1 acpi= [HW,ACPI,X86,ARM64,RISCV64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64,riscv64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64 and RISCV64, ONLY "acpi=off", "acpi=on" or 14 "acpi=force" are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 { vendor | video | native | none } 26 If set to vendor, prefer vendor-specific driver 27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 28 of the ACPI video.ko driver. 29 If set to video, use the ACPI video.ko driver. 30 If set to native, use the device's native backlight mode. 31 If set to none, disable the ACPI backlight interface. 32 33 acpi_force_32bit_fadt_addr 34 force FADT to use 32 bit addresses rather than the 35 64 bit X_* addresses. Some firmware have broken 64 36 bit addresses for force ACPI ignore these and use 37 the older legacy 32 bit addresses. 38 39 acpica_no_return_repair [HW, ACPI] 40 Disable AML predefined validation mechanism 41 This mechanism can repair the evaluation result to make 42 the return objects more ACPI specification compliant. 43 This option is useful for developers to identify the 44 root cause of an AML interpreter issue when the issue 45 has something to do with the repair mechanism. 46 47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 49 Format: <int> 50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 51 debug output. Bits in debug_layer correspond to a 52 _COMPONENT in an ACPI source file, e.g., 53 #define _COMPONENT ACPI_EVENTS 54 Bits in debug_level correspond to a level in 55 ACPI_DEBUG_PRINT statements, e.g., 56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 57 The debug_level mask defaults to "info". See 58 Documentation/firmware-guide/acpi/debug.rst for more information about 59 debug layers and levels. 60 61 Enable processor driver info messages: 62 acpi.debug_layer=0x20000000 63 Enable AML "Debug" output, i.e., stores to the Debug 64 object while interpreting AML: 65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 66 Enable all messages related to ACPI hardware: 67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 68 69 Some values produce so much output that the system is 70 unusable. The "log_buf_len" parameter may be useful 71 if you need to capture more output. 72 73 acpi_enforce_resources= [ACPI] 74 { strict | lax | no } 75 Check for resource conflicts between native drivers 76 and ACPI OperationRegions (SystemIO and SystemMemory 77 only). IO ports and memory declared in ACPI might be 78 used by the ACPI subsystem in arbitrary AML code and 79 can interfere with legacy drivers. 80 strict (default): access to resources claimed by ACPI 81 is denied; legacy drivers trying to access reserved 82 resources will fail to bind to device using them. 83 lax: access to resources claimed by ACPI is allowed; 84 legacy drivers trying to access reserved resources 85 will bind successfully but a warning message is logged. 86 no: ACPI OperationRegions are not marked as reserved, 87 no further checks are performed. 88 89 acpi_force_table_verification [HW,ACPI] 90 Enable table checksum verification during early stage. 91 By default, this is disabled due to x86 early mapping 92 size limitation. 93 94 acpi_irq_balance [HW,ACPI] 95 ACPI will balance active IRQs 96 default in APIC mode 97 98 acpi_irq_nobalance [HW,ACPI] 99 ACPI will not move active IRQs (default) 100 default in PIC mode 101 102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 103 Format: <irq>,<irq>... 104 105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 106 use by PCI 107 Format: <irq>,<irq>... 108 109 acpi_mask_gpe= [HW,ACPI] 110 Due to the existence of _Lxx/_Exx, some GPEs triggered 111 by unsupported hardware/firmware features can result in 112 GPE floodings that cannot be automatically disabled by 113 the GPE dispatcher. 114 This facility can be used to prevent such uncontrolled 115 GPE floodings. 116 Format: <byte> or <bitmap-list> 117 118 acpi_no_auto_serialize [HW,ACPI] 119 Disable auto-serialization of AML methods 120 AML control methods that contain the opcodes to create 121 named objects will be marked as "Serialized" by the 122 auto-serialization feature. 123 This feature is enabled by default. 124 This option allows to turn off the feature. 125 126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 127 kernels. 128 129 acpi_no_static_ssdt [HW,ACPI] 130 Disable installation of static SSDTs at early boot time 131 By default, SSDTs contained in the RSDT/XSDT will be 132 installed automatically and they will appear under 133 /sys/firmware/acpi/tables. 134 This option turns off this feature. 135 Note that specifying this option does not affect 136 dynamic table installation which will install SSDT 137 tables to /sys/firmware/acpi/tables/dynamic. 138 139 acpi_no_watchdog [HW,ACPI,WDT] 140 Ignore the ACPI-based watchdog interface (WDAT) and let 141 a native driver control the watchdog device instead. 142 143 acpi_rsdp= [ACPI,EFI,KEXEC] 144 Pass the RSDP address to the kernel, mostly used 145 on machines running EFI runtime service to boot the 146 second kernel for kdump. 147 148 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 149 Format: To spoof as Windows 98: ="Microsoft Windows" 150 151 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 152 of 2 which is mandated by ACPI 6) as the supported ACPI 153 specification revision (when using this switch, it may 154 be necessary to carry out a cold reboot _twice_ in a 155 row to make it take effect on the platform firmware). 156 157 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 158 acpi_osi="string1" # add string1 159 acpi_osi="!string2" # remove string2 160 acpi_osi=!* # remove all strings 161 acpi_osi=! # disable all built-in OS vendor 162 strings 163 acpi_osi=!! # enable all built-in OS vendor 164 strings 165 acpi_osi= # disable all strings 166 167 'acpi_osi=!' can be used in combination with single or 168 multiple 'acpi_osi="string1"' to support specific OS 169 vendor string(s). Note that such command can only 170 affect the default state of the OS vendor strings, thus 171 it cannot affect the default state of the feature group 172 strings and the current state of the OS vendor strings, 173 specifying it multiple times through kernel command line 174 is meaningless. This command is useful when one do not 175 care about the state of the feature group strings which 176 should be controlled by the OSPM. 177 Examples: 178 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 179 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 180 can make '_OSI("Windows 2000")' TRUE. 181 182 'acpi_osi=' cannot be used in combination with other 183 'acpi_osi=' command lines, the _OSI method will not 184 exist in the ACPI namespace. NOTE that such command can 185 only affect the _OSI support state, thus specifying it 186 multiple times through kernel command line is also 187 meaningless. 188 Examples: 189 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 190 FALSE. 191 192 'acpi_osi=!*' can be used in combination with single or 193 multiple 'acpi_osi="string1"' to support specific 194 string(s). Note that such command can affect the 195 current state of both the OS vendor strings and the 196 feature group strings, thus specifying it multiple times 197 through kernel command line is meaningful. But it may 198 still not able to affect the final state of a string if 199 there are quirks related to this string. This command 200 is useful when one want to control the state of the 201 feature group strings to debug BIOS issues related to 202 the OSPM features. 203 Examples: 204 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 205 '_OSI("Module Device")' FALSE. 206 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 207 '_OSI("Module Device")' TRUE. 208 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 209 equivalent to 210 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 211 and 212 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 213 they all will make '_OSI("Windows 2000")' TRUE. 214 215 acpi_pm_good [X86] 216 Override the pmtimer bug detection: force the kernel 217 to assume that this machine's pmtimer latches its value 218 and always returns good values. 219 220 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 221 Format: { level | edge | high | low } 222 223 acpi_skip_timer_override [HW,ACPI] 224 Recognize and ignore IRQ0/pin2 Interrupt Override. 225 For broken nForce2 BIOS resulting in XT-PIC timer. 226 227 acpi_sleep= [HW,ACPI] Sleep options 228 Format: { s3_bios, s3_mode, s3_beep, s4_hwsig, 229 s4_nohwsig, old_ordering, nonvs, 230 sci_force_enable, nobl } 231 See Documentation/power/video.rst for information on 232 s3_bios and s3_mode. 233 s3_beep is for debugging; it makes the PC's speaker beep 234 as soon as the kernel's real-mode entry point is called. 235 s4_hwsig causes the kernel to check the ACPI hardware 236 signature during resume from hibernation, and gracefully 237 refuse to resume if it has changed. This complies with 238 the ACPI specification but not with reality, since 239 Windows does not do this and many laptops do change it 240 on docking. So the default behaviour is to allow resume 241 and simply warn when the signature changes, unless the 242 s4_hwsig option is enabled. 243 s4_nohwsig prevents ACPI hardware signature from being 244 used (or even warned about) during resume. 245 old_ordering causes the ACPI 1.0 ordering of the _PTS 246 control method, with respect to putting devices into 247 low power states, to be enforced (the ACPI 2.0 ordering 248 of _PTS is used by default). 249 nonvs prevents the kernel from saving/restoring the 250 ACPI NVS memory during suspend/hibernation and resume. 251 sci_force_enable causes the kernel to set SCI_EN directly 252 on resume from S1/S3 (which is against the ACPI spec, 253 but some broken systems don't work without it). 254 nobl causes the internal blacklist of systems known to 255 behave incorrectly in some ways with respect to system 256 suspend and resume to be ignored (use wisely). 257 258 acpi_use_timer_override [HW,ACPI] 259 Use timer override. For some broken Nvidia NF5 boards 260 that require a timer override, but don't have HPET 261 262 add_efi_memmap [EFI; X86] Include EFI memory map in 263 kernel's map of available physical RAM. 264 265 agp= [AGP] 266 { off | try_unsupported } 267 off: disable AGP support 268 try_unsupported: try to drive unsupported chipsets 269 (may crash computer or cause data corruption) 270 271 ALSA [HW,ALSA] 272 See Documentation/sound/alsa-configuration.rst 273 274 alignment= [KNL,ARM] 275 Allow the default userspace alignment fault handler 276 behaviour to be specified. Bit 0 enables warnings, 277 bit 1 enables fixups, and bit 2 sends a segfault. 278 279 align_va_addr= [X86-64] 280 Align virtual addresses by clearing slice [14:12] when 281 allocating a VMA at process creation time. This option 282 gives you up to 3% performance improvement on AMD F15h 283 machines (where it is enabled by default) for a 284 CPU-intensive style benchmark, and it can vary highly in 285 a microbenchmark depending on workload and compiler. 