5115 "Maybe" means test if the RCU implementation 5116 under test support RCU priority boosting. 5117 5118 rcutorture.test_boost_duration= [KNL] 5119 Duration (s) of each individual boost test. 5120 5121 rcutorture.test_boost_interval= [KNL] 5122 Interval (s) between each boost test. 5123 5124 rcutorture.test_no_idle_hz= [KNL] 5125 Test RCU's dyntick-idle handling. See also the 5126 rcutorture.shuffle_interval parameter. 5127 5128 rcutorture.torture_type= [KNL] 5129 Specify the RCU implementation to test. 5130 5131 rcutorture.verbose= [KNL] 5132 Enable additional printk() statements. 5133 5134 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 5135 Dump ftrace buffer after reporting RCU CPU 5136 stall warning. 5137 5138 rcupdate.rcu_cpu_stall_suppress= [KNL] 5139 Suppress RCU CPU stall warning messages. 5140 5141 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 5142 Suppress RCU CPU stall warning messages and 5143 rcutorture writer stall warnings that occur 5144 during early boot, that is, during the time 5145 before the init task is spawned. 5146 5147 rcupdate.rcu_cpu_stall_timeout= [KNL] 5148 Set timeout for RCU CPU stall warning messages. 5149 The value is in seconds and the maximum allowed 5150 value is 300 seconds. 5151 5152 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 5153 Set timeout for expedited RCU CPU stall warning 5154 messages. The value is in milliseconds 5155 and the maximum allowed value is 21000 5156 milliseconds. Please note that this value is 5157 adjusted to an arch timer tick resolution. 5158 Setting this to zero causes the value from 5159 rcupdate.rcu_cpu_stall_timeout to be used (after 5160 conversion from seconds to milliseconds). 5161 5162 rcupdate.rcu_cpu_stall_cputime= [KNL] 5163 Provide statistics on the cputime and count of 5164 interrupts and tasks during the sampling period. For 5165 multiple continuous RCU stalls, all sampling periods 5166 begin at half of the first RCU stall timeout. 5167 5168 rcupdate.rcu_exp_stall_task_details= [KNL] 5169 Print stack dumps of any tasks blocking the 5170 current expedited RCU grace period during an 5171 expedited RCU CPU stall warning. 5172 5173 rcupdate.rcu_expedited= [KNL] 5174 Use expedited grace-period primitives, for 5175 example, synchronize_rcu_expedited() instead 5176 of synchronize_rcu(). This reduces latency, 5177 but can increase CPU utilization, degrade 5178 real-time latency, and degrade energy efficiency. 5179 No effect on CONFIG_TINY_RCU kernels. 5180 5181 rcupdate.rcu_normal= [KNL] 5182 Use only normal grace-period primitives, 5183 for example, synchronize_rcu() instead of 5184 synchronize_rcu_expedited(). This improves 5185 real-time latency, CPU utilization, and 5186 energy efficiency, but can expose users to 5187 increased grace-period latency. This parameter 5188 overrides rcupdate.rcu_expedited. No effect on 5189 CONFIG_TINY_RCU kernels. 5190 5191 rcupdate.rcu_normal_after_boot= [KNL] 5192 Once boot has completed (that is, after 5193 rcu_end_inkernel_boot() has been invoked), use 5194 only normal grace-period primitives. No effect 5195 on CONFIG_TINY_RCU kernels. 5196 5197 But note that CONFIG_PREEMPT_RT=y kernels enables 5198 this kernel boot parameter, forcibly setting 5199 it to the value one, that is, converting any 5200 post-boot attempt at an expedited RCU grace 5201 period to instead use normal non-expedited 5202 grace-period processing. 5203 5204 rcupdate.rcu_task_collapse_lim= [KNL] 5205 Set the maximum number of callbacks present 5206 at the beginning of a grace period that allows 5207 the RCU Tasks flavors to collapse back to using 5208 a single callback queue. This switching only 5209 occurs when rcupdate.rcu_task_enqueue_lim is 5210 set to the default value of -1. 5211 5212 rcupdate.rcu_task_contend_lim= [KNL] 5213 Set the minimum number of callback-queuing-time 5214 lock-contention events per jiffy required to 5215 cause the RCU Tasks flavors to switch to per-CPU 5216 callback queuing. This switching only occurs 5217 when rcupdate.rcu_task_enqueue_lim is set to 5218 the default value of -1. 5219 5220 rcupdate.rcu_task_enqueue_lim= [KNL] 5221 Set the number of callback queues to use for the 5222 RCU Tasks family of RCU flavors. The default 5223 of -1 allows this to be automatically (and 5224 dynamically) adjusted. This parameter is intended 5225 for use in testing. 5226 5227 rcupdate.rcu_task_ipi_delay= [KNL] 5228 Set time in jiffies during which RCU tasks will 5229 avoid sending IPIs, starting with the beginning 5230 of a given grace period. Setting a large 5231 number avoids disturbing real-time workloads, 5232 but lengthens grace periods. 5233 5234 rcupdate.rcu_task_stall_info= [KNL] 5235 Set initial timeout in jiffies for RCU task stall 5236 informational messages, which give some indication 5237 of the problem for those not patient enough to 5238 wait for ten minutes. Informational messages are 5239 only printed prior to the stall-warning message 5240 for a given grace period. Disable with a value 5241 less than or equal to zero. Defaults to ten 5242 seconds. A change in value does not take effect 5243 until the beginning of the next grace period. 5244 5245 rcupdate.rcu_task_stall_info_mult= [KNL] 5246 Multiplier for time interval between successive 5247 RCU task stall informational messages for a given 5248 RCU tasks grace period. This value is clamped 5249 to one through ten, inclusive. It defaults to 5250 the value three, so that the first informational 5251 message is printed 10 seconds into the grace 5252 period, the second at 40 seconds, the third at 5253 160 seconds, and then the stall warning at 600 5254 seconds would prevent a fourth at 640 seconds. 5255 5256 rcupdate.rcu_task_stall_timeout= [KNL] 5257 Set timeout in jiffies for RCU task stall 5258 warning messages. Disable with a value less 5259 than or equal to zero. Defaults to ten minutes. 5260 A change in value does not take effect until 5261 the beginning of the next grace period. 5262 5263 rcupdate.rcu_self_test= [KNL] 5264 Run the RCU early boot self tests 5265 5266 rdinit= [KNL] 5267 Format: <full_path> 5268 Run specified binary instead of /init from the ramdisk, 5269 used for early userspace startup. See initrd. 5270 5271 rdrand= [X86] 5272 force - Override the decision by the kernel to hide the 5273 advertisement of RDRAND support (this affects 5274 certain AMD processors because of buggy BIOS 5275 support, specifically around the suspend/resume 5276 path). 5277 5278 rdt= [HW,X86,RDT] 5279 Turn on/off individual RDT features. List is: 5280 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 5281 mba, smba, bmec. 5282 E.g. to turn on cmt and turn off mba use: 5283 rdt=cmt,!mba 5284 5285 reboot= [KNL] 5286 Format (x86 or x86_64): 5287 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 5288 [[,]s[mp]#### \ 5289 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 5290 [[,]f[orce] 5291 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 5292 (prefix with 'panic_' to set mode for panic 5293 reboot only), 5294 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 5295 reboot_force is either force or not specified, 5296 reboot_cpu is s[mp]#### with #### being the processor 5297 to be used for rebooting. 5298 5299 refscale.holdoff= [KNL] 5300 Set test-start holdoff period. The purpose of 5301 this parameter is to delay the start of the 5302 test until boot completes in order to avoid 5303 interference. 5304 5305 refscale.loops= [KNL] 5306 Set the number of loops over the synchronization 5307 primitive under test. Increasing this number 5308 reduces noise due to loop start/end overhead, 5309 but the default has already reduced the per-pass 5310 noise to a handful of picoseconds on ca. 2020 5311 x86 laptops. 5312 5313 refscale.nreaders= [KNL] 5314 Set number of readers. The default value of -1 5315 selects N, where N is roughly 75% of the number 5316 of CPUs. A value of zero is an interesting choice. 5317 5318 refscale.nruns= [KNL] 5319 Set number of runs, each of which is dumped onto 5320 the console log. 5321 5322 refscale.readdelay= [KNL] 5323 Set the read-side critical-section duration, 5324 measured in microseconds. 5325 5326 refscale.scale_type= [KNL] 5327 Specify the read-protection implementation to test. 5328 5329 refscale.shutdown= [KNL] 5330 Shut down the system at the end of the performance 5331 test. This defaults to 1 (shut it down) when 5332 refscale is built into the kernel and to 0 (leave 5333 it running) when refscale is built as a module. 5334 5335 refscale.verbose= [KNL] 5336 Enable additional printk() statements. 5337 5338 refscale.verbose_batched= [KNL] 5339 Batch the additional printk() statements. If zero 5340 (the default) or negative, print everything. Otherwise, 5341 print every Nth verbose statement, where N is the value 5342 specified. 5343 5344 relax_domain_level= 5345 [KNL, SMP] Set scheduler's default relax_domain_level. 5346 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 5347 5348 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 5349 Format: <base1>,<size1>[,<base2>,<size2>,...] 5350 Reserve I/O ports or memory so the kernel won't use 5351 them. If <base> is less than 0x10000, the region 5352 is assumed to be I/O ports; otherwise it is memory. 5353 5354 reservetop= [X86-32] 5355 Format: nn[KMG] 5356 Reserves a hole at the top of the kernel virtual 5357 address space. 