286 287 32: only for 32-bit processes 288 64: only for 64-bit processes 289 on: enable for both 32- and 64-bit processes 290 off: disable for both 32- and 64-bit processes 291 292 alloc_snapshot [FTRACE] 293 Allocate the ftrace snapshot buffer on boot up when the 294 main buffer is allocated. This is handy if debugging 295 and you need to use tracing_snapshot() on boot up, and 296 do not want to use tracing_snapshot_alloc() as it needs 297 to be done where GFP_KERNEL allocations are allowed. 298 299 allow_mismatched_32bit_el0 [ARM64] 300 Allow execve() of 32-bit applications and setting of the 301 PER_LINUX32 personality on systems where only a strict 302 subset of the CPUs support 32-bit EL0. When this 303 parameter is present, the set of CPUs supporting 32-bit 304 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 305 and hot-unplug operations may be restricted. 306 307 See Documentation/arch/arm64/asymmetric-32bit.rst for more 308 information. 309 310 amd_iommu= [HW,X86-64] 311 Pass parameters to the AMD IOMMU driver in the system. 312 Possible values are: 313 fullflush - Deprecated, equivalent to iommu.strict=1 314 off - do not initialize any AMD IOMMU found in 315 the system 316 force_isolation - Force device isolation for all 317 devices. The IOMMU driver is not 318 allowed anymore to lift isolation 319 requirements as needed. This option 320 does not override iommu=pt 321 force_enable - Force enable the IOMMU on platforms known 322 to be buggy with IOMMU enabled. Use this 323 option with care. 324 pgtbl_v1 - Use v1 page table for DMA-API (Default). 325 pgtbl_v2 - Use v2 page table for DMA-API. 326 irtcachedis - Disable Interrupt Remapping Table (IRT) caching. 327 328 amd_iommu_dump= [HW,X86-64] 329 Enable AMD IOMMU driver option to dump the ACPI table 330 for AMD IOMMU. With this option enabled, AMD IOMMU 331 driver will print ACPI tables for AMD IOMMU during 332 IOMMU initialization. 333 334 amd_iommu_intr= [HW,X86-64] 335 Specifies one of the following AMD IOMMU interrupt 336 remapping modes: 337 legacy - Use legacy interrupt remapping mode. 338 vapic - Use virtual APIC mode, which allows IOMMU 339 to inject interrupts directly into guest. 340 This mode requires kvm-amd.avic=1. 341 (Default when IOMMU HW support is present.) 342 343 amd_pstate= [X86] 344 disable 345 Do not enable amd_pstate as the default 346 scaling driver for the supported processors 347 passive 348 Use amd_pstate with passive mode as a scaling driver. 349 In this mode autonomous selection is disabled. 350 Driver requests a desired performance level and platform 351 tries to match the same performance level if it is 352 satisfied by guaranteed performance level. 353 active 354 Use amd_pstate_epp driver instance as the scaling driver, 355 driver provides a hint to the hardware if software wants 356 to bias toward performance (0x0) or energy efficiency (0xff) 357 to the CPPC firmware. then CPPC power algorithm will 358 calculate the runtime workload and adjust the realtime cores 359 frequency. 360 guided 361 Activate guided autonomous mode. Driver requests minimum and 362 maximum performance level and the platform autonomously 363 selects a performance level in this range and appropriate 364 to the current workload. 365 366 amijoy.map= [HW,JOY] Amiga joystick support 367 Map of devices attached to JOY0DAT and JOY1DAT 368 Format: <a>,<b> 369 See also Documentation/input/joydev/joystick.rst 370 371 analog.map= [HW,JOY] Analog joystick and gamepad support 372 Specifies type or capabilities of an analog joystick 373 connected to one of 16 gameports 374 Format: <type1>,<type2>,..<type16> 375 376 apc= [HW,SPARC] 377 Power management functions (SPARCstation-4/5 + deriv.) 378 Format: noidle 379 Disable APC CPU standby support. SPARCstation-Fox does 380 not play well with APC CPU idle - disable it if you have 381 APC and your system crashes randomly. 382 383 apic= [APIC,X86] Advanced Programmable Interrupt Controller 384 Change the output verbosity while booting 385 Format: { quiet (default) | verbose | debug } 386 Change the amount of debugging information output 387 when initialising the APIC and IO-APIC components. 388 For X86-32, this can also be used to specify an APIC 389 driver name. 390 Format: apic=driver_name 391 Examples: apic=bigsmp 392 393 apic_extnmi= [APIC,X86] External NMI delivery setting 394 Format: { bsp (default) | all | none } 395 bsp: External NMI is delivered only to CPU 0 396 all: External NMIs are broadcast to all CPUs as a 397 backup of CPU 0 398 none: External NMI is masked for all CPUs. This is 399 useful so that a dump capture kernel won't be 400 shot down by NMI 401 402 autoconf= [IPV6] 403 See Documentation/networking/ipv6.rst. 404 405 apm= [APM] Advanced Power Management 406 See header of arch/x86/kernel/apm_32.c. 407 408 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 409 Format: { "0" | "1" } 410 See security/apparmor/Kconfig help text 411 0 -- disable. 412 1 -- enable. 413 Default value is set via kernel config option. 414 415 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 416 Format: <io>,<irq>,<nodeID> 417 418 arm64.nobti [ARM64] Unconditionally disable Branch Target 419 Identification support 420 421 arm64.nomops [ARM64] Unconditionally disable Memory Copy and Memory 422 Set instructions support 423 424 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 425 support 426 427 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 428 support 429 430 arm64.nosme [ARM64] Unconditionally disable Scalable Matrix 431 Extension support 432 433 arm64.nosve [ARM64] Unconditionally disable Scalable Vector 434 Extension support 435 436 ataflop= [HW,M68k] 437 438 atarimouse= [HW,MOUSE] Atari Mouse 439 440 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 441 EzKey and similar keyboards 442 443 atkbd.reset= [HW] Reset keyboard during initialization 444 445 atkbd.set= [HW] Select keyboard code set 446 Format: <int> (2 = AT (default), 3 = PS/2) 447 448 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 449 keyboards 450 451 atkbd.softraw= [HW] Choose between synthetic and real raw mode 452 Format: <bool> (0 = real, 1 = synthetic (default)) 453 454 atkbd.softrepeat= [HW] 455 Use software keyboard repeat 456 457 audit= [KNL] Enable the audit sub-system 458 Format: { "0" | "1" | "off" | "on" } 459 0 | off - kernel audit is disabled and can not be 460 enabled until the next reboot 461 unset - kernel audit is initialized but disabled and 462 will be fully enabled by the userspace auditd. 463 1 | on - kernel audit is initialized and partially 464 enabled, storing at most audit_backlog_limit 465 messages in RAM until it is fully enabled by the 466 userspace auditd. 467 Default: unset 468 469 audit_backlog_limit= [KNL] Set the audit queue size limit. 470 Format: <int> (must be >=0) 471 Default: 64 472 473 bau= [X86_UV] Enable the BAU on SGI UV. The default 474 behavior is to disable the BAU (i.e. bau=0). 475 Format: { "0" | "1" } 476 0 - Disable the BAU. 477 1 - Enable the BAU. 478 unset - Disable the BAU. 479 480 baycom_epp= [HW,AX25] 481 Format: <io>,<mode> 482 483 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 484 Format: <io>,<mode> 485 See header of drivers/net/hamradio/baycom_par.c. 486 487 baycom_ser_fdx= [HW,AX25] 488 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 489 Format: <io>,<irq>,<mode>[,<baud>] 490 See header of drivers/net/hamradio/baycom_ser_fdx.c. 491 492 baycom_ser_hdx= [HW,AX25] 493 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 494 Format: <io>,<irq>,<mode> 495 See header of drivers/net/hamradio/baycom_ser_hdx.c. 496 497 bert_disable [ACPI] 498 Disable BERT OS support on buggy BIOSes. 499 500 bgrt_disable [ACPI][X86] 501 Disable BGRT to avoid flickering OEM logo. 502 503 blkdevparts= Manual partition parsing of block device(s) for 504 embedded devices based on command line input. 505 See Documentation/block/cmdline-partition.rst 506 507 boot_delay= Milliseconds to delay each printk during boot. 508 Only works if CONFIG_BOOT_PRINTK_DELAY is enabled, 509 and you may also have to specify "lpj=". Boot_delay 510 values larger than 10 seconds (10000) are assumed 511 erroneous and ignored. 512 Format: integer 513 514 bootconfig [KNL] 515 Extended command line options can be added to an initrd 516 and this will cause the kernel to look for it. 517 518 See Documentation/admin-guide/bootconfig.rst 519 520 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 521 bttv.radio= Most important insmod options are available as 522 kernel args too. 523 bttv.pll= See Documentation/admin-guide/media/bttv.rst 524 bttv.tuner= 525 526 bulk_remove=off [PPC] This parameter disables the use of the pSeries 527 firmware feature for flushing multiple hpte entries 528 at a time. 529 530 c101= [NET] Moxa C101 synchronous serial card 531 532 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 533 Sometimes CPU hardware bugs make them report the cache 534 size incorrectly. The kernel will attempt work arounds 535 to fix known problems, but for some CPUs it is not 536 possible to determine what the correct size should be. 537 This option provides an override for these situations. 538 539 carrier_timeout= 540 [NET] Specifies amount of time (in seconds) that 541 the kernel should wait for a network carrier. By default 542 it waits 120 seconds. 543 544 ca_keys= [KEYS] This parameter identifies a specific key(s) on 545 the system trusted keyring to be used for certificate 546 trust validation. 547 format: { id:<keyid> | builtin } 548 549 cca= [MIPS] Override the kernel pages' cache coherency 550 algorithm. Accepted values range from 0 to 7 551 inclusive. See arch/mips/include/asm/pgtable-bits.h 552 for platform specific values (SB1, Loongson3 and 553 others). 554 555 ccw_timeout_log [S390] 556 See Documentation/arch/s390/common_io.rst for details. 557 558 cgroup_disable= [KNL] Disable a particular controller or optional feature 559 Format: {name of the controller(s) or feature(s) to disable} 560 The effects of cgroup_disable=foo are: 561 - foo isn't auto-mounted if you mount all cgroups in 562 a single hierarchy 563 - foo isn't visible as an individually mountable 564 subsystem 565 - if foo is an optional feature then the feature is 566 disabled and corresponding cgroup files are not 567 created 568 {Currently only "memory" controller deal with this and 569 cut the overhead, others just disable the usage. So 570 only cgroup_disable=memory is actually worthy} 571 Specifying "pressure" disables per-cgroup pressure 572 stall information accounting feature 573 574 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 575 Format: { { controller | "all" | "named" } 576 [,{ controller | "all" | "named" }...] } 577 Like cgroup_disable, but only applies to cgroup v1; 578 the blacklisted controllers remain available in cgroup2. 579 "all" blacklists all controllers and "named" disables 580 named mounts. Specifying both "all" and "named" disables 581 all v1 hierarchies. 582 583 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 584 Format: <string> 585 nosocket -- Disable socket memory accounting. 