5358 5359 reset_devices [KNL] Force drivers to reset the underlying device 5360 during initialization. 5361 5362 resume= [SWSUSP] 5363 Specify the partition device for software suspend 5364 Format: 5365 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 5366 5367 resume_offset= [SWSUSP] 5368 Specify the offset from the beginning of the partition 5369 given by "resume=" at which the swap header is located, 5370 in <PAGE_SIZE> units (needed only for swap files). 5371 See Documentation/power/swsusp-and-swap-files.rst 5372 5373 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 5374 read the resume files 5375 5376 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 5377 Useful for devices that are detected asynchronously 5378 (e.g. USB and MMC devices). 5379 5380 retain_initrd [RAM] Keep initrd memory after extraction 5381 5382 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 5383 Speculative Code Execution with Return Instructions) 5384 vulnerability. 5385 5386 AMD-based UNRET and IBPB mitigations alone do not stop 5387 sibling threads from influencing the predictions of other 5388 sibling threads. For that reason, STIBP is used on pro- 5389 cessors that support it, and mitigate SMT on processors 5390 that don't. 5391 5392 off - no mitigation 5393 auto - automatically select a migitation 5394 auto,nosmt - automatically select a mitigation, 5395 disabling SMT if necessary for 5396 the full mitigation (only on Zen1 5397 and older without STIBP). 5398 ibpb - On AMD, mitigate short speculation 5399 windows on basic block boundaries too. 5400 Safe, highest perf impact. It also 5401 enables STIBP if present. Not suitable 5402 on Intel. 5403 ibpb,nosmt - Like "ibpb" above but will disable SMT 5404 when STIBP is not available. This is 5405 the alternative for systems which do not 5406 have STIBP. 5407 unret - Force enable untrained return thunks, 5408 only effective on AMD f15h-f17h based 5409 systems. 5410 unret,nosmt - Like unret, but will disable SMT when STIBP 5411 is not available. This is the alternative for 5412 systems which do not have STIBP. 5413 5414 Selecting 'auto' will choose a mitigation method at run 5415 time according to the CPU. 5416 5417 Not specifying this option is equivalent to retbleed=auto. 5418 5419 rfkill.default_state= 5420 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 5421 etc. communication is blocked by default. 5422 1 Unblocked. 5423 5424 rfkill.master_switch_mode= 5425 0 The "airplane mode" button does nothing. 5426 1 The "airplane mode" button toggles between everything 5427 blocked and the previous configuration. 5428 2 The "airplane mode" button toggles between everything 5429 blocked and everything unblocked. 5430 5431 rhash_entries= [KNL,NET] 5432 Set number of hash buckets for route cache 5433 5434 ring3mwait=disable 5435 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 5436 CPUs. 5437 5438 ro [KNL] Mount root device read-only on boot 5439 5440 rodata= [KNL] 5441 on Mark read-only kernel memory as read-only (default). 5442 off Leave read-only kernel memory writable for debugging. 5443 full Mark read-only kernel memory and aliases as read-only 5444 [arm64] 5445 5446 rockchip.usb_uart 5447 Enable the uart passthrough on the designated usb port 5448 on Rockchip SoCs. When active, the signals of the 5449 debug-uart get routed to the D+ and D- pins of the usb 5450 port and the regular usb controller gets disabled. 5451 5452 root= [KNL] Root filesystem 5453 See name_to_dev_t comment in init/do_mounts.c. 5454 5455 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 5456 mount the root filesystem 5457 5458 rootflags= [KNL] Set root filesystem mount option string 5459 5460 rootfstype= [KNL] Set root filesystem type 5461 5462 rootwait [KNL] Wait (indefinitely) for root device to show up. 5463 Useful for devices that are detected asynchronously 5464 (e.g. USB and MMC devices). 5465 5466 rproc_mem=nn[KMG][@address] 5467 [KNL,ARM,CMA] Remoteproc physical memory block. 5468 Memory area to be used by remote processor image, 5469 managed by CMA. 5470 5471 rw [KNL] Mount root device read-write on boot 5472 5473 S [KNL] Run init in single mode 5474 5475 s390_iommu= [HW,S390] 5476 Set s390 IOTLB flushing mode 5477 strict 5478 With strict flushing every unmap operation will result in 5479 an IOTLB flush. Default is lazy flushing before reuse, 5480 which is faster. 5481 5482 s390_iommu_aperture= [KNL,S390] 5483 Specifies the size of the per device DMA address space 5484 accessible through the DMA and IOMMU APIs as a decimal 5485 factor of the size of main memory. 5486 The default is 1 meaning that one can concurrently use 5487 as many DMA addresses as physical memory is installed, 5488 if supported by hardware, and thus map all of memory 5489 once. With a value of 2 one can map all of memory twice 5490 and so on. As a special case a factor of 0 imposes no 5491 restrictions other than those given by hardware at the 5492 cost of significant additional memory use for tables. 5493 5494 sa1100ir [NET] 5495 See drivers/net/irda/sa1100_ir.c. 5496 5497 sched_verbose [KNL] Enables verbose scheduler debug messages. 5498 5499 schedstats= [KNL,X86] Enable or disable scheduled statistics. 5500 Allowed values are enable and disable. This feature 5501 incurs a small amount of overhead in the scheduler 5502 but is useful for debugging and performance tuning. 5503 5504 sched_thermal_decay_shift= 5505 [KNL, SMP] Set a decay shift for scheduler thermal 5506 pressure signal. Thermal pressure signal follows the 5507 default decay period of other scheduler pelt 5508 signals(usually 32 ms but configurable). Setting 5509 sched_thermal_decay_shift will left shift the decay 5510 period for the thermal pressure signal by the shift 5511 value. 5512 i.e. with the default pelt decay period of 32 ms 5513 sched_thermal_decay_shift thermal pressure decay pr 5514 1 64 ms 5515 2 128 ms 5516 and so on. 5517 Format: integer between 0 and 10 5518 Default is 0. 5519 5520 scftorture.holdoff= [KNL] 5521 Number of seconds to hold off before starting 5522 test. Defaults to zero for module insertion and 5523 to 10 seconds for built-in smp_call_function() 5524 tests. 5525 5526 scftorture.longwait= [KNL] 5527 Request ridiculously long waits randomly selected 5528 up to the chosen limit in seconds. Zero (the 5529 default) disables this feature. Please note 5530 that requesting even small non-zero numbers of 5531 seconds can result in RCU CPU stall warnings, 5532 softlockup complaints, and so on. 5533 5534 scftorture.nthreads= [KNL] 5535 Number of kthreads to spawn to invoke the 5536 smp_call_function() family of functions. 5537 The default of -1 specifies a number of kthreads 5538 equal to the number of CPUs. 5539 5540 scftorture.onoff_holdoff= [KNL] 5541 Number seconds to wait after the start of the 5542 test before initiating CPU-hotplug operations. 5543 5544 scftorture.onoff_interval= [KNL] 5545 Number seconds to wait between successive 5546 CPU-hotplug operations. Specifying zero (which 5547 is the default) disables CPU-hotplug operations. 5548 5549 scftorture.shutdown_secs= [KNL] 5550 The number of seconds following the start of the 5551 test after which to shut down the system. The 5552 default of zero avoids shutting down the system. 5553 Non-zero values are useful for automated tests. 5554 5555 scftorture.stat_interval= [KNL] 5556 The number of seconds between outputting the 5557 current test statistics to the console. A value 5558 of zero disables statistics output. 5559 5560 scftorture.stutter_cpus= [KNL] 5561 The number of jiffies to wait between each change 5562 to the set of CPUs under test. 5563 5564 scftorture.use_cpus_read_lock= [KNL] 5565 Use use_cpus_read_lock() instead of the default 5566 preempt_disable() to disable CPU hotplug 5567 while invoking one of the smp_call_function*() 5568 functions. 5569 5570 scftorture.verbose= [KNL] 5571 Enable additional printk() statements. 5572 5573 scftorture.weight_single= [KNL] 5574 The probability weighting to use for the 5575 smp_call_function_single() function with a zero 5576 "wait" parameter. A value of -1 selects the 5577 default if all other weights are -1. However, 5578 if at least one weight has some other value, a 5579 value of -1 will instead select a weight of zero. 5580 5581 scftorture.weight_single_wait= [KNL] 5582 The probability weighting to use for the 5583 smp_call_function_single() function with a 5584 non-zero "wait" parameter. See weight_single. 5585 5586 scftorture.weight_many= [KNL] 5587 The probability weighting to use for the 5588 smp_call_function_many() function with a zero 5589 "wait" parameter. See weight_single. 5590 Note well that setting a high probability for 5591 this weighting can place serious IPI load 5592 on the system. 5593 5594 scftorture.weight_many_wait= [KNL] 5595 The probability weighting to use for the 5596 smp_call_function_many() function with a 5597 non-zero "wait" parameter. See weight_single 5598 and weight_many. 5599 5600 scftorture.weight_all= [KNL] 5601 The probability weighting to use for the 5602 smp_call_function_all() function with a zero 5603 "wait" parameter. See weight_single and 5604 weight_many. 5605 5606 scftorture.weight_all_wait= [KNL] 5607 The probability weighting to use for the 5608 smp_call_function_all() function with a 5609 non-zero "wait" parameter. See weight_single 5610 and weight_many. 5611 5612 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 5613 xtime_lock contention on larger systems, and/or RCU lock 5614 contention on all systems with CONFIG_MAXSMP set. 5615 Format: { "0" | "1" } 5616 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 5617 1 -- enable. 5618 Note: increases power consumption, thus should only be 5619 enabled if running jitter sensitive (HPC/RT) workloads. 5620 5621 security= [SECURITY] Choose a legacy "major" security module to 5622 enable at boot. This has been deprecated by the 5623 "lsm=" parameter. 5624 5625 selinux= [SELINUX] Disable or enable SELinux at boot time. 5626 Format: { "0" | "1" } 5627 See security/selinux/Kconfig help text. 5628 0 -- disable. 5629 1 -- enable. 5630 Default value is 1. 5631 5632 serialnumber [BUGS=X86-32] 5633 5634 sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 5635 5636 shapers= [NET] 5637 Maximal number of shapers. 5638 5639 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 5640 Limit apic dumping. The parameter defines the maximal 5641 number of local apics being dumped. Also it is possible 5642 to set it to "all" by meaning -- no limit here. 5643 Format: { 1 (default) | 2 | ... | all }. 5644 The parameter valid if only apic=debug or 5645 apic=verbose is specified. 5646 Example: apic=debug show_lapic=all 5647 5648 simeth= [IA-64] 5649 simscsi= 5650 5651 slram= [HW,MTD] 5652 5653 slab_merge [MM] 5654 Enable merging of slabs with similar size when the 5655 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 5656 5657 slab_nomerge [MM] 5658 Disable merging of slabs with similar size. May be 5659 necessary if there is some reason to distinguish 5660 allocs to different slabs, especially in hardened 5661 environments where the risk of heap overflows and 5662 layout control by attackers can usually be 5663 frustrated by disabling merging. This will reduce 5664 most of the exposure of a heap attack to a single 5665 cache (risks via metadata attacks are mostly 5666 unchanged). Debug options disable merging on their 5667 own. 5668 For more information see Documentation/mm/slub.rst. 5669 5670 slab_max_order= [MM, SLAB] 5671 Determines the maximum allowed order for slabs. 5672 A high setting may cause OOMs due to memory 5673 fragmentation. Defaults to 1 for systems with 5674 more than 32MB of RAM, 0 otherwise. 5675 5676 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB] 5677 Enabling slub_debug allows one to determine the 5678 culprit if slab objects become corrupted. Enabling 5679 slub_debug can create guard zones around objects and 5680 may poison objects when not in use. Also tracks the 5681 last alloc / free. For more information see 5682 Documentation/mm/slub.rst. 5683 5684 slub_max_order= [MM, SLUB] 5685 Determines the maximum allowed order for slabs. 5686 A high setting may cause OOMs due to memory 5687 fragmentation. For more information see 5688 Documentation/mm/slub.rst. 5689 5690 slub_min_objects= [MM, SLUB] 5691 The minimum number of objects per slab. SLUB will 5692 increase the slab order up to slub_max_order to 5693 generate a sufficiently large slab able to contain 5694 the number of objects indicated. The higher the number 5695 of objects the smaller the overhead of tracking slabs 5696 and the less frequently locks need to be acquired. 5697 For more information see Documentation/mm/slub.rst. 5698 5699 slub_min_order= [MM, SLUB] 5700 Determines the minimum page order for slabs. Must be 5701 lower than slub_max_order. 5702 For more information see Documentation/mm/slub.rst. 5703 5704 slub_merge [MM, SLUB] 5705 Same with slab_merge. 5706 5707 slub_nomerge [MM, SLUB] 5708 Same with slab_nomerge. This is supported for legacy. 5709 See slab_nomerge for more information. 5710 5711 smart2= [HW] 5712 Format: <io1>[,<io2>[,...,<io8>]] 5713 5714 smp.csd_lock_timeout= [KNL] 5715 Specify the period of time in milliseconds 5716 that smp_call_function() and friends will wait 5717 for a CPU to release the CSD lock. This is 5718 useful when diagnosing bugs involving CPUs 5719 disabling interrupts for extended periods 5720 of time. Defaults to 5,000 milliseconds, and 5721 setting a value of zero disables this feature. 5722 This feature may be more efficiently disabled 5723 using the csdlock_debug- kernel parameter. 5724 5725 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 5726 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 5727 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 5728 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 5729 smsc-ircc2.ircc_irq= [HW] IRQ line 5730 smsc-ircc2.ircc_dma= [HW] DMA channel 5731 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 5732 0: Toshiba Satellite 1800 (GP data pin select) 5733 1: Fast pin select (default) 5734 2: ATC IRMode 5735 5736 smt= [KNL,S390] Set the maximum number of threads (logical 5737 CPUs) to use per physical CPU on systems capable of 5738 symmetric multithreading (SMT). Will be capped to the 5739 actual hardware limit. 5740 Format: <integer> 5741 Default: -1 (no limit) 5742 5743 softlockup_panic= 5744 [KNL] Should the soft-lockup detector generate panics. 5745 Format: 0 | 1 5746 5747 A value of 1 instructs the soft-lockup detector 5748 to panic the machine when a soft-lockup occurs. It is 5749 also controlled by the kernel.softlockup_panic sysctl 5750 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 5751 respective build-time switch to that functionality. 5752 5753 softlockup_all_cpu_backtrace= 5754 [KNL] Should the soft-lockup detector generate 5755 backtraces on all cpus. 5756 Format: 0 | 1 5757 5758 sonypi.*= [HW] Sony Programmable I/O Control Device driver 5759 See Documentation/admin-guide/laptops/sonypi.rst 5760 5761 spectre_v2= [X86] Control mitigation of Spectre variant 2 5762 (indirect branch speculation) vulnerability. 5763 The default operation protects the kernel from 5764 user space attacks. 5765 5766 on - unconditionally enable, implies 5767 spectre_v2_user=on 5768 off - unconditionally disable, implies 5769 spectre_v2_user=off 5770 auto - kernel detects whether your CPU model is 5771 vulnerable 5772 5773 Selecting 'on' will, and 'auto' may, choose a 5774 mitigation method at run time according to the 5775 CPU, the available microcode, the setting of the 5776 CONFIG_RETPOLINE configuration option, and the 5777 compiler with which the kernel was built. 5778 5779 Selecting 'on' will also enable the mitigation 5780 against user space to user space task attacks. 5781 5782 Selecting 'off' will disable both the kernel and 5783 the user space protections. 5784 5785 Specific mitigations can also be selected manually: 5786 5787 retpoline - replace indirect branches 5788 retpoline,generic - Retpolines 5789 retpoline,lfence - LFENCE; indirect branch 5790 retpoline,amd - alias for retpoline,lfence 5791 eibrs - Enhanced/Auto IBRS 5792 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 5793 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 5794 ibrs - use IBRS to protect kernel 5795 5796 Not specifying this option is equivalent to 5797 spectre_v2=auto. 5798 5799 spectre_v2_user= 5800 [X86] Control mitigation of Spectre variant 2 5801 (indirect branch speculation) vulnerability between 5802 user space tasks 5803 5804 on - Unconditionally enable mitigations. Is 5805 enforced by spectre_v2=on 5806 5807 off - Unconditionally disable mitigations. Is 5808 enforced by spectre_v2=off 5809 5810 prctl - Indirect branch speculation is enabled, 5811 but mitigation can be enabled via prctl 5812 per thread. The mitigation control state 5813 is inherited on fork. 5814 5815 prctl,ibpb 5816 - Like "prctl" above, but only STIBP is 5817 controlled per thread. IBPB is issued 5818 always when switching between different user 5819 space processes. 5820 5821 seccomp 5822 - Same as "prctl" above, but all seccomp 5823 threads will enable the mitigation unless 5824 they explicitly opt out. 5825 5826 seccomp,ibpb 5827 - Like "seccomp" above, but only STIBP is 5828 controlled per thread. IBPB is issued 5829 always when switching between different 5830 user space processes. 5831 5832 auto - Kernel selects the mitigation depending on 5833 the available CPU features and vulnerability. 5834 5835 Default mitigation: "prctl" 5836 5837 Not specifying this option is equivalent to 5838 spectre_v2_user=auto. 5839 5840 spec_store_bypass_disable= 5841 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 5842 (Speculative Store Bypass vulnerability) 5843 5844 Certain CPUs are vulnerable to an exploit against a 5845 a common industry wide performance optimization known 5846 as "Speculative Store Bypass" in which recent stores 5847 to the same memory location may not be observed by 5848 later loads during speculative execution. The idea 5849 is that such stores are unlikely and that they can 5850 be detected prior to instruction retirement at the 5851 end of a particular speculation execution window. 5852 5853 In vulnerable processors, the speculatively forwarded 5854 store can be used in a cache side channel attack, for 5855 example to read memory to which the attacker does not 5856 directly have access (e.g. inside sandboxed code). 