586 nokmem -- Disable kernel memory accounting. 587 nobpf -- Disable BPF memory accounting. 588 589 checkreqprot= [SELINUX] Set initial checkreqprot flag value. 590 Format: { "0" | "1" } 591 See security/selinux/Kconfig help text. 592 0 -- check protection applied by kernel (includes 593 any implied execute protection). 594 1 -- check protection requested by application. 595 Default value is set via a kernel config option. 596 Value can be changed at runtime via 597 /sys/fs/selinux/checkreqprot. 598 Setting checkreqprot to 1 is deprecated. 599 600 cio_ignore= [S390] 601 See Documentation/arch/s390/common_io.rst for details. 602 603 clearcpuid=X[,X...] [X86] 604 Disable CPUID feature X for the kernel. See 605 arch/x86/include/asm/cpufeatures.h for the valid bit 606 numbers X. Note the Linux-specific bits are not necessarily 607 stable over kernel options, but the vendor-specific 608 ones should be. 609 X can also be a string as appearing in the flags: line 610 in /proc/cpuinfo which does not have the above 611 instability issue. However, not all features have names 612 in /proc/cpuinfo. 613 Note that using this option will taint your kernel. 614 Also note that user programs calling CPUID directly 615 or using the feature without checking anything 616 will still see it. This just prevents it from 617 being used by the kernel or shown in /proc/cpuinfo. 618 Also note the kernel might malfunction if you disable 619 some critical bits. 620 621 clk_ignore_unused 622 [CLK] 623 Prevents the clock framework from automatically gating 624 clocks that have not been explicitly enabled by a Linux 625 device driver but are enabled in hardware at reset or 626 by the bootloader/firmware. Note that this does not 627 force such clocks to be always-on nor does it reserve 628 those clocks in any way. This parameter is useful for 629 debug and development, but should not be needed on a 630 platform with proper driver support. For more 631 information, see Documentation/driver-api/clk.rst. 632 633 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 634 [Deprecated] 635 Forces specified clocksource (if available) to be used 636 when calculating gettimeofday(). If specified 637 clocksource is not available, it defaults to PIT. 638 Format: { pit | tsc | cyclone | pmtmr } 639 640 clocksource= Override the default clocksource 641 Format: <string> 642 Override the default clocksource and use the clocksource 643 with the name specified. 644 Some clocksource names to choose from, depending on 645 the platform: 646 [all] jiffies (this is the base, fallback clocksource) 647 [ACPI] acpi_pm 648 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 649 pxa_timer,timer3,32k_counter,timer0_1 650 [X86-32] pit,hpet,tsc; 651 scx200_hrt on Geode; cyclone on IBM x440 652 [MIPS] MIPS 653 [PARISC] cr16 654 [S390] tod 655 [SH] SuperH 656 [SPARC64] tick 657 [X86-64] hpet,tsc 658 659 clocksource.arm_arch_timer.evtstrm= 660 [ARM,ARM64] 661 Format: <bool> 662 Enable/disable the eventstream feature of the ARM 663 architected timer so that code using WFE-based polling 664 loops can be debugged more effectively on production 665 systems. 666 667 clocksource.verify_n_cpus= [KNL] 668 Limit the number of CPUs checked for clocksources 669 marked with CLOCK_SOURCE_VERIFY_PERCPU that 670 are marked unstable due to excessive skew. 671 A negative value says to check all CPUs, while 672 zero says not to check any. Values larger than 673 nr_cpu_ids are silently truncated to nr_cpu_ids. 674 The actual CPUs are chosen randomly, with 675 no replacement if the same CPU is chosen twice. 676 677 clocksource-wdtest.holdoff= [KNL] 678 Set the time in seconds that the clocksource 679 watchdog test waits before commencing its tests. 680 Defaults to zero when built as a module and to 681 10 seconds when built into the kernel. 682 683 cma=nn[MG]@[start[MG][-end[MG]]] 684 [KNL,CMA] 685 Sets the size of kernel global memory area for 686 contiguous memory allocations and optionally the 687 placement constraint by the physical address range of 688 memory allocations. A value of 0 disables CMA 689 altogether. For more information, see 690 kernel/dma/contiguous.c 691 692 cma_pernuma=nn[MG] 693 [KNL,CMA] 694 Sets the size of kernel per-numa memory area for 695 contiguous memory allocations. A value of 0 disables 696 per-numa CMA altogether. And If this option is not 697 specified, the default value is 0. 698 With per-numa CMA enabled, DMA users on node nid will 699 first try to allocate buffer from the pernuma area 700 which is located in node nid, if the allocation fails, 701 they will fallback to the global default memory area. 702 703 numa_cma=<node>:nn[MG][,<node>:nn[MG]] 704 [KNL,CMA] 705 Sets the size of kernel numa memory area for 706 contiguous memory allocations. It will reserve CMA 707 area for the specified node. 708 709 With numa CMA enabled, DMA users on node nid will 710 first try to allocate buffer from the numa area 711 which is located in node nid, if the allocation fails, 712 they will fallback to the global default memory area. 713 714 cmo_free_hint= [PPC] Format: { yes | no } 715 Specify whether pages are marked as being inactive 716 when they are freed. This is used in CMO environments 717 to determine OS memory pressure for page stealing by 718 a hypervisor. 719 Default: yes 720 721 coherent_pool=nn[KMG] [ARM,KNL] 722 Sets the size of memory pool for coherent, atomic dma 723 allocations, by default set to 256K. 724 725 com20020= [HW,NET] ARCnet - COM20020 chipset 726 Format: 727 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 728 729 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 730 Format: <io>[,<irq>] 731 732 com90xx= [HW,NET] 733 ARCnet - COM90xx chipset (memory-mapped buffers) 734 Format: <io>[,<irq>[,<memstart>]] 735 736 condev= [HW,S390] console device 737 conmode= 738 739 con3215_drop= [S390] 3215 console drop mode. 740 Format: y|n|Y|N|1|0 741 When set to true, drop data on the 3215 console when 742 the console buffer is full. In this case the 743 operator using a 3270 terminal emulator (for example 744 x3270) does not have to enter the clear key for the 745 console output to advance and the kernel to continue. 746 This leads to a much faster boot time when a 3270 747 terminal emulator is active. If no 3270 terminal 748 emulator is used, this parameter has no effect. 749 750 console= [KNL] Output console device and options. 751 752 tty<n> Use the virtual console device <n>. 753 754 ttyS<n>[,options] 755 ttyUSB0[,options] 756 Use the specified serial port. The options are of 757 the form "bbbbpnf", where "bbbb" is the baud rate, 758 "p" is parity ("n", "o", or "e"), "n" is number of 759 bits, and "f" is flow control ("r" for RTS or 760 omit it). Default is "9600n8". 761 762 See Documentation/admin-guide/serial-console.rst for more 763 information. See 764 Documentation/networking/netconsole.rst for an 765 alternative. 766 767 uart[8250],io,<addr>[,options] 768 uart[8250],mmio,<addr>[,options] 769 uart[8250],mmio16,<addr>[,options] 770 uart[8250],mmio32,<addr>[,options] 771 uart[8250],0x<addr>[,options] 772 Start an early, polled-mode console on the 8250/16550 773 UART at the specified I/O port or MMIO address, 774 switching to the matching ttyS device later. 775 MMIO inter-register address stride is either 8-bit 776 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 777 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 778 to be equivalent to 'mmio'. 'options' are specified in 779 the same format described for ttyS above; if unspecified, 780 the h/w is not re-initialized. 781 782 hvc<n> Use the hypervisor console device <n>. This is for 783 both Xen and PowerPC hypervisors. 784 785 { null | "" } 786 Use to disable console output, i.e., to have kernel 787 console messages discarded. 788 This must be the only console= parameter used on the 789 kernel command line. 790 791 If the device connected to the port is not a TTY but a braille 792 device, prepend "brl," before the device type, for instance 793 console=brl,ttyS0 794 For now, only VisioBraille is supported. 795 796 console_msg_format= 797 [KNL] Change console messages format 798 default 799 By default we print messages on consoles in 800 "[time stamp] text\n" format (time stamp may not be 801 printed, depending on CONFIG_PRINTK_TIME or 802 `printk_time' param). 803 syslog 804 Switch to syslog format: "<%u>[time stamp] text\n" 805 IOW, each message will have a facility and loglevel 806 prefix. The format is similar to one used by syslog() 807 syscall, or to executing "dmesg -S --raw" or to reading 808 from /proc/kmsg. 809 810 consoleblank= [KNL] The console blank (screen saver) timeout in 811 seconds. A value of 0 disables the blank timer. 812 Defaults to 0. 813 814 coredump_filter= 815 [KNL] Change the default value for 816 /proc/<pid>/coredump_filter. 817 See also Documentation/filesystems/proc.rst. 818 819 coresight_cpu_debug.enable 820 [ARM,ARM64] 821 Format: <bool> 822 Enable/disable the CPU sampling based debugging. 823 0: default value, disable debugging 824 1: enable debugging at boot time 825 826 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 827 Format: 828 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 829 830 cpuidle.off=1 [CPU_IDLE] 831 disable the cpuidle sub-system 832 833 cpuidle.governor= 834 [CPU_IDLE] Name of the cpuidle governor to use. 835 836 cpufreq.off=1 [CPU_FREQ] 837 disable the cpufreq sub-system 838 839 cpufreq.default_governor= 840 [CPU_FREQ] Name of the default cpufreq governor or 841 policy to use. This governor must be registered in the 842 kernel before the cpufreq driver probes. 843 844 cpu_init_udelay=N 845 [X86] Delay for N microsec between assert and de-assert 846 of APIC INIT to start processors. This delay occurs 847 on every CPU online, such as boot, and resume from suspend. 848 Default: 10000 849 850 cpuhp.parallel= 851 [SMP] Enable/disable parallel bringup of secondary CPUs 852 Format: <bool> 853 Default is enabled if CONFIG_HOTPLUG_PARALLEL=y. Otherwise 854 the parameter has no effect. 855 856 crash_kexec_post_notifiers 857 Run kdump after running panic-notifiers and dumping 858 kmsg. This only for the users who doubt kdump always 859 succeeds in any situation. 860 Note that this also increases risks of kdump failure, 861 because some panic notifiers can make the crashed 862 kernel more unstable. 863 864 crashkernel=size[KMG][@offset[KMG]] 865 [KNL] Using kexec, Linux can switch to a 'crash kernel' 866 upon panic. This parameter reserves the physical 867 memory region [offset, offset + size] for that kernel 868 image. If '@offset' is omitted, then a suitable offset 869 is selected automatically. 870 [KNL, X86-64, ARM64, RISCV] Select a region under 4G first, and 871 fall back to reserve region above 4G when '@offset' 872 hasn't been specified. 873 See Documentation/admin-guide/kdump/kdump.rst for further details. 874 875 crashkernel=range1:size1[,range2:size2,...][@offset] 876 [KNL] Same as above, but depends on the memory 877 in the running system. The syntax of range is 878 start-[end] where start and end are both 879 a memory unit (amount[KMG]). See also 880 Documentation/admin-guide/kdump/kdump.rst for an example. 