5857 5858 This parameter controls whether the Speculative Store 5859 Bypass optimization is used. 5860 5861 On x86 the options are: 5862 5863 on - Unconditionally disable Speculative Store Bypass 5864 off - Unconditionally enable Speculative Store Bypass 5865 auto - Kernel detects whether the CPU model contains an 5866 implementation of Speculative Store Bypass and 5867 picks the most appropriate mitigation. If the 5868 CPU is not vulnerable, "off" is selected. If the 5869 CPU is vulnerable the default mitigation is 5870 architecture and Kconfig dependent. See below. 5871 prctl - Control Speculative Store Bypass per thread 5872 via prctl. Speculative Store Bypass is enabled 5873 for a process by default. The state of the control 5874 is inherited on fork. 5875 seccomp - Same as "prctl" above, but all seccomp threads 5876 will disable SSB unless they explicitly opt out. 5877 5878 Default mitigations: 5879 X86: "prctl" 5880 5881 On powerpc the options are: 5882 5883 on,auto - On Power8 and Power9 insert a store-forwarding 5884 barrier on kernel entry and exit. On Power7 5885 perform a software flush on kernel entry and 5886 exit. 5887 off - No action. 5888 5889 Not specifying this option is equivalent to 5890 spec_store_bypass_disable=auto. 5891 5892 spia_io_base= [HW,MTD] 5893 spia_fio_base= 5894 spia_pedr= 5895 spia_peddr= 5896 5897 split_lock_detect= 5898 [X86] Enable split lock detection or bus lock detection 5899 5900 When enabled (and if hardware support is present), atomic 5901 instructions that access data across cache line 5902 boundaries will result in an alignment check exception 5903 for split lock detection or a debug exception for 5904 bus lock detection. 5905 5906 off - not enabled 5907 5908 warn - the kernel will emit rate-limited warnings 5909 about applications triggering the #AC 5910 exception or the #DB exception. This mode is 5911 the default on CPUs that support split lock 5912 detection or bus lock detection. Default 5913 behavior is by #AC if both features are 5914 enabled in hardware. 5915 5916 fatal - the kernel will send SIGBUS to applications 5917 that trigger the #AC exception or the #DB 5918 exception. Default behavior is by #AC if 5919 both features are enabled in hardware. 5920 5921 ratelimit:N - 5922 Set system wide rate limit to N bus locks 5923 per second for bus lock detection. 5924 0 < N <= 1000. 5925 5926 N/A for split lock detection. 5927 5928 5929 If an #AC exception is hit in the kernel or in 5930 firmware (i.e. not while executing in user mode) 5931 the kernel will oops in either "warn" or "fatal" 5932 mode. 5933 5934 #DB exception for bus lock is triggered only when 5935 CPL > 0. 5936 5937 srbds= [X86,INTEL] 5938 Control the Special Register Buffer Data Sampling 5939 (SRBDS) mitigation. 5940 5941 Certain CPUs are vulnerable to an MDS-like 5942 exploit which can leak bits from the random 5943 number generator. 5944 5945 By default, this issue is mitigated by 5946 microcode. However, the microcode fix can cause 5947 the RDRAND and RDSEED instructions to become 5948 much slower. Among other effects, this will 5949 result in reduced throughput from /dev/urandom. 5950 5951 The microcode mitigation can be disabled with 5952 the following option: 5953 5954 off: Disable mitigation and remove 5955 performance impact to RDRAND and RDSEED 5956 5957 srcutree.big_cpu_lim [KNL] 5958 Specifies the number of CPUs constituting a 5959 large system, such that srcu_struct structures 5960 should immediately allocate an srcu_node array. 5961 This kernel-boot parameter defaults to 128, 5962 but takes effect only when the low-order four 5963 bits of srcutree.convert_to_big is equal to 3 5964 (decide at boot). 5965 5966 srcutree.convert_to_big [KNL] 5967 Specifies under what conditions an SRCU tree 5968 srcu_struct structure will be converted to big 5969 form, that is, with an rcu_node tree: 5970 5971 0: Never. 5972 1: At init_srcu_struct() time. 5973 2: When rcutorture decides to. 5974 3: Decide at boot time (default). 5975 0x1X: Above plus if high contention. 5976 5977 Either way, the srcu_node tree will be sized based 5978 on the actual runtime number of CPUs (nr_cpu_ids) 5979 instead of the compile-time CONFIG_NR_CPUS. 5980 5981 srcutree.counter_wrap_check [KNL] 5982 Specifies how frequently to check for 5983 grace-period sequence counter wrap for the 5984 srcu_data structure's ->srcu_gp_seq_needed field. 5985 The greater the number of bits set in this kernel 5986 parameter, the less frequently counter wrap will 5987 be checked for. Note that the bottom two bits 5988 are ignored. 5989 5990 srcutree.exp_holdoff [KNL] 5991 Specifies how many nanoseconds must elapse 5992 since the end of the last SRCU grace period for 5993 a given srcu_struct until the next normal SRCU 5994 grace period will be considered for automatic 5995 expediting. Set to zero to disable automatic 5996 expediting. 5997 5998 srcutree.srcu_max_nodelay [KNL] 5999 Specifies the number of no-delay instances 6000 per jiffy for which the SRCU grace period 6001 worker thread will be rescheduled with zero 6002 delay. Beyond this limit, worker thread will 6003 be rescheduled with a sleep delay of one jiffy. 6004 6005 srcutree.srcu_max_nodelay_phase [KNL] 6006 Specifies the per-grace-period phase, number of 6007 non-sleeping polls of readers. Beyond this limit, 6008 grace period worker thread will be rescheduled 6009 with a sleep delay of one jiffy, between each 6010 rescan of the readers, for a grace period phase. 6011 6012 srcutree.srcu_retry_check_delay [KNL] 6013 Specifies number of microseconds of non-sleeping 6014 delay between each non-sleeping poll of readers. 6015 6016 srcutree.small_contention_lim [KNL] 6017 Specifies the number of update-side contention 6018 events per jiffy will be tolerated before 6019 initiating a conversion of an srcu_struct 6020 structure to big form. Note that the value of 6021 srcutree.convert_to_big must have the 0x10 bit 6022 set for contention-based conversions to occur. 6023 6024 ssbd= [ARM64,HW] 6025 Speculative Store Bypass Disable control 6026 6027 On CPUs that are vulnerable to the Speculative 6028 Store Bypass vulnerability and offer a 6029 firmware based mitigation, this parameter 6030 indicates how the mitigation should be used: 6031 6032 force-on: Unconditionally enable mitigation for 6033 for both kernel and userspace 6034 force-off: Unconditionally disable mitigation for 6035 for both kernel and userspace 6036 kernel: Always enable mitigation in the 6037 kernel, and offer a prctl interface 6038 to allow userspace to register its 6039 interest in being mitigated too. 6040 6041 stack_guard_gap= [MM] 6042 override the default stack gap protection. The value 6043 is in page units and it defines how many pages prior 6044 to (for stacks growing down) resp. after (for stacks 6045 growing up) the main stack are reserved for no other 6046 mapping. Default value is 256 pages. 6047 6048 stack_depot_disable= [KNL] 6049 Setting this to true through kernel command line will 6050 disable the stack depot thereby saving the static memory 6051 consumed by the stack hash table. By default this is set 6052 to false. 6053 6054 stacktrace [FTRACE] 6055 Enabled the stack tracer on boot up. 6056 6057 stacktrace_filter=[function-list] 6058 [FTRACE] Limit the functions that the stack tracer 6059 will trace at boot up. function-list is a comma-separated 6060 list of functions. This list can be changed at run 6061 time by the stack_trace_filter file in the debugfs 6062 tracing directory. Note, this enables stack tracing 6063 and the stacktrace above is not needed. 6064 6065 sti= [PARISC,HW] 6066 Format: <num> 6067 Set the STI (builtin display/keyboard on the HP-PARISC 6068 machines) console (graphic card) which should be used 6069 as the initial boot-console. 6070 See also comment in drivers/video/console/sticore.c. 6071 6072 sti_font= [HW] 6073 See comment in drivers/video/console/sticore.c. 6074 6075 stifb= [HW] 6076 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 6077 6078 strict_sas_size= 6079 [X86] 6080 Format: <bool> 6081 Enable or disable strict sigaltstack size checks 6082 against the required signal frame size which 6083 depends on the supported FPU features. This can 6084 be used to filter out binaries which have 6085 not yet been made aware of AT_MINSIGSTKSZ. 6086 6087 stress_hpt [PPC] 6088 Limits the number of kernel HPT entries in the hash 6089 page table to increase the rate of hash page table 6090 faults on kernel addresses. 6091 6092 stress_slb [PPC] 6093 Limits the number of kernel SLB entries, and flushes 6094 them frequently to increase the rate of SLB faults 6095 on kernel addresses. 6096 6097 sunrpc.min_resvport= 6098 sunrpc.max_resvport= 6099 [NFS,SUNRPC] 6100 SunRPC servers often require that client requests 6101 originate from a privileged port (i.e. a port in the 6102 range 0 < portnr < 1024). 6103 An administrator who wishes to reserve some of these 6104 ports for other uses may adjust the range that the 6105 kernel's sunrpc client considers to be privileged 6106 using these two parameters to set the minimum and 6107 maximum port values. 6108 6109 sunrpc.svc_rpc_per_connection_limit= 6110 [NFS,SUNRPC] 6111 Limit the number of requests that the server will 6112 process in parallel from a single connection. 6113 The default value is 0 (no limit). 6114 6115 sunrpc.pool_mode= 6116 [NFS] 6117 Control how the NFS server code allocates CPUs to 6118 service thread pools. Depending on how many NICs 6119 you have and where their interrupts are bound, this 6120 option will affect which CPUs will do NFS serving. 6121 Note: this parameter cannot be changed while the 6122 NFS server is running. 6123 6124 auto the server chooses an appropriate mode 6125 automatically using heuristics 6126 global a single global pool contains all CPUs 6127 percpu one pool for each CPU 6128 pernode one pool for each NUMA node (equivalent 6129 to global on non-NUMA machines) 6130 6131 sunrpc.tcp_slot_table_entries= 6132 sunrpc.udp_slot_table_entries= 6133 [NFS,SUNRPC] 6134 Sets the upper limit on the number of simultaneous 6135 RPC calls that can be sent from the client to a 6136 server. Increasing these values may allow you to 6137 improve throughput, but will also increase the 6138 amount of memory reserved for use by the client. 6139 6140 suspend.pm_test_delay= 6141 [SUSPEND] 6142 Sets the number of seconds to remain in a suspend test 6143 mode before resuming the system (see 6144 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 6145 is set. Default value is 5. 6146 6147 svm= [PPC] 6148 Format: { on | off | y | n | 1 | 0 } 6149 This parameter controls use of the Protected 6150 Execution Facility on pSeries. 6151 6152 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 6153 Format: { <int> [,<int>] | force | noforce } 6154 <int> -- Number of I/O TLB slabs 6155 <int> -- Second integer after comma. Number of swiotlb 6156 areas with their own lock. Will be rounded up 6157 to a power of 2. 6158 force -- force using of bounce buffers even if they 6159 wouldn't be automatically used by the kernel 6160 noforce -- Never use bounce buffers (for debugging) 6161 6162 switches= [HW,M68k] 6163 6164 sysctl.*= [KNL] 6165 Set a sysctl parameter, right before loading the init 6166 process, as if the value was written to the respective 6167 /proc/sys/... file. Both '.' and '/' are recognized as 6168 separators. Unrecognized parameters and invalid values 6169 are reported in the kernel log. Sysctls registered 6170 later by a loaded module cannot be set this way. 6171 Example: sysctl.vm.swappiness=40 6172 6173 sysrq_always_enabled 6174 [KNL] 6175 Ignore sysrq setting - this boot parameter will 6176 neutralize any effect of /proc/sys/kernel/sysrq. 6177 Useful for debugging. 6178 6179 tcpmhash_entries= [KNL,NET] 6180 Set the number of tcp_metrics_hash slots. 6181 Default value is 8192 or 16384 depending on total 6182 ram pages. This is used to specify the TCP metrics 6183 cache size. See Documentation/networking/ip-sysctl.rst 6184 "tcp_no_metrics_save" section for more details. 6185 6186 tdfx= [HW,DRM] 6187 6188 test_suspend= [SUSPEND] 6189 Format: { "mem" | "standby" | "freeze" }[,N] 6190 Specify "mem" (for Suspend-to-RAM) or "standby" (for 6191 standby suspend) or "freeze" (for suspend type freeze) 6192 as the system sleep state during system startup with 6193 the optional capability to repeat N number of times. 6194 The system is woken from this state using a 6195 wakeup-capable RTC alarm. 6196 6197 thash_entries= [KNL,NET] 6198 Set number of hash buckets for TCP connection 6199 6200 thermal.act= [HW,ACPI] 6201 -1: disable all active trip points in all thermal zones 6202 <degrees C>: override all lowest active trip points 6203 6204 thermal.crt= [HW,ACPI] 6205 -1: disable all critical trip points in all thermal zones 6206 <degrees C>: override all critical trip points 6207 6208 thermal.nocrt= [HW,ACPI] 6209 Set to disable actions on ACPI thermal zone 6210 critical and hot trip points. 6211 6212 thermal.off= [HW,ACPI] 6213 1: disable ACPI thermal control 6214 6215 thermal.psv= [HW,ACPI] 6216 -1: disable all passive trip points 6217 <degrees C>: override all passive trip points to this 6218 value 6219 6220 thermal.tzp= [HW,ACPI] 6221 Specify global default ACPI thermal zone polling rate 6222 <deci-seconds>: poll all this frequency 6223 0: no polling (default) 6224 6225 threadirqs [KNL] 6226 Force threading of all interrupt handlers except those 6227 marked explicitly IRQF_NO_THREAD. 6228 6229 topology= [S390] 6230 Format: {off | on} 6231 Specify if the kernel should make use of the cpu 6232 topology information if the hardware supports this. 6233 The scheduler will make use of this information and 6234 e.g. base its process migration decisions on it. 6235 Default is on. 6236 6237 topology_updates= [KNL, PPC, NUMA] 6238 Format: {off} 6239 Specify if the kernel should ignore (off) 6240 topology updates sent by the hypervisor to this 6241 LPAR. 6242 6243 torture.disable_onoff_at_boot= [KNL] 6244 Prevent the CPU-hotplug component of torturing 6245 until after init has spawned. 6246 6247 torture.ftrace_dump_at_shutdown= [KNL] 6248 Dump the ftrace buffer at torture-test shutdown, 6249 even if there were no errors. This can be a 6250 very costly operation when many torture tests 6251 are running concurrently, especially on systems 6252 with rotating-rust storage. 6253 6254 torture.verbose_sleep_frequency= [KNL] 6255 Specifies how many verbose printk()s should be 6256 emitted between each sleep. The default of zero 6257 disables verbose-printk() sleeping. 6258 6259 torture.verbose_sleep_duration= [KNL] 6260 Duration of each verbose-printk() sleep in jiffies. 6261 6262 tp720= [HW,PS2] 6263 6264 tpm_suspend_pcr=[HW,TPM] 6265 Format: integer pcr id 6266 Specify that at suspend time, the tpm driver 6267 should extend the specified pcr with zeros, 6268 as a workaround for some chips which fail to 6269 flush the last written pcr on TPM_SaveState. 6270 This will guarantee that all the other pcrs 6271 are saved. 6272 6273 tp_printk [FTRACE] 6274 Have the tracepoints sent to printk as well as the 6275 tracing ring buffer. This is useful for early boot up 6276 where the system hangs or reboots and does not give the 6277 option for reading the tracing buffer or performing a 6278 ftrace_dump_on_oops. 6279 6280 To turn off having tracepoints sent to printk, 6281 echo 0 > /proc/sys/kernel/tracepoint_printk 6282 Note, echoing 1 into this file without the 6283 tracepoint_printk kernel cmdline option has no effect. 6284 6285 The tp_printk_stop_on_boot (see below) can also be used 6286 to stop the printing of events to console at 6287 late_initcall_sync. 6288 6289 ** CAUTION ** 6290 6291 Having tracepoints sent to printk() and activating high 6292 frequency tracepoints such as irq or sched, can cause 6293 the system to live lock. 6294 6295 tp_printk_stop_on_boot [FTRACE] 6296 When tp_printk (above) is set, it can cause a lot of noise 6297 on the console. It may be useful to only include the 6298 printing of events during boot up, as user space may 6299 make the system inoperable. 6300 6301 This command line option will stop the printing of events 6302 to console at the late_initcall_sync() time frame. 6303 6304 trace_buf_size=nn[KMG] 6305 [FTRACE] will set tracing buffer size on each cpu. 6306 6307 trace_clock= [FTRACE] Set the clock used for tracing events 6308 at boot up. 6309 local - Use the per CPU time stamp counter 6310 (converted into nanoseconds). Fast, but 6311 depending on the architecture, may not be 6312 in sync between CPUs. 6313 global - Event time stamps are synchronize across 6314 CPUs. May be slower than the local clock, 6315 but better for some race conditions. 6316 counter - Simple counting of events (1, 2, ..) 6317 note, some counts may be skipped due to the 6318 infrastructure grabbing the clock more than 6319 once per event. 6320 uptime - Use jiffies as the time stamp. 6321 perf - Use the same clock that perf uses. 6322 mono - Use ktime_get_mono_fast_ns() for time stamps. 6323 mono_raw - Use ktime_get_raw_fast_ns() for time 6324 stamps. 6325 boot - Use ktime_get_boot_fast_ns() for time stamps. 6326 Architectures may add more clocks. See 6327 Documentation/trace/ftrace.rst for more details. 6328 6329 trace_event=[event-list] 6330 [FTRACE] Set and start specified trace events in order 6331 to facilitate early boot debugging. The event-list is a 6332 comma-separated list of trace events to enable. See 6333 also Documentation/trace/events.rst 6334 6335 trace_instance=[instance-info] 6336 [FTRACE] Create a ring buffer instance early in boot up. 6337 This will be listed in: 6338 6339 /sys/kernel/tracing/instances 6340 6341 Events can be enabled at the time the instance is created 6342 via: 6343 6344 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 6345 6346 Note, the "<system*>:" portion is optional if the event is 6347 unique. 6348 6349 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 6350 6351 will enable the "sched_switch" event (note, the "sched:" is optional, and 6352 the same thing would happen if it was left off). The irq_handler_entry 6353 event, and all events under the "initcall" system. 6354 6355 trace_options=[option-list] 6356 [FTRACE] Enable or disable tracer options at boot. 6357 The option-list is a comma delimited list of options 6358 that can be enabled or disabled just as if you were 6359 to echo the option name into 6360 6361 /sys/kernel/tracing/trace_options 6362 6363 For example, to enable stacktrace option (to dump the 6364 stack trace of each event), add to the command line: 6365 6366 trace_options=stacktrace 6367 6368 See also Documentation/trace/ftrace.rst "trace options" 6369 section. 