881 882 crashkernel=size[KMG],high 883 [KNL, X86-64, ARM64, RISCV] range could be above 4G. 884 Allow kernel to allocate physical memory region from top, 885 so could be above 4G if system have more than 4G ram 886 installed. Otherwise memory region will be allocated 887 below 4G, if available. 888 It will be ignored if crashkernel=X is specified. 889 crashkernel=size[KMG],low 890 [KNL, X86-64, ARM64, RISCV] range under 4G. When crashkernel=X,high 891 is passed, kernel could allocate physical memory region 892 above 4G, that cause second kernel crash on system 893 that require some amount of low memory, e.g. swiotlb 894 requires at least 64M+32K low memory, also enough extra 895 low memory is needed to make sure DMA buffers for 32-bit 896 devices won't run out. Kernel would try to allocate 897 default size of memory below 4G automatically. The default 898 size is platform dependent. 899 --> x86: max(swiotlb_size_or_default() + 8MiB, 256MiB) 900 --> arm64: 128MiB 901 --> riscv: 128MiB 902 This one lets the user specify own low range under 4G 903 for second kernel instead. 904 0: to disable low allocation. 905 It will be ignored when crashkernel=X,high is not used 906 or memory reserved is below 4G. 907 908 cryptomgr.notests 909 [KNL] Disable crypto self-tests 910 911 cs89x0_dma= [HW,NET] 912 Format: <dma> 913 914 cs89x0_media= [HW,NET] 915 Format: { rj45 | aui | bnc } 916 917 csdlock_debug= [KNL] Enable or disable debug add-ons of cross-CPU 918 function call handling. When switched on, 919 additional debug data is printed to the console 920 in case a hanging CPU is detected, and that 921 CPU is pinged again in order to try to resolve 922 the hang situation. The default value of this 923 option depends on the CSD_LOCK_WAIT_DEBUG_DEFAULT 924 Kconfig option. 925 926 dasd= [HW,NET] 927 See header of drivers/s390/block/dasd_devmap.c. 928 929 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 930 (one device per port) 931 Format: <port#>,<type> 932 See also Documentation/input/devices/joystick-parport.rst 933 934 debug [KNL] Enable kernel debugging (events log level). 935 936 debug_boot_weak_hash 937 [KNL] Enable printing [hashed] pointers early in the 938 boot sequence. If enabled, we use a weak hash instead 939 of siphash to hash pointers. Use this option if you are 940 seeing instances of '(___ptrval___)') and need to see a 941 value (hashed pointer) instead. Cryptographically 942 insecure, please do not use on production kernels. 943 944 debug_locks_verbose= 945 [KNL] verbose locking self-tests 946 Format: <int> 947 Print debugging info while doing the locking API 948 self-tests. 949 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 950 (no extra messages), setting it to -1 (all bits set) 951 will print _a_lot_ more information - normally only 952 useful to lockdep developers. 953 954 debug_objects [KNL] Enable object debugging 955 956 debug_guardpage_minorder= 957 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 958 parameter allows control of the order of pages that will 959 be intentionally kept free (and hence protected) by the 960 buddy allocator. Bigger value increase the probability 961 of catching random memory corruption, but reduce the 962 amount of memory for normal system use. The maximum 963 possible value is MAX_ORDER/2. Setting this parameter 964 to 1 or 2 should be enough to identify most random 965 memory corruption problems caused by bugs in kernel or 966 driver code when a CPU writes to (or reads from) a 967 random memory location. Note that there exists a class 968 of memory corruptions problems caused by buggy H/W or 969 F/W or by drivers badly programming DMA (basically when 970 memory is written at bus level and the CPU MMU is 971 bypassed) which are not detectable by 972 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 973 tracking down these problems. 974 975 debug_pagealloc= 976 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 977 enables the feature at boot time. By default, it is 978 disabled and the system will work mostly the same as a 979 kernel built without CONFIG_DEBUG_PAGEALLOC. 980 Note: to get most of debug_pagealloc error reports, it's 981 useful to also enable the page_owner functionality. 982 on: enable the feature 983 984 debugfs= [KNL] This parameter enables what is exposed to userspace 985 and debugfs internal clients. 986 Format: { on, no-mount, off } 987 on: All functions are enabled. 988 no-mount: 989 Filesystem is not registered but kernel clients can 990 access APIs and a crashkernel can be used to read 991 its content. There is nothing to mount. 992 off: Filesystem is not registered and clients 993 get a -EPERM as result when trying to register files 994 or directories within debugfs. 995 This is equivalent of the runtime functionality if 996 debugfs was not enabled in the kernel at all. 997 Default value is set in build-time with a kernel configuration. 998 999 debugpat [X86] Enable PAT debugging 1000 1001 default_hugepagesz= 1002 [HW] The size of the default HugeTLB page. This is 1003 the size represented by the legacy /proc/ hugepages 1004 APIs. In addition, this is the default hugetlb size 1005 used for shmget(), mmap() and mounting hugetlbfs 1006 filesystems. If not specified, defaults to the 1007 architecture's default huge page size. Huge page 1008 sizes are architecture dependent. See also 1009 Documentation/admin-guide/mm/hugetlbpage.rst. 1010 Format: size[KMG] 1011 1012 deferred_probe_timeout= 1013 [KNL] Debugging option to set a timeout in seconds for 1014 deferred probe to give up waiting on dependencies to 1015 probe. Only specific dependencies (subsystems or 1016 drivers) that have opted in will be ignored. A timeout 1017 of 0 will timeout at the end of initcalls. If the time 1018 out hasn't expired, it'll be restarted by each 1019 successful driver registration. This option will also 1020 dump out devices still on the deferred probe list after 1021 retrying. 1022 1023 delayacct [KNL] Enable per-task delay accounting 1024 1025 dell_smm_hwmon.ignore_dmi= 1026 [HW] Continue probing hardware even if DMI data 1027 indicates that the driver is running on unsupported 1028 hardware. 1029 1030 dell_smm_hwmon.force= 1031 [HW] Activate driver even if SMM BIOS signature does 1032 not match list of supported models and enable otherwise 1033 blacklisted features. 1034 1035 dell_smm_hwmon.power_status= 1036 [HW] Report power status in /proc/i8k 1037 (disabled by default). 1038 1039 dell_smm_hwmon.restricted= 1040 [HW] Allow controlling fans only if SYS_ADMIN 1041 capability is set. 1042 1043 dell_smm_hwmon.fan_mult= 1044 [HW] Factor to multiply fan speed with. 1045 1046 dell_smm_hwmon.fan_max= 1047 [HW] Maximum configurable fan speed. 1048 1049 dfltcc= [HW,S390] 1050 Format: { on | off | def_only | inf_only | always } 1051 on: s390 zlib hardware support for compression on 1052 level 1 and decompression (default) 1053 off: No s390 zlib hardware support 1054 def_only: s390 zlib hardware support for deflate 1055 only (compression on level 1) 1056 inf_only: s390 zlib hardware support for inflate 1057 only (decompression) 1058 always: Same as 'on' but ignores the selected compression 1059 level always using hardware support (used for debugging) 1060 1061 dhash_entries= [KNL] 1062 Set number of hash buckets for dentry cache. 1063 1064 disable_1tb_segments [PPC] 1065 Disables the use of 1TB hash page table segments. This 1066 causes the kernel to fall back to 256MB segments which 1067 can be useful when debugging issues that require an SLB 1068 miss to occur. 1069 1070 disable= [IPV6] 1071 See Documentation/networking/ipv6.rst. 1072 1073 disable_radix [PPC] 1074 Disable RADIX MMU mode on POWER9 1075 1076 disable_tlbie [PPC] 1077 Disable TLBIE instruction. Currently does not work 1078 with KVM, with HASH MMU, or with coherent accelerators. 1079 1080 disable_cpu_apicid= [X86,APIC,SMP] 1081 Format: <int> 1082 The number of initial APIC ID for the 1083 corresponding CPU to be disabled at boot, 1084 mostly used for the kdump 2nd kernel to 1085 disable BSP to wake up multiple CPUs without 1086 causing system reset or hang due to sending 1087 INIT from AP to BSP. 1088 1089 disable_ddw [PPC/PSERIES] 1090 Disable Dynamic DMA Window support. Use this 1091 to workaround buggy firmware. 1092 1093 disable_ipv6= [IPV6] 1094 See Documentation/networking/ipv6.rst. 1095 1096 disable_mtrr_cleanup [X86] 1097 The kernel tries to adjust MTRR layout from continuous 1098 to discrete, to make X server driver able to add WB 1099 entry later. This parameter disables that. 1100 1101 disable_mtrr_trim [X86, Intel and AMD only] 1102 By default the kernel will trim any uncacheable 1103 memory out of your available memory pool based on 1104 MTRR settings. This parameter disables that behavior, 1105 possibly causing your machine to run very slowly. 1106 1107 disable_timer_pin_1 [X86] 1108 Disable PIN 1 of APIC timer 1109 Can be useful to work around chipset bugs. 1110 1111 dis_ucode_ldr [X86] Disable the microcode loader. 1112 1113 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1114 this option disables the debugging code at boot. 1115 1116 dma_debug_entries=<number> 1117 This option allows to tune the number of preallocated 1118 entries for DMA-API debugging code. One entry is 1119 required per DMA-API allocation. Use this if the 1120 DMA-API debugging code disables itself because the 1121 architectural default is too low. 1122 1123 dma_debug_driver=<driver_name> 1124 With this option the DMA-API debugging driver 1125 filter feature can be enabled at boot time. Just 1126 pass the driver to filter for as the parameter. 1127 The filter can be disabled or changed to another 1128 driver later using sysfs. 1129 1130 reg_file_data_sampling= 1131 [X86] Controls mitigation for Register File Data 1132 Sampling (RFDS) vulnerability. RFDS is a CPU 1133 vulnerability which may allow userspace to infer 1134 kernel data values previously stored in floating point 1135 registers, vector registers, or integer registers. 1136 RFDS only affects Intel Atom processors. 1137 1138 on: Turns ON the mitigation. 1139 off: Turns OFF the mitigation. 1140 1141 This parameter overrides the compile time default set 1142 by CONFIG_MITIGATION_RFDS. Mitigation cannot be 1143 disabled when other VERW based mitigations (like MDS) 1144 are enabled. In order to disable RFDS mitigation all 1145 VERW based mitigations need to be disabled. 1146 1147 For details see: 1148 Documentation/admin-guide/hw-vuln/reg-file-data-sampling.rst 1149 1150 driver_async_probe= [KNL] 1151 List of driver names to be probed asynchronously. * 1152 matches with all driver names. If * is specified, the 1153 rest of the listed driver names are those that will NOT 1154 match the *. 1155 Format: <driver_name1>,<driver_name2>... 1156 1157 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1158 Broken monitors, graphic adapters, KVMs and EDIDless 1159 panels may send no or incorrect EDID data sets. 1160 This parameter allows to specify an EDID data sets 1161 in the /lib/firmware directory that are used instead. 1162 Generic built-in EDID data sets are used, if one of 1163 edid/1024x768.bin, edid/1280x1024.bin, 1164 edid/1680x1050.bin, or edid/1920x1080.bin is given 1165 and no file with the same name exists. Details and 1166 instructions how to build your own EDID data are 1167 available in Documentation/admin-guide/edid.rst. An EDID 1168 data set will only be used for a particular connector, 1169 if its name and a colon are prepended to the EDID 1170 name. Each connector may use a unique EDID data 1171 set by separating the files with a comma. An EDID 1172 data set with no connector name will be used for 1173 any connectors not explicitly specified. 1174 1175 dscc4.setup= [NET] 1176 1177 dt_cpu_ftrs= [PPC] 1178 Format: {"off" | "known"} 1179 Control how the dt_cpu_ftrs device-tree binding is 1180 used for CPU feature discovery and setup (if it 1181 exists). 1182 off: Do not use it, fall back to legacy cpu table. 1183 known: Do not pass through unknown features to guests 1184 or userspace, only those that the kernel is aware of. 1185 1186 dump_apple_properties [X86] 1187 Dump name and content of EFI device properties on 1188 x86 Macs. Useful for driver authors to determine 1189 what data is available or for reverse-engineering. 1190 1191 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1192 <module>.dyndbg[="val"] 1193 Enable debug messages at boot time. See 1194 Documentation/admin-guide/dynamic-debug-howto.rst 1195 for details. 1196 1197 early_ioremap_debug [KNL] 1198 Enable debug messages in early_ioremap support. This 1199 is useful for tracking down temporary early mappings 1200 which are not unmapped. 1201 1202 earlycon= [KNL] Output early console device and options. 1203 1204 When used with no options, the early console is 1205 determined by stdout-path property in device tree's 1206 chosen node or the ACPI SPCR table if supported by 1207 the platform. 1208 1209 cdns,<addr>[,options] 1210 Start an early, polled-mode console on a Cadence 1211 (xuartps) serial port at the specified address. Only 1212 supported option is baud rate. If baud rate is not 1213 specified, the serial port must already be setup and 1214 configured. 1215 1216 uart[8250],io,<addr>[,options[,uartclk]] 1217 uart[8250],mmio,<addr>[,options[,uartclk]] 1218 uart[8250],mmio32,<addr>[,options[,uartclk]] 1219 uart[8250],mmio32be,<addr>[,options[,uartclk]] 1220 uart[8250],0x<addr>[,options] 1221 Start an early, polled-mode console on the 8250/16550 1222 UART at the specified I/O port or MMIO address. 1223 MMIO inter-register address stride is either 8-bit 1224 (mmio) or 32-bit (mmio32 or mmio32be). 1225 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1226 to be equivalent to 'mmio'. 'options' are specified 1227 in the same format described for "console=ttyS<n>"; if 1228 unspecified, the h/w is not initialized. 'uartclk' is 1229 the uart clock frequency; if unspecified, it is set 1230 to 'BASE_BAUD' * 16. 1231 1232 pl011,<addr> 1233 pl011,mmio32,<addr> 1234 Start an early, polled-mode console on a pl011 serial 1235 port at the specified address. The pl011 serial port 1236 must already be setup and configured. Options are not 1237 yet supported. If 'mmio32' is specified, then only 1238 the driver will use only 32-bit accessors to read/write 1239 the device registers. 1240 1241 liteuart,<addr> 1242 Start an early console on a litex serial port at the 1243 specified address. The serial port must already be 1244 setup and configured. Options are not yet supported. 1245 1246 meson,<addr> 1247 Start an early, polled-mode console on a meson serial 1248 port at the specified address. The serial port must 1249 already be setup and configured. Options are not yet 1250 supported. 1251 1252 msm_serial,<addr> 1253 Start an early, polled-mode console on an msm serial 1254 port at the specified address. The serial port 1255 must already be setup and configured. Options are not 1256 yet supported. 1257 1258 msm_serial_dm,<addr> 1259 Start an early, polled-mode console on an msm serial 1260 dm port at the specified address. The serial port 1261 must already be setup and configured. Options are not 1262 yet supported. 1263 1264 owl,<addr> 1265 Start an early, polled-mode console on a serial port 1266 of an Actions Semi SoC, such as S500 or S900, at the 1267 specified address. The serial port must already be 1268 setup and configured. Options are not yet supported. 1269 1270 rda,<addr> 1271 Start an early, polled-mode console on a serial port 1272 of an RDA Micro SoC, such as RDA8810PL, at the 1273 specified address. The serial port must already be 1274 setup and configured. Options are not yet supported. 1275 1276 sbi 1277 Use RISC-V SBI (Supervisor Binary Interface) for early 1278 console. 1279 1280 smh Use ARM semihosting calls for early console. 1281 1282 s3c2410,<addr> 1283 s3c2412,<addr> 1284 s3c2440,<addr> 1285 s3c6400,<addr> 1286 s5pv210,<addr> 1287 exynos4210,<addr> 1288 Use early console provided by serial driver available 1289 on Samsung SoCs, requires selecting proper type and 1290 a correct base address of the selected UART port. The 1291 serial port must already be setup and configured. 1292 Options are not yet supported. 1293 1294 lantiq,<addr> 1295 Start an early, polled-mode console on a lantiq serial 1296 (lqasc) port at the specified address. The serial port 1297 must already be setup and configured. Options are not 1298 yet supported. 1299 1300 lpuart,<addr> 1301 lpuart32,<addr> 1302 Use early console provided by Freescale LP UART driver 1303 found on Freescale Vybrid and QorIQ LS1021A processors. 1304 A valid base address must be provided, and the serial 1305 port must already be setup and configured. 1306 1307 ec_imx21,<addr> 1308 ec_imx6q,<addr> 1309 Start an early, polled-mode, output-only console on the 1310 Freescale i.MX UART at the specified address. The UART 1311 must already be setup and configured. 1312 1313 ar3700_uart,<addr> 1314 Start an early, polled-mode console on the 1315 Armada 3700 serial port at the specified 1316 address. The serial port must already be setup 1317 and configured. Options are not yet supported. 1318 1319 qcom_geni,<addr> 1320 Start an early, polled-mode console on a Qualcomm 1321 Generic Interface (GENI) based serial port at the 1322 specified address. The serial port must already be 1323 setup and configured. Options are not yet supported. 1324 1325 efifb,[options] 1326 Start an early, unaccelerated console on the EFI 1327 memory mapped framebuffer (if available). On cache 1328 coherent non-x86 systems that use system memory for 1329 the framebuffer, pass the 'ram' option so that it is 1330 mapped with the correct attributes. 1331 1332 linflex,<addr> 1333 Use early console provided by Freescale LINFlexD UART 1334 serial driver for NXP S32V234 SoCs. A valid base 1335 address must be provided, and the serial port must 1336 already be setup and configured. 1337 1338 earlyprintk= [X86,SH,ARM,M68k,S390] 1339 earlyprintk=vga 1340 earlyprintk=sclp 1341 earlyprintk=xen 1342 earlyprintk=serial[,ttySn[,baudrate]] 1343 earlyprintk=serial[,0x...[,baudrate]] 1344 earlyprintk=ttySn[,baudrate] 1345 earlyprintk=dbgp[debugController#] 1346 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1347 earlyprintk=xdbc[xhciController#] 1348 1349 earlyprintk is useful when the kernel crashes before 1350 the normal console is initialized. It is not enabled by 1351 default because it has some cosmetic problems. 1352 1353 Append ",keep" to not disable it when the real console 1354 takes over. 1355 1356 Only one of vga, serial, or usb debug port can 1357 be used at a time. 1358 1359 Currently only ttyS0 and ttyS1 may be specified by 1360 name. Other I/O ports may be explicitly specified 1361 on some architectures (x86 and arm at least) by 1362 replacing ttySn with an I/O port address, like this: 1363 earlyprintk=serial,0x1008,115200 1364 You can find the port for a given device in 1365 /proc/tty/driver/serial: 1366 2: uart:ST16650V2 port:00001008 irq:18 ... 1367 1368 Interaction with the standard serial driver is not 1369 very good. 1370 1371 The VGA output is eventually overwritten by 1372 the real console. 1373 1374 The xen option can only be used in Xen domains. 1375 1376 The sclp output can only be used on s390. 1377 1378 The optional "force" to "pciserial" enables use of a 1379 PCI device even when its classcode is not of the 1380 UART class. 1381 1382 edac_report= [HW,EDAC] Control how to report EDAC event 1383 Format: {"on" | "off" | "force"} 1384 on: enable EDAC to report H/W event. May be overridden 1385 by other higher priority error reporting module. 1386 off: disable H/W event reporting through EDAC. 1387 force: enforce the use of EDAC to report H/W event. 1388 default: on. 1389 1390 edd= [EDD] 1391 Format: {"off" | "on" | "skip[mbr]"} 1392 1393 efi= [EFI] 1394 Format: { "debug", "disable_early_pci_dma", 1395 "nochunk", "noruntime", "nosoftreserve", 1396 "novamap", "no_disable_early_pci_dma" } 1397 debug: enable misc debug output. 1398 disable_early_pci_dma: disable the busmaster bit on all 1399 PCI bridges while in the EFI boot stub. 1400 nochunk: disable reading files in "chunks" in the EFI 1401 boot stub, as chunking can cause problems with some 1402 firmware implementations. 1403 noruntime : disable EFI runtime services support 1404 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1405 attribute may cause the kernel to reserve the 1406 memory range for a memory mapping driver to 1407 claim. Specify efi=nosoftreserve to disable this 1408 reservation and treat the memory by its base type 1409 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1410 novamap: do not call SetVirtualAddressMap(). 1411 no_disable_early_pci_dma: Leave the busmaster bit set 1412 on all PCI bridges while in the EFI boot stub 1413 1414 efi_no_storage_paranoia [EFI; X86] 1415 Using this parameter you can use more than 50% of 1416 your efi variable storage. Use this parameter only if 1417 you are really sure that your UEFI does sane gc and 1418 fulfills the spec otherwise your board may brick. 1419 1420 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1421 Add arbitrary attribute to specific memory range by 1422 updating original EFI memory map. 1423 Region of memory which aa attribute is added to is 1424 from ss to ss+nn. 1425 1426 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1427 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1428 attribute is added to range 0x100000000-0x180000000 and 1429 0x10a0000000-0x1120000000. 1430 1431 If efi_fake_mem=8G@9G:0x40000 is specified, the 1432 EFI_MEMORY_SP(0x40000) attribute is added to 1433 range 0x240000000-0x43fffffff. 1434 1435 Using this parameter you can do debugging of EFI memmap 1436 related features. For example, you can do debugging of 1437 Address Range Mirroring feature even if your box 1438 doesn't support it, or mark specific memory as 1439 "soft reserved". 