6370 6371 trace_trigger=[trigger-list] 6372 [FTRACE] Add a event trigger on specific events. 6373 Set a trigger on top of a specific event, with an optional 6374 filter. 6375 6376 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..." 6377 Where more than one trigger may be specified that are comma deliminated. 6378 6379 For example: 6380 6381 trace_trigger="sched_switch.stacktrace if prev_state == 2" 6382 6383 The above will enable the "stacktrace" trigger on the "sched_switch" 6384 event but only trigger it if the "prev_state" of the "sched_switch" 6385 event is "2" (TASK_UNINTERUPTIBLE). 6386 6387 See also "Event triggers" in Documentation/trace/events.rst 6388 6389 6390 traceoff_on_warning 6391 [FTRACE] enable this option to disable tracing when a 6392 warning is hit. This turns off "tracing_on". Tracing can 6393 be enabled again by echoing '1' into the "tracing_on" 6394 file located in /sys/kernel/tracing/ 6395 6396 This option is useful, as it disables the trace before 6397 the WARNING dump is called, which prevents the trace to 6398 be filled with content caused by the warning output. 6399 6400 This option can also be set at run time via the sysctl 6401 option: kernel/traceoff_on_warning 6402 6403 transparent_hugepage= 6404 [KNL] 6405 Format: [always|madvise|never] 6406 Can be used to control the default behavior of the system 6407 with respect to transparent hugepages. 6408 See Documentation/admin-guide/mm/transhuge.rst 6409 for more details. 6410 6411 trusted.source= [KEYS] 6412 Format: <string> 6413 This parameter identifies the trust source as a backend 6414 for trusted keys implementation. Supported trust 6415 sources: 6416 - "tpm" 6417 - "tee" 6418 - "caam" 6419 If not specified then it defaults to iterating through 6420 the trust source list starting with TPM and assigns the 6421 first trust source as a backend which is initialized 6422 successfully during iteration. 6423 6424 trusted.rng= [KEYS] 6425 Format: <string> 6426 The RNG used to generate key material for trusted keys. 6427 Can be one of: 6428 - "kernel" 6429 - the same value as trusted.source: "tpm" or "tee" 6430 - "default" 6431 If not specified, "default" is used. In this case, 6432 the RNG's choice is left to each individual trust source. 6433 6434 tsc= Disable clocksource stability checks for TSC. 6435 Format: <string> 6436 [x86] reliable: mark tsc clocksource as reliable, this 6437 disables clocksource verification at runtime, as well 6438 as the stability checks done at bootup. Used to enable 6439 high-resolution timer mode on older hardware, and in 6440 virtualized environment. 6441 [x86] noirqtime: Do not use TSC to do irq accounting. 6442 Used to run time disable IRQ_TIME_ACCOUNTING on any 6443 platforms where RDTSC is slow and this accounting 6444 can add overhead. 6445 [x86] unstable: mark the TSC clocksource as unstable, this 6446 marks the TSC unconditionally unstable at bootup and 6447 avoids any further wobbles once the TSC watchdog notices. 6448 [x86] nowatchdog: disable clocksource watchdog. Used 6449 in situations with strict latency requirements (where 6450 interruptions from clocksource watchdog are not 6451 acceptable). 6452 [x86] recalibrate: force recalibration against a HW timer 6453 (HPET or PM timer) on systems whose TSC frequency was 6454 obtained from HW or FW using either an MSR or CPUID(0x15). 6455 Warn if the difference is more than 500 ppm. 6456 [x86] watchdog: Use TSC as the watchdog clocksource with 6457 which to check other HW timers (HPET or PM timer), but 6458 only on systems where TSC has been deemed trustworthy. 6459 This will be suppressed by an earlier tsc=nowatchdog and 6460 can be overridden by a later tsc=nowatchdog. A console 6461 message will flag any such suppression or overriding. 6462 6463 tsc_early_khz= [X86] Skip early TSC calibration and use the given 6464 value instead. Useful when the early TSC frequency discovery 6465 procedure is not reliable, such as on overclocked systems 6466 with CPUID.16h support and partial CPUID.15h support. 6467 Format: <unsigned int> 6468 6469 tsx= [X86] Control Transactional Synchronization 6470 Extensions (TSX) feature in Intel processors that 6471 support TSX control. 6472 6473 This parameter controls the TSX feature. The options are: 6474 6475 on - Enable TSX on the system. Although there are 6476 mitigations for all known security vulnerabilities, 6477 TSX has been known to be an accelerator for 6478 several previous speculation-related CVEs, and 6479 so there may be unknown security risks associated 6480 with leaving it enabled. 6481 6482 off - Disable TSX on the system. (Note that this 6483 option takes effect only on newer CPUs which are 6484 not vulnerable to MDS, i.e., have 6485 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 6486 the new IA32_TSX_CTRL MSR through a microcode 6487 update. This new MSR allows for the reliable 6488 deactivation of the TSX functionality.) 6489 6490 auto - Disable TSX if X86_BUG_TAA is present, 6491 otherwise enable TSX on the system. 6492 6493 Not specifying this option is equivalent to tsx=off. 6494 6495 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6496 for more details. 6497 6498 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 6499 Abort (TAA) vulnerability. 6500 6501 Similar to Micro-architectural Data Sampling (MDS) 6502 certain CPUs that support Transactional 6503 Synchronization Extensions (TSX) are vulnerable to an 6504 exploit against CPU internal buffers which can forward 6505 information to a disclosure gadget under certain 6506 conditions. 6507 6508 In vulnerable processors, the speculatively forwarded 6509 data can be used in a cache side channel attack, to 6510 access data to which the attacker does not have direct 6511 access. 6512 6513 This parameter controls the TAA mitigation. The 6514 options are: 6515 6516 full - Enable TAA mitigation on vulnerable CPUs 6517 if TSX is enabled. 6518 6519 full,nosmt - Enable TAA mitigation and disable SMT on 6520 vulnerable CPUs. If TSX is disabled, SMT 6521 is not disabled because CPU is not 6522 vulnerable to cross-thread TAA attacks. 6523 off - Unconditionally disable TAA mitigation 6524 6525 On MDS-affected machines, tsx_async_abort=off can be 6526 prevented by an active MDS mitigation as both vulnerabilities 6527 are mitigated with the same mechanism so in order to disable 6528 this mitigation, you need to specify mds=off too. 6529 6530 Not specifying this option is equivalent to 6531 tsx_async_abort=full. On CPUs which are MDS affected 6532 and deploy MDS mitigation, TAA mitigation is not 6533 required and doesn't provide any additional 6534 mitigation. 6535 6536 For details see: 6537 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6538 6539 turbografx.map[2|3]= [HW,JOY] 6540 TurboGraFX parallel port interface 6541 Format: 6542 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 6543 See also Documentation/input/devices/joystick-parport.rst 6544 6545 udbg-immortal [PPC] When debugging early kernel crashes that 6546 happen after console_init() and before a proper 6547 console driver takes over, this boot options might 6548 help "seeing" what's going on. 6549 6550 uhash_entries= [KNL,NET] 6551 Set number of hash buckets for UDP/UDP-Lite connections 6552 6553 uhci-hcd.ignore_oc= 6554 [USB] Ignore overcurrent events (default N). 6555 Some badly-designed motherboards generate lots of 6556 bogus events, for ports that aren't wired to 6557 anything. Set this parameter to avoid log spamming. 6558 Note that genuine overcurrent events won't be 6559 reported either. 6560 6561 unknown_nmi_panic 6562 [X86] Cause panic on unknown NMI. 6563 6564 usbcore.authorized_default= 6565 [USB] Default USB device authorization: 6566 (default -1 = authorized except for wireless USB, 6567 0 = not authorized, 1 = authorized, 2 = authorized 6568 if device connected to internal port) 6569 6570 usbcore.autosuspend= 6571 [USB] The autosuspend time delay (in seconds) used 6572 for newly-detected USB devices (default 2). This 6573 is the time required before an idle device will be 6574 autosuspended. Devices for which the delay is set 6575 to a negative value won't be autosuspended at all. 6576 6577 usbcore.usbfs_snoop= 6578 [USB] Set to log all usbfs traffic (default 0 = off). 6579 6580 usbcore.usbfs_snoop_max= 6581 [USB] Maximum number of bytes to snoop in each URB 6582 (default = 65536). 6583 6584 usbcore.blinkenlights= 6585 [USB] Set to cycle leds on hubs (default 0 = off). 6586 6587 usbcore.old_scheme_first= 6588 [USB] Start with the old device initialization 6589 scheme (default 0 = off). 6590 6591 usbcore.usbfs_memory_mb= 6592 [USB] Memory limit (in MB) for buffers allocated by 6593 usbfs (default = 16, 0 = max = 2047). 6594 6595 usbcore.use_both_schemes= 6596 [USB] Try the other device initialization scheme 6597 if the first one fails (default 1 = enabled). 6598 6599 usbcore.initial_descriptor_timeout= 6600 [USB] Specifies timeout for the initial 64-byte 6601 USB_REQ_GET_DESCRIPTOR request in milliseconds 6602 (default 5000 = 5.0 seconds). 6603 6604 usbcore.nousb [USB] Disable the USB subsystem 6605 6606 usbcore.quirks= 6607 [USB] A list of quirk entries to augment the built-in 6608 usb core quirk list. List entries are separated by 6609 commas. Each entry has the form 6610 VendorID:ProductID:Flags. The IDs are 4-digit hex 6611 numbers and Flags is a set of letters. Each letter 6612 will change the built-in quirk; setting it if it is 6613 clear and clearing it if it is set. The letters have 6614 the following meanings: 6615 a = USB_QUIRK_STRING_FETCH_255 (string 6616 descriptors must not be fetched using 6617 a 255-byte read); 6618 b = USB_QUIRK_RESET_RESUME (device can't resume 6619 correctly so reset it instead); 6620 c = USB_QUIRK_NO_SET_INTF (device can't handle 6621 Set-Interface requests); 6622 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 6623 handle its Configuration or Interface 6624 strings); 6625 e = USB_QUIRK_RESET (device can't be reset 6626 (e.g morph devices), don't use reset); 6627 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 6628 more interface descriptions than the 6629 bNumInterfaces count, and can't handle 6630 talking to these interfaces); 6631 g = USB_QUIRK_DELAY_INIT (device needs a pause 6632 during initialization, after we read 6633 the device descriptor); 6634 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 6635 high speed and super speed interrupt 6636 endpoints, the USB 2.0 and USB 3.0 spec 6637 require the interval in microframes (1 6638 microframe = 125 microseconds) to be 6639 calculated as interval = 2 ^ 6640 (bInterval-1). 6641 Devices with this quirk report their 6642 bInterval as the result of this 6643 calculation instead of the exponent 6644 variable used in the calculation); 6645 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 6646 handle device_qualifier descriptor 6647 requests); 6648 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 6649 generates spurious wakeup, ignore 6650 remote wakeup capability); 6651 k = USB_QUIRK_NO_LPM (device can't handle Link 6652 Power Management); 6653 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 6654 (Device reports its bInterval as linear 6655 frames instead of the USB 2.0 6656 calculation); 6657 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 6658 to be disconnected before suspend to 6659 prevent spurious wakeup); 6660 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 6661 pause after every control message); 6662 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 6663 delay after resetting its port); 6664 Example: quirks=0781:5580:bk,0a5c:5834:gij 6665 6666 usbhid.mousepoll= 6667 [USBHID] The interval which mice are to be polled at. 6668 6669 usbhid.jspoll= 6670 [USBHID] The interval which joysticks are to be polled at. 6671 6672 usbhid.kbpoll= 6673 [USBHID] The interval which keyboards are to be polled at. 6674 6675 usb-storage.delay_use= 6676 [UMS] The delay in seconds before a new device is 6677 scanned for Logical Units (default 1). 6678 6679 usb-storage.quirks= 6680 [UMS] A list of quirks entries to supplement or 6681 override the built-in unusual_devs list. List 6682 entries are separated by commas. Each entry has 6683 the form VID:PID:Flags where VID and PID are Vendor 6684 and Product ID values (4-digit hex numbers) and 6685 Flags is a set of characters, each corresponding 6686 to a common usb-storage quirk flag as follows: 6687 a = SANE_SENSE (collect more than 18 bytes 6688 of sense data, not on uas); 6689 b = BAD_SENSE (don't collect more than 18 6690 bytes of sense data, not on uas); 6691 c = FIX_CAPACITY (decrease the reported 6692 device capacity by one sector); 6693 d = NO_READ_DISC_INFO (don't use 6694 READ_DISC_INFO command, not on uas); 6695 e = NO_READ_CAPACITY_16 (don't use 6696 READ_CAPACITY_16 command); 6697 f = NO_REPORT_OPCODES (don't use report opcodes 6698 command, uas only); 6699 g = MAX_SECTORS_240 (don't transfer more than 6700 240 sectors at a time, uas only); 6701 h = CAPACITY_HEURISTICS (decrease the 6702 reported device capacity by one 6703 sector if the number is odd); 6704 i = IGNORE_DEVICE (don't bind to this 6705 device); 6706 j = NO_REPORT_LUNS (don't use report luns 6707 command, uas only); 6708 k = NO_SAME (do not use WRITE_SAME, uas only) 6709 l = NOT_LOCKABLE (don't try to lock and 6710 unlock ejectable media, not on uas); 6711 m = MAX_SECTORS_64 (don't transfer more 6712 than 64 sectors = 32 KB at a time, 6713 not on uas); 6714 n = INITIAL_READ10 (force a retry of the 6715 initial READ(10) command, not on uas); 6716 o = CAPACITY_OK (accept the capacity 6717 reported by the device, not on uas); 6718 p = WRITE_CACHE (the device cache is ON 6719 by default, not on uas); 6720 r = IGNORE_RESIDUE (the device reports 6721 bogus residue values, not on uas); 6722 s = SINGLE_LUN (the device has only one 6723 Logical Unit); 6724 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 6725 commands, uas only); 6726 u = IGNORE_UAS (don't bind to the uas driver); 6727 w = NO_WP_DETECT (don't test whether the 6728 medium is write-protected). 6729 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 6730 even if the device claims no cache, 6731 not on uas) 6732 Example: quirks=0419:aaf5:rl,0421:0433:rc 6733 6734 user_debug= [KNL,ARM] 6735 Format: <int> 6736 See arch/arm/Kconfig.debug help text. 6737 1 - undefined instruction events 6738 2 - system calls 6739 4 - invalid data aborts 6740 8 - SIGSEGV faults 6741 16 - SIGBUS faults 6742 Example: user_debug=31 6743 6744 userpte= 6745 [X86] Flags controlling user PTE allocations. 6746 6747 nohigh = do not allocate PTE pages in 6748 HIGHMEM regardless of setting 6749 of CONFIG_HIGHPTE. 6750 6751 vdso= [X86,SH,SPARC] 6752 On X86_32, this is an alias for vdso32=. Otherwise: 6753 6754 vdso=1: enable VDSO (the default) 6755 vdso=0: disable VDSO mapping 6756 6757 vdso32= [X86] Control the 32-bit vDSO 6758 vdso32=1: enable 32-bit VDSO 6759 vdso32=0 or vdso32=2: disable 32-bit VDSO 6760 6761 See the help text for CONFIG_COMPAT_VDSO for more 6762 details. If CONFIG_COMPAT_VDSO is set, the default is 6763 vdso32=0; otherwise, the default is vdso32=1. 6764 6765 For compatibility with older kernels, vdso32=2 is an 6766 alias for vdso32=0. 6767 6768 Try vdso32=0 if you encounter an error that says: 6769 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 6770 6771 vector= [IA-64,SMP] 6772 vector=percpu: enable percpu vector domain 6773 6774 video= [FB] Frame buffer configuration 6775 See Documentation/fb/modedb.rst. 6776 6777 video.brightness_switch_enabled= [ACPI] 6778 Format: [0|1] 6779 If set to 1, on receiving an ACPI notify event 6780 generated by hotkey, video driver will adjust brightness 6781 level and then send out the event to user space through 6782 the allocated input device. If set to 0, video driver 6783 will only send out the event without touching backlight 6784 brightness level. 6785 default: 1 6786 6787 virtio_mmio.device= 6788 [VMMIO] Memory mapped virtio (platform) device. 6789 6790 <size>@<baseaddr>:<irq>[:<id>] 6791 where: 6792 <size> := size (can use standard suffixes 6793 like K, M and G) 6794 <baseaddr> := physical base address 6795 <irq> := interrupt number (as passed to 6796 request_irq()) 6797 <id> := (optional) platform device id 6798 example: 6799 virtio_mmio.device=1K@0x100b0000:48:7 6800 6801 Can be used multiple times for multiple devices. 6802 6803 vga= [BOOT,X86-32] Select a particular video mode 6804 See Documentation/arch/x86/boot.rst and 6805 Documentation/admin-guide/svga.rst. 6806 Use vga=ask for menu. 6807 This is actually a boot loader parameter; the value is 6808 passed to the kernel using a special protocol. 6809 6810 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 6811 May slow down system boot speed, especially when 6812 enabled on systems with a large amount of memory. 6813 All options are enabled by default, and this 6814 interface is meant to allow for selectively 6815 enabling or disabling specific virtual memory 6816 debugging features. 6817 6818 Available options are: 6819 P Enable page structure init time poisoning 6820 - Disable all of the above options 6821 6822 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 6823 size of <nn>. This can be used to increase the 6824 minimum size (128MB on x86). It can also be used to 6825 decrease the size and leave more room for directly 6826 mapped kernel RAM. 6827 6828 vmcp_cma=nn[MG] [KNL,S390] 6829 Sets the memory size reserved for contiguous memory 6830 allocations for the vmcp device driver. 6831 6832 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 6833 Format: <command> 6834 6835 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 6836 Format: <command> 6837 6838 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 6839 Format: <command> 6840 6841 vsyscall= [X86-64] 6842 Controls the behavior of vsyscalls (i.e. calls to 6843 fixed addresses of 0xffffffffff600x00 from legacy 6844 code). Most statically-linked binaries and older 6845 versions of glibc use these calls. Because these 6846 functions are at fixed addresses, they make nice 6847 targets for exploits that can control RIP. 6848 6849 emulate Vsyscalls turn into traps and are emulated 6850 reasonably safely. The vsyscall page is 6851 readable. 6852 6853 xonly [default] Vsyscalls turn into traps and are 6854 emulated reasonably safely. The vsyscall 6855 page is not readable. 6856 6857 none Vsyscalls don't work at all. This makes 6858 them quite hard to use for exploits but 6859 might break your system. 