1440 1441 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1442 that is to be dynamically loaded by Linux. If there are 1443 multiple variables with the same name but with different 1444 vendor GUIDs, all of them will be loaded. See 1445 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1446 1447 1448 eisa_irq_edge= [PARISC,HW] 1449 See header of drivers/parisc/eisa.c. 1450 1451 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1452 Format: ekgdboc=kbd 1453 1454 This is designed to be used in conjunction with 1455 the boot argument: earlyprintk=vga 1456 1457 This parameter works in place of the kgdboc parameter 1458 but can only be used if the backing tty is available 1459 very early in the boot process. For early debugging 1460 via a serial port see kgdboc_earlycon instead. 1461 1462 elanfreq= [X86-32] 1463 See comment before function elanfreq_setup() in 1464 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1465 1466 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1467 Specifies physical address of start of kernel core 1468 image elf header and optionally the size. Generally 1469 kexec loader will pass this option to capture kernel. 1470 See Documentation/admin-guide/kdump/kdump.rst for details. 1471 1472 enable_mtrr_cleanup [X86] 1473 The kernel tries to adjust MTRR layout from continuous 1474 to discrete, to make X server driver able to add WB 1475 entry later. This parameter enables that. 1476 1477 enable_timer_pin_1 [X86] 1478 Enable PIN 1 of APIC timer 1479 Can be useful to work around chipset bugs 1480 (in particular on some ATI chipsets). 1481 The kernel tries to set a reasonable default. 1482 1483 enforcing= [SELINUX] Set initial enforcing status. 1484 Format: {"0" | "1"} 1485 See security/selinux/Kconfig help text. 1486 0 -- permissive (log only, no denials). 1487 1 -- enforcing (deny and log). 1488 Default value is 0. 1489 Value can be changed at runtime via 1490 /sys/fs/selinux/enforce. 1491 1492 erst_disable [ACPI] 1493 Disable Error Record Serialization Table (ERST) 1494 support. 1495 1496 ether= [HW,NET] Ethernet cards parameters 1497 This option is obsoleted by the "netdev=" option, which 1498 has equivalent usage. See its documentation for details. 1499 1500 evm= [EVM] 1501 Format: { "fix" } 1502 Permit 'security.evm' to be updated regardless of 1503 current integrity status. 1504 1505 early_page_ext [KNL] Enforces page_ext initialization to earlier 1506 stages so cover more early boot allocations. 1507 Please note that as side effect some optimizations 1508 might be disabled to achieve that (e.g. parallelized 1509 memory initialization is disabled) so the boot process 1510 might take longer, especially on systems with a lot of 1511 memory. Available with CONFIG_PAGE_EXTENSION=y. 1512 1513 failslab= 1514 fail_usercopy= 1515 fail_page_alloc= 1516 fail_make_request=[KNL] 1517 General fault injection mechanism. 1518 Format: <interval>,<probability>,<space>,<times> 1519 See also Documentation/fault-injection/. 1520 1521 fb_tunnels= [NET] 1522 Format: { initns | none } 1523 See Documentation/admin-guide/sysctl/net.rst for 1524 fb_tunnels_only_for_init_ns 1525 1526 floppy= [HW] 1527 See Documentation/admin-guide/blockdev/floppy.rst. 1528 1529 force_pal_cache_flush 1530 [IA-64] Avoid check_sal_cache_flush which may hang on 1531 buggy SAL_CACHE_FLUSH implementations. Using this 1532 parameter will force ia64_sal_cache_flush to call 1533 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1534 1535 forcepae [X86-32] 1536 Forcefully enable Physical Address Extension (PAE). 1537 Many Pentium M systems disable PAE but may have a 1538 functionally usable PAE implementation. 1539 Warning: use of this parameter will taint the kernel 1540 and may cause unknown problems. 1541 1542 ftrace=[tracer] 1543 [FTRACE] will set and start the specified tracer 1544 as early as possible in order to facilitate early 1545 boot debugging. 1546 1547 ftrace_boot_snapshot 1548 [FTRACE] On boot up, a snapshot will be taken of the 1549 ftrace ring buffer that can be read at: 1550 /sys/kernel/tracing/snapshot. 1551 This is useful if you need tracing information from kernel 1552 boot up that is likely to be overridden by user space 1553 start up functionality. 1554 1555 Optionally, the snapshot can also be defined for a tracing 1556 instance that was created by the trace_instance= command 1557 line parameter. 1558 1559 trace_instance=foo,sched_switch ftrace_boot_snapshot=foo 1560 1561 The above will cause the "foo" tracing instance to trigger 1562 a snapshot at the end of boot up. 1563 1564 ftrace_dump_on_oops[=orig_cpu] 1565 [FTRACE] will dump the trace buffers on oops. 1566 If no parameter is passed, ftrace will dump 1567 buffers of all CPUs, but if you pass orig_cpu, it will 1568 dump only the buffer of the CPU that triggered the 1569 oops. 1570 1571 ftrace_filter=[function-list] 1572 [FTRACE] Limit the functions traced by the function 1573 tracer at boot up. function-list is a comma-separated 1574 list of functions. This list can be changed at run 1575 time by the set_ftrace_filter file in the debugfs 1576 tracing directory. 1577 1578 ftrace_notrace=[function-list] 1579 [FTRACE] Do not trace the functions specified in 1580 function-list. This list can be changed at run time 1581 by the set_ftrace_notrace file in the debugfs 1582 tracing directory. 1583 1584 ftrace_graph_filter=[function-list] 1585 [FTRACE] Limit the top level callers functions traced 1586 by the function graph tracer at boot up. 1587 function-list is a comma-separated list of functions 1588 that can be changed at run time by the 1589 set_graph_function file in the debugfs tracing directory. 1590 1591 ftrace_graph_notrace=[function-list] 1592 [FTRACE] Do not trace from the functions specified in 1593 function-list. This list is a comma-separated list of 1594 functions that can be changed at run time by the 1595 set_graph_notrace file in the debugfs tracing directory. 1596 1597 ftrace_graph_max_depth=<uint> 1598 [FTRACE] Used with the function graph tracer. This is 1599 the max depth it will trace into a function. This value 1600 can be changed at run time by the max_graph_depth file 1601 in the tracefs tracing directory. default: 0 (no limit) 1602 1603 fw_devlink= [KNL] Create device links between consumer and supplier 1604 devices by scanning the firmware to infer the 1605 consumer/supplier relationships. This feature is 1606 especially useful when drivers are loaded as modules as 1607 it ensures proper ordering of tasks like device probing 1608 (suppliers first, then consumers), supplier boot state 1609 clean up (only after all consumers have probed), 1610 suspend/resume & runtime PM (consumers first, then 1611 suppliers). 1612 Format: { off | permissive | on | rpm } 1613 off -- Don't create device links from firmware info. 1614 permissive -- Create device links from firmware info 1615 but use it only for ordering boot state clean 1616 up (sync_state() calls). 1617 on -- Create device links from firmware info and use it 1618 to enforce probe and suspend/resume ordering. 1619 rpm -- Like "on", but also use to order runtime PM. 1620 1621 fw_devlink.strict=<bool> 1622 [KNL] Treat all inferred dependencies as mandatory 1623 dependencies. This only applies for fw_devlink=on|rpm. 1624 Format: <bool> 1625 1626 fw_devlink.sync_state = 1627 [KNL] When all devices that could probe have finished 1628 probing, this parameter controls what to do with 1629 devices that haven't yet received their sync_state() 1630 calls. 1631 Format: { strict | timeout } 1632 strict -- Default. Continue waiting on consumers to 1633 probe successfully. 1634 timeout -- Give up waiting on consumers and call 1635 sync_state() on any devices that haven't yet 1636 received their sync_state() calls after 1637 deferred_probe_timeout has expired or by 1638 late_initcall() if !CONFIG_MODULES. 1639 1640 gamecon.map[2|3]= 1641 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1642 support via parallel port (up to 5 devices per port) 1643 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1644 See also Documentation/input/devices/joystick-parport.rst 1645 1646 gamma= [HW,DRM] 1647 1648 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART 1649 Format: off | on 1650 default: on 1651 1652 gather_data_sampling= 1653 [X86,INTEL] Control the Gather Data Sampling (GDS) 1654 mitigation. 1655 1656 Gather Data Sampling is a hardware vulnerability which 1657 allows unprivileged speculative access to data which was 1658 previously stored in vector registers. 1659 1660 This issue is mitigated by default in updated microcode. 1661 The mitigation may have a performance impact but can be 1662 disabled. On systems without the microcode mitigation 1663 disabling AVX serves as a mitigation. 1664 1665 force: Disable AVX to mitigate systems without 1666 microcode mitigation. No effect if the microcode 1667 mitigation is present. Known to cause crashes in 1668 userspace with buggy AVX enumeration. 1669 1670 off: Disable GDS mitigation. 1671 1672 gcov_persist= [GCOV] When non-zero (default), profiling data for 1673 kernel modules is saved and remains accessible via 1674 debugfs, even when the module is unloaded/reloaded. 1675 When zero, profiling data is discarded and associated 1676 debugfs files are removed at module unload time. 1677 1678 goldfish [X86] Enable the goldfish android emulator platform. 1679 Don't use this when you are not running on the 1680 android emulator 1681 1682 gpio-mockup.gpio_mockup_ranges 1683 [HW] Sets the ranges of gpiochip of for this device. 1684 Format: <start1>,<end1>,<start2>,<end2>... 1685 gpio-mockup.gpio_mockup_named_lines 1686 [HW] Let the driver know GPIO lines should be named. 1687 1688 gpt [EFI] Forces disk with valid GPT signature but 1689 invalid Protective MBR to be treated as GPT. If the 1690 primary GPT is corrupted, it enables the backup/alternate 1691 GPT to be used instead. 1692 1693 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1694 the "Enable 0" bit of the configuration register. 1695 Format: 0 | 1 1696 Default: 0 1697 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1698 the "Enable 0" bit of the configuration register. 1699 Format: 0 | 1 1700 Default: 0 1701 grcan.select= [HW] Select which physical interface to use. 1702 Format: 0 | 1 1703 Default: 0 1704 grcan.txsize= [HW] Sets the size of the tx buffer. 1705 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1706 Default: 1024 1707 grcan.rxsize= [HW] Sets the size of the rx buffer. 1708 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1709 Default: 1024 1710 1711 hardened_usercopy= 1712 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 1713 hardening is enabled for this boot. Hardened 1714 usercopy checking is used to protect the kernel 1715 from reading or writing beyond known memory 1716 allocation boundaries as a proactive defense 1717 against bounds-checking flaws in the kernel's 1718 copy_to_user()/copy_from_user() interface. 1719 on Perform hardened usercopy checks (default). 1720 off Disable hardened usercopy checks. 