6860 6861 vt.color= [VT] Default text color. 6862 Format: 0xYX, X = foreground, Y = background. 6863 Default: 0x07 = light gray on black. 6864 6865 vt.cur_default= [VT] Default cursor shape. 6866 Format: 0xCCBBAA, where AA, BB, and CC are the same as 6867 the parameters of the <Esc>[?A;B;Cc escape sequence; 6868 see VGA-softcursor.txt. Default: 2 = underline. 6869 6870 vt.default_blu= [VT] 6871 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 6872 Change the default blue palette of the console. 6873 This is a 16-member array composed of values 6874 ranging from 0-255. 6875 6876 vt.default_grn= [VT] 6877 Format: <green0>,<green1>,<green2>,...,<green15> 6878 Change the default green palette of the console. 6879 This is a 16-member array composed of values 6880 ranging from 0-255. 6881 6882 vt.default_red= [VT] 6883 Format: <red0>,<red1>,<red2>,...,<red15> 6884 Change the default red palette of the console. 6885 This is a 16-member array composed of values 6886 ranging from 0-255. 6887 6888 vt.default_utf8= 6889 [VT] 6890 Format=<0|1> 6891 Set system-wide default UTF-8 mode for all tty's. 6892 Default is 1, i.e. UTF-8 mode is enabled for all 6893 newly opened terminals. 6894 6895 vt.global_cursor_default= 6896 [VT] 6897 Format=<-1|0|1> 6898 Set system-wide default for whether a cursor 6899 is shown on new VTs. Default is -1, 6900 i.e. cursors will be created by default unless 6901 overridden by individual drivers. 0 will hide 6902 cursors, 1 will display them. 6903 6904 vt.italic= [VT] Default color for italic text; 0-15. 6905 Default: 2 = green. 6906 6907 vt.underline= [VT] Default color for underlined text; 0-15. 6908 Default: 3 = cyan. 6909 6910 watchdog timers [HW,WDT] For information on watchdog timers, 6911 see Documentation/watchdog/watchdog-parameters.rst 6912 or other driver-specific files in the 6913 Documentation/watchdog/ directory. 6914 6915 watchdog_thresh= 6916 [KNL] 6917 Set the hard lockup detector stall duration 6918 threshold in seconds. The soft lockup detector 6919 threshold is set to twice the value. A value of 0 6920 disables both lockup detectors. Default is 10 6921 seconds. 6922 6923 workqueue.watchdog_thresh= 6924 If CONFIG_WQ_WATCHDOG is configured, workqueue can 6925 warn stall conditions and dump internal state to 6926 help debugging. 0 disables workqueue stall 6927 detection; otherwise, it's the stall threshold 6928 duration in seconds. The default value is 30 and 6929 it can be updated at runtime by writing to the 6930 corresponding sysfs file. 6931 6932 workqueue.disable_numa 6933 By default, all work items queued to unbound 6934 workqueues are affine to the NUMA nodes they're 6935 issued on, which results in better behavior in 6936 general. If NUMA affinity needs to be disabled for 6937 whatever reason, this option can be used. Note 6938 that this also can be controlled per-workqueue for 6939 workqueues visible under /sys/bus/workqueue/. 6940 6941 workqueue.power_efficient 6942 Per-cpu workqueues are generally preferred because 6943 they show better performance thanks to cache 6944 locality; unfortunately, per-cpu workqueues tend to 6945 be more power hungry than unbound workqueues. 6946 6947 Enabling this makes the per-cpu workqueues which 6948 were observed to contribute significantly to power 6949 consumption unbound, leading to measurably lower 6950 power usage at the cost of small performance 6951 overhead. 6952 6953 The default value of this parameter is determined by 6954 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 6955 6956 workqueue.debug_force_rr_cpu 6957 Workqueue used to implicitly guarantee that work 6958 items queued without explicit CPU specified are put 6959 on the local CPU. This guarantee is no longer true 6960 and while local CPU is still preferred work items 6961 may be put on foreign CPUs. This debug option 6962 forces round-robin CPU selection to flush out 6963 usages which depend on the now broken guarantee. 6964 When enabled, memory and cache locality will be 6965 impacted. 6966 6967 writecombine= [LOONGARCH] Control the MAT (Memory Access Type) of 6968 ioremap_wc(). 6969 6970 on - Enable writecombine, use WUC for ioremap_wc() 6971 off - Disable writecombine, use SUC for ioremap_wc() 6972 6973 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 6974 default x2apic cluster mode on platforms 6975 supporting x2apic. 6976 6977 xen_512gb_limit [KNL,X86-64,XEN] 6978 Restricts the kernel running paravirtualized under Xen 6979 to use only up to 512 GB of RAM. The reason to do so is 6980 crash analysis tools and Xen tools for doing domain 6981 save/restore/migration must be enabled to handle larger 6982 domains. 6983 6984 xen_emul_unplug= [HW,X86,XEN] 6985 Unplug Xen emulated devices 6986 Format: [unplug0,][unplug1] 6987 ide-disks -- unplug primary master IDE devices 6988 aux-ide-disks -- unplug non-primary-master IDE devices 6989 nics -- unplug network devices 6990 all -- unplug all emulated devices (NICs and IDE disks) 6991 unnecessary -- unplugging emulated devices is 6992 unnecessary even if the host did not respond to 6993 the unplug protocol 6994 never -- do not unplug even if version check succeeds 6995 6996 xen_legacy_crash [X86,XEN] 6997 Crash from Xen panic notifier, without executing late 6998 panic() code such as dumping handler. 6999 7000 xen_msr_safe= [X86,XEN] 7001 Format: <bool> 7002 Select whether to always use non-faulting (safe) MSR 7003 access functions when running as Xen PV guest. The 7004 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 7005 7006 xen_nopvspin [X86,XEN] 7007 Disables the qspinlock slowpath using Xen PV optimizations. 7008 This parameter is obsoleted by "nopvspin" parameter, which 7009 has equivalent effect for XEN platform. 7010 7011 xen_nopv [X86] 7012 Disables the PV optimizations forcing the HVM guest to 7013 run as generic HVM guest with no PV drivers. 7014 This option is obsoleted by the "nopv" option, which 7015 has equivalent effect for XEN platform. 7016 7017 xen_no_vector_callback 7018 [KNL,X86,XEN] Disable the vector callback for Xen 7019 event channel interrupts. 7020 7021 xen_scrub_pages= [XEN] 7022 Boolean option to control scrubbing pages before giving them back 7023 to Xen, for use by other domains. Can be also changed at runtime 7024 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 7025 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 7026 7027 xen_timer_slop= [X86-64,XEN] 7028 Set the timer slop (in nanoseconds) for the virtual Xen 7029 timers (default is 100000). This adjusts the minimum 7030 delta of virtualized Xen timers, where lower values 7031 improve timer resolution at the expense of processing 7032 more timer interrupts. 7033 7034 xen.balloon_boot_timeout= [XEN] 7035 The time (in seconds) to wait before giving up to boot 7036 in case initial ballooning fails to free enough memory. 7037 Applies only when running as HVM or PVH guest and 7038 started with less memory configured than allowed at 7039 max. Default is 180. 7040 7041 xen.event_eoi_delay= [XEN] 7042 How long to delay EOI handling in case of event 7043 storms (jiffies). Default is 10. 7044 7045 xen.event_loop_timeout= [XEN] 7046 After which time (jiffies) the event handling loop 7047 should start to delay EOI handling. Default is 2. 7048 7049 xen.fifo_events= [XEN] 7050 Boolean parameter to disable using fifo event handling 7051 even if available. Normally fifo event handling is 7052 preferred over the 2-level event handling, as it is 7053 fairer and the number of possible event channels is 7054 much higher. Default is on (use fifo events). 7055 7056 xirc2ps_cs= [NET,PCMCIA] 7057 Format: 7058 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 7059 7060 xive= [PPC] 7061 By default on POWER9 and above, the kernel will 7062 natively use the XIVE interrupt controller. This option 7063 allows the fallback firmware mode to be used: 7064 7065 off Fallback to firmware control of XIVE interrupt 7066 controller on both pseries and powernv 7067 platforms. Only useful on POWER9 and above. 7068 7069 xive.store-eoi=off [PPC] 7070 By default on POWER10 and above, the kernel will use 7071 stores for EOI handling when the XIVE interrupt mode 7072 is active. This option allows the XIVE driver to use 7073 loads instead, as on POWER9. 7074 7075 xhci-hcd.quirks [USB,KNL] 7076 A hex value specifying bitmask with supplemental xhci 7077 host controller quirks. Meaning of each bit can be 7078 consulted in header drivers/usb/host/xhci.h. 7079 7080 xmon [PPC] 7081 Format: { early | on | rw | ro | off } 7082 Controls if xmon debugger is enabled. Default is off. 7083 Passing only "xmon" is equivalent to "xmon=early". 7084 early Call xmon as early as possible on boot; xmon 7085 debugger is called from setup_arch(). 7086 on xmon debugger hooks will be installed so xmon 7087 is only called on a kernel crash. Default mode, 7088 i.e. either "ro" or "rw" mode, is controlled 7089 with CONFIG_XMON_DEFAULT_RO_MODE. 7090 rw xmon debugger hooks will be installed so xmon 7091 is called only on a kernel crash, mode is write, 7092 meaning SPR registers, memory and, other data 7093 can be written using xmon commands. 7094 ro same as "rw" option above but SPR registers, 7095 memory, and other data can't be written using 7096 xmon commands. 7097 off xmon is disabled. 7098 7099