1721 1722 hardlockup_all_cpu_backtrace= 1723 [KNL] Should the hard-lockup detector generate 1724 backtraces on all cpus. 1725 Format: 0 | 1 1726 1727 hashdist= [KNL,NUMA] Large hashes allocated during boot 1728 are distributed across NUMA nodes. Defaults on 1729 for 64-bit NUMA, off otherwise. 1730 Format: 0 | 1 (for off | on) 1731 1732 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1733 1734 hd= [EIDE] (E)IDE hard drive subsystem geometry 1735 Format: <cyl>,<head>,<sect> 1736 1737 hest_disable [ACPI] 1738 Disable Hardware Error Source Table (HEST) support; 1739 corresponding firmware-first mode error processing 1740 logic will be disabled. 1741 1742 hibernate= [HIBERNATION] 1743 noresume Don't check if there's a hibernation image 1744 present during boot. 1745 nocompress Don't compress/decompress hibernation images. 1746 no Disable hibernation and resume. 1747 protect_image Turn on image protection during restoration 1748 (that will set all pages holding image data 1749 during restoration read-only). 1750 1751 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1752 size of <nn>. This works even on boxes that have no 1753 highmem otherwise. This also works to reduce highmem 1754 size on bigger boxes. 1755 1756 highres= [KNL] Enable/disable high resolution timer mode. 1757 Valid parameters: "on", "off" 1758 Default: "on" 1759 1760 hlt [BUGS=ARM,SH] 1761 1762 hostname= [KNL] Set the hostname (aka UTS nodename). 1763 Format: <string> 1764 This allows setting the system's hostname during early 1765 startup. This sets the name returned by gethostname. 1766 Using this parameter to set the hostname makes it 1767 possible to ensure the hostname is correctly set before 1768 any userspace processes run, avoiding the possibility 1769 that a process may call gethostname before the hostname 1770 has been explicitly set, resulting in the calling 1771 process getting an incorrect result. The string must 1772 not exceed the maximum allowed hostname length (usually 1773 64 characters) and will be truncated otherwise. 1774 1775 hpet= [X86-32,HPET] option to control HPET usage 1776 Format: { enable (default) | disable | force | 1777 verbose } 1778 disable: disable HPET and use PIT instead 1779 force: allow force enabled of undocumented chips (ICH4, 1780 VIA, nVidia) 1781 verbose: show contents of HPET registers during setup 1782 1783 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1784 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1785 1786 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1787 If this follows hugepagesz (below), it specifies 1788 the number of pages of hugepagesz to be allocated. 1789 If this is the first HugeTLB parameter on the command 1790 line, it specifies the number of pages to allocate for 1791 the default huge page size. If using node format, the 1792 number of pages to allocate per-node can be specified. 1793 See also Documentation/admin-guide/mm/hugetlbpage.rst. 1794 Format: <integer> or (node format) 1795 <node>:<integer>[,<node>:<integer>] 1796 1797 hugepagesz= 1798 [HW] The size of the HugeTLB pages. This is used in 1799 conjunction with hugepages (above) to allocate huge 1800 pages of a specific size at boot. The pair 1801 hugepagesz=X hugepages=Y can be specified once for 1802 each supported huge page size. Huge page sizes are 1803 architecture dependent. See also 1804 Documentation/admin-guide/mm/hugetlbpage.rst. 1805 Format: size[KMG] 1806 1807 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation 1808 of gigantic hugepages. Or using node format, the size 1809 of a CMA area per node can be specified. 1810 Format: nn[KMGTPE] or (node format) 1811 <node>:nn[KMGTPE][,<node>:nn[KMGTPE]] 1812 1813 Reserve a CMA area of given size and allocate gigantic 1814 hugepages using the CMA allocator. If enabled, the 1815 boot-time allocation of gigantic hugepages is skipped. 1816 1817 hugetlb_free_vmemmap= 1818 [KNL] Requires CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP 1819 enabled. 1820 Control if HugeTLB Vmemmap Optimization (HVO) is enabled. 1821 Allows heavy hugetlb users to free up some more 1822 memory (7 * PAGE_SIZE for each 2MB hugetlb page). 1823 Format: { on | off (default) } 1824 1825 on: enable HVO 1826 off: disable HVO 1827 1828 Built with CONFIG_HUGETLB_PAGE_OPTIMIZE_VMEMMAP_DEFAULT_ON=y, 1829 the default is on. 1830 1831 Note that the vmemmap pages may be allocated from the added 1832 memory block itself when memory_hotplug.memmap_on_memory is 1833 enabled, those vmemmap pages cannot be optimized even if this 1834 feature is enabled. Other vmemmap pages not allocated from 1835 the added memory block itself do not be affected. 1836 1837 hung_task_panic= 1838 [KNL] Should the hung task detector generate panics. 1839 Format: 0 | 1 1840 1841 A value of 1 instructs the kernel to panic when a 1842 hung task is detected. The default value is controlled 1843 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1844 option. The value selected by this boot parameter can 1845 be changed later by the kernel.hung_task_panic sysctl. 1846 1847 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1848 terminal devices. Valid values: 0..8 1849 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1850 If specified, z/VM IUCV HVC accepts connections 1851 from listed z/VM user IDs only. 1852 1853 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1854 which allow the hypervisor to 'idle' the 1855 guest on lock contention. 1856 1857 i2c_bus= [HW] Override the default board specific I2C bus speed 1858 or register an additional I2C bus that is not 1859 registered from board initialization code. 1860 Format: 1861 <bus_id>,<clkrate> 1862 1863 i8042.debug [HW] Toggle i8042 debug mode 1864 i8042.unmask_kbd_data 1865 [HW] Enable printing of interrupt data from the KBD port 1866 (disabled by default, and as a pre-condition 1867 requires that i8042.debug=1 be enabled) 1868 i8042.direct [HW] Put keyboard port into non-translated mode 1869 i8042.dumbkbd [HW] Pretend that controller can only read data from 1870 keyboard and cannot control its state 1871 (Don't attempt to blink the leds) 1872 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1873 i8042.nokbd [HW] Don't check/create keyboard port 1874 i8042.noloop [HW] Disable the AUX Loopback command while probing 1875 for the AUX port 1876 i8042.nomux [HW] Don't check presence of an active multiplexing 1877 controller 1878 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1879 controllers 1880 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1881 i8042.reset [HW] Reset the controller during init, cleanup and 1882 suspend-to-ram transitions, only during s2r 1883 transitions, or never reset 1884 Format: { 1 | Y | y | 0 | N | n } 1885 1, Y, y: always reset controller 1886 0, N, n: don't ever reset controller 1887 Default: only on s2r transitions on x86; most other 1888 architectures force reset to be always executed 1889 i8042.unlock [HW] Unlock (ignore) the keylock 1890 i8042.kbdreset [HW] Reset device connected to KBD port 1891 i8042.probe_defer 1892 [HW] Allow deferred probing upon i8042 probe errors 1893 1894 i810= [HW,DRM] 1895 1896 i915.invert_brightness= 1897 [DRM] Invert the sense of the variable that is used to 1898 set the brightness of the panel backlight. Normally a 1899 brightness value of 0 indicates backlight switched off, 1900 and the maximum of the brightness value sets the backlight 1901 to maximum brightness. If this parameter is set to 0 1902 (default) and the machine requires it, or this parameter 1903 is set to 1, a brightness value of 0 sets the backlight 1904 to maximum brightness, and the maximum of the brightness 1905 value switches the backlight off. 1906 -1 -- never invert brightness 1907 0 -- machine default 1908 1 -- force brightness inversion 1909 1910 icn= [HW,ISDN] 1911 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1912 1913 1914 idle= [X86] 1915 Format: idle=poll, idle=halt, idle=nomwait 1916 Poll forces a polling idle loop that can slightly 1917 improve the performance of waking up a idle CPU, but 1918 will use a lot of power and make the system run hot. 1919 Not recommended. 1920 idle=halt: Halt is forced to be used for CPU idle. 1921 In such case C2/C3 won't be used again. 1922 idle=nomwait: Disable mwait for CPU C-states 1923 1924 idxd.sva= [HW] 1925 Format: <bool> 1926 Allow force disabling of Shared Virtual Memory (SVA) 1927 support for the idxd driver. By default it is set to 1928 true (1). 1929 1930 idxd.tc_override= [HW] 1931 Format: <bool> 1932 Allow override of default traffic class configuration 1933 for the device. By default it is set to false (0). 1934 1935 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1936 Format: { strict | legacy | 2008 | relaxed } 1937 Default: strict 1938 1939 Choose which programs will be accepted for execution 1940 based on the IEEE 754 NaN encoding(s) supported by 1941 the FPU and the NaN encoding requested with the value 1942 of an ELF file header flag individually set by each 1943 binary. Hardware implementations are permitted to 1944 support either or both of the legacy and the 2008 NaN 1945 encoding mode. 1946 1947 Available settings are as follows: 1948 strict accept binaries that request a NaN encoding 1949 supported by the FPU 1950 legacy only accept legacy-NaN binaries, if supported 1951 by the FPU 1952 2008 only accept 2008-NaN binaries, if supported 1953 by the FPU 1954 relaxed accept any binaries regardless of whether 1955 supported by the FPU 1956 1957 The FPU emulator is always able to support both NaN 1958 encodings, so if no FPU hardware is present or it has 1959 been disabled with 'nofpu', then the settings of 1960 'legacy' and '2008' strap the emulator accordingly, 1961 'relaxed' straps the emulator for both legacy-NaN and 1962 2008-NaN, whereas 'strict' enables legacy-NaN only on 1963 legacy processors and both NaN encodings on MIPS32 or 1964 MIPS64 CPUs. 1965 1966 The setting for ABS.fmt/NEG.fmt instruction execution 1967 mode generally follows that for the NaN encoding, 1968 except where unsupported by hardware. 1969 1970 ignore_loglevel [KNL] 1971 Ignore loglevel setting - this will print /all/ 1972 kernel messages to the console. Useful for debugging. 1973 We also add it as printk module parameter, so users 1974 could change it dynamically, usually by 1975 /sys/module/printk/parameters/ignore_loglevel. 1976 1977 ignore_rlimit_data 1978 Ignore RLIMIT_DATA setting for data mappings, 1979 print warning at first misuse. Can be changed via 1980 /sys/module/kernel/parameters/ignore_rlimit_data. 1981 1982 ihash_entries= [KNL] 1983 Set number of hash buckets for inode cache. 1984 1985 ima_appraise= [IMA] appraise integrity measurements 1986 Format: { "off" | "enforce" | "fix" | "log" } 1987 default: "enforce" 1988 1989 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1990 The builtin appraise policy appraises all files 1991 owned by uid=0. 1992 1993 ima_canonical_fmt [IMA] 1994 Use the canonical format for the binary runtime 1995 measurements, instead of host native format. 1996 1997 ima_hash= [IMA] 1998 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1999 | sha512 | ... } 2000 default: "sha1" 2001 2002 The list of supported hash algorithms is defined 2003 in crypto/hash_info.h. 2004 2005 ima_policy= [IMA] 2006 The builtin policies to load during IMA setup. 2007 Format: "tcb | appraise_tcb | secure_boot | 2008 fail_securely | critical_data" 2009 2010 The "tcb" policy measures all programs exec'd, files 2011 mmap'd for exec, and all files opened with the read 2012 mode bit set by either the effective uid (euid=0) or 2013 uid=0. 2014 2015 The "appraise_tcb" policy appraises the integrity of 2016 all files owned by root. 2017 2018 The "secure_boot" policy appraises the integrity 2019 of files (eg. kexec kernel image, kernel modules, 2020 firmware, policy, etc) based on file signatures. 2021 2022 The "fail_securely" policy forces file signature 2023 verification failure also on privileged mounted 2024 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 2025 flag. 2026 2027 The "critical_data" policy measures kernel integrity 2028 critical data. 2029 2030 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 2031 Load a policy which meets the needs of the Trusted 2032 Computing Base. This means IMA will measure all 2033 programs exec'd, files mmap'd for exec, and all files 2034 opened for read by uid=0. 2035 2036 ima_template= [IMA] 2037 Select one of defined IMA measurements template formats. 2038 Formats: { "ima" | "ima-ng" | "ima-ngv2" | "ima-sig" | 2039 "ima-sigv2" } 2040 Default: "ima-ng" 2041 2042 ima_template_fmt= 2043 [IMA] Define a custom template format. 2044 Format: { "field1|...|fieldN" } 2045 2046 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 2047 Format: <min_file_size> 2048 Set the minimal file size for using asynchronous hash. 2049 If left unspecified, ahash usage is disabled. 2050 2051 ahash performance varies for different data sizes on 2052 different crypto accelerators. This option can be used 2053 to achieve the best performance for a particular HW. 2054 2055 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 2056 Format: <bufsize> 2057 Set hashing buffer size. Default: 4k. 2058 2059 ahash performance varies for different chunk sizes on 2060 different crypto accelerators. This option can be used 2061 to achieve best performance for particular HW. 2062 2063 init= [KNL] 2064 Format: <full_path> 2065 Run specified binary instead of /sbin/init as init 2066 process. 2067 2068 initcall_debug [KNL] Trace initcalls as they are executed. Useful 2069 for working out where the kernel is dying during 2070 startup. 2071 2072 initcall_blacklist= [KNL] Do not execute a comma-separated list of 2073 initcall functions. Useful for debugging built-in 2074 modules and initcalls. 2075 2076 initramfs_async= [KNL] 2077 Format: <bool> 2078 Default: 1 2079 This parameter controls whether the initramfs 2080 image is unpacked asynchronously, concurrently 2081 with devices being probed and 2082 initialized. This should normally just work, 2083 but as a debugging aid, one can get the 2084 historical behaviour of the initramfs 2085 unpacking being completed before device_ and 2086 late_ initcalls. 2087 2088 initrd= [BOOT] Specify the location of the initial ramdisk 2089 2090 initrdmem= [KNL] Specify a physical address and size from which to 2091 load the initrd. If an initrd is compiled in or 2092 specified in the bootparams, it takes priority over this 2093 setting. 2094 Format: ss[KMG],nn[KMG] 2095 Default is 0, 0 2096 2097 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 2098 zeroes. 2099 Format: 0 | 1 2100 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 2101 2102 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 2103 Format: 0 | 1 2104 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 2105 2106 init_pkru= [X86] Specify the default memory protection keys rights 2107 register contents for all processes. 0x55555554 by 2108 default (disallow access to all but pkey 0). Can 2109 override in debugfs after boot. 2110 2111 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 2112 Format: <irq> 2113 2114 int_pln_enable [X86] Enable power limit notification interrupt 2115 2116 integrity_audit=[IMA] 2117 Format: { "0" | "1" } 2118 0 -- basic integrity auditing messages. (Default) 2119 1 -- additional integrity auditing messages. 2120 2121 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 2122 on 2123 Enable intel iommu driver. 2124 off 2125 Disable intel iommu driver. 2126 igfx_off [Default Off] 2127 By default, gfx is mapped as normal device. If a gfx 2128 device has a dedicated DMAR unit, the DMAR unit is 2129 bypassed by not enabling DMAR with this option. In 2130 this case, gfx device will use physical address for 2131 DMA. 2132 strict [Default Off] 2133 Deprecated, equivalent to iommu.strict=1. 2134 sp_off [Default Off] 2135 By default, super page will be supported if Intel IOMMU 2136 has the capability. With this option, super page will 2137 not be supported. 2138 sm_on 2139 Enable the Intel IOMMU scalable mode if the hardware 2140 advertises that it has support for the scalable mode 2141 translation. 2142 sm_off 2143 Disallow use of the Intel IOMMU scalable mode. 2144 tboot_noforce [Default Off] 2145 Do not force the Intel IOMMU enabled under tboot. 2146 By default, tboot will force Intel IOMMU on, which 2147 could harm performance of some high-throughput 2148 devices like 40GBit network cards, even if identity 2149 mapping is enabled. 2150 Note that using this option lowers the security 2151 provided by tboot because it makes the system 2152 vulnerable to DMA attacks. 2153 2154 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 2155 0 disables intel_idle and fall back on acpi_idle. 2156 1 to 9 specify maximum depth of C-state. 2157 2158 intel_pstate= [X86] 2159 disable 2160 Do not enable intel_pstate as the default 2161 scaling driver for the supported processors 2162 active 2163 Use intel_pstate driver to bypass the scaling 2164 governors layer of cpufreq and provides it own 2165 algorithms for p-state selection. There are two 2166 P-state selection algorithms provided by 2167 intel_pstate in the active mode: powersave and 2168 performance. The way they both operate depends 2169 on whether or not the hardware managed P-states 2170 (HWP) feature has been enabled in the processor 2171 and possibly on the processor model. 2172 passive 2173 Use intel_pstate as a scaling driver, but configure it 2174 to work with generic cpufreq governors (instead of 2175 enabling its internal governor). This mode cannot be 2176 used along with the hardware-managed P-states (HWP) 2177 feature. 2178 force 2179 Enable intel_pstate on systems that prohibit it by default 2180 in favor of acpi-cpufreq. Forcing the intel_pstate driver 2181 instead of acpi-cpufreq may disable platform features, such 2182 as thermal controls and power capping, that rely on ACPI 2183 P-States information being indicated to OSPM and therefore 2184 should be used with caution. This option does not work with 2185 processors that aren't supported by the intel_pstate driver 2186 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 2187 no_hwp 2188 Do not enable hardware P state control (HWP) 2189 if available. 2190 hwp_only 2191 Only load intel_pstate on systems which support 2192 hardware P state control (HWP) if available. 2193 support_acpi_ppc 2194 Enforce ACPI _PPC performance limits. If the Fixed ACPI 2195 Description Table, specifies preferred power management 2196 profile as "Enterprise Server" or "Performance Server", 2197 then this feature is turned on by default. 2198 per_cpu_perf_limits 2199 Allow per-logical-CPU P-State performance control limits using 2200 cpufreq sysfs interface 2201 2202 intremap= [X86-64, Intel-IOMMU] 2203 on enable Interrupt Remapping (default) 2204 off disable Interrupt Remapping 2205 nosid disable Source ID checking 2206 no_x2apic_optout 2207 BIOS x2APIC opt-out request will be ignored 2208 nopost disable Interrupt Posting 2209 2210 iomem= Disable strict checking of access to MMIO memory 2211 strict regions from userspace. 2212 relaxed 2213 2214 iommu= [X86] 2215 off 2216 force 2217 noforce 2218 biomerge 2219 panic 2220 nopanic 2221 merge 2222 nomerge 2223 soft 2224 pt [X86] 2225 nopt [X86] 2226 nobypass [PPC/POWERNV] 2227 Disable IOMMU bypass, using IOMMU for PCI devices. 2228 2229 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices. 2230 Format: { "0" | "1" } 2231 0 - Try to allocate a 32-bit DMA address first, before 2232 falling back to the full range if needed. 2233 1 - Allocate directly from the full usable range, 2234 forcing Dual Address Cycle for PCI cards supporting 2235 greater than 32-bit addressing. 2236 2237 iommu.strict= [ARM64, X86] Configure TLB invalidation behaviour 2238 Format: { "0" | "1" } 2239 0 - Lazy mode. 2240 Request that DMA unmap operations use deferred 2241 invalidation of hardware TLBs, for increased 2242 throughput at the cost of reduced device isolation. 2243 Will fall back to strict mode if not supported by 2244 the relevant IOMMU driver. 2245 1 - Strict mode. 2246 DMA unmap operations invalidate IOMMU hardware TLBs 2247 synchronously. 2248 unset - Use value of CONFIG_IOMMU_DEFAULT_DMA_{LAZY,STRICT}. 2249 Note: on x86, strict mode specified via one of the 2250 legacy driver-specific options takes precedence. 2251 2252 iommu.passthrough= 2253 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 2254 Format: { "0" | "1" } 2255 0 - Use IOMMU translation for DMA. 2256 1 - Bypass the IOMMU for DMA. 2257 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2258 2259 io7= [HW] IO7 for Marvel-based Alpha systems 2260 See comment before marvel_specify_io7 in 2261 arch/alpha/kernel/core_marvel.c. 2262 2263 io_delay= [X86] I/O delay method 2264 0x80 2265 Standard port 0x80 based delay 2266 0xed 2267 Alternate port 0xed based delay (needed on some systems) 2268 udelay 2269 Simple two microseconds delay 2270 none 2271 No delay 2272 2273 ip= [IP_PNP] 2274 See Documentation/admin-guide/nfs/nfsroot.rst. 2275 2276 ipcmni_extend [KNL] Extend the maximum number of unique System V 2277 IPC identifiers from 32,768 to 16,777,216. 2278 2279 irqaffinity= [SMP] Set the default irq affinity mask 2280 The argument is a cpu list, as described above. 2281 2282 irqchip.gicv2_force_probe= 2283 [ARM, ARM64] 2284 Format: <bool> 2285 Force the kernel to look for the second 4kB page 2286 of a GICv2 controller even if the memory range 2287 exposed by the device tree is too small. 2288 2289 irqchip.gicv3_nolpi= 2290 [ARM, ARM64] 2291 Force the kernel to ignore the availability of 2292 LPIs (and by consequence ITSs). Intended for system 2293 that use the kernel as a bootloader, and thus want 2294 to let secondary kernels in charge of setting up 2295 LPIs. 2296 2297 irqchip.gicv3_pseudo_nmi= [ARM64] 2298 Enables support for pseudo-NMIs in the kernel. This 2299 requires the kernel to be built with 2300 CONFIG_ARM64_PSEUDO_NMI. 2301 2302 irqfixup [HW] 2303 When an interrupt is not handled search all handlers 2304 for it. Intended to get systems with badly broken 2305 firmware running. 2306 2307 irqpoll [HW] 2308 When an interrupt is not handled search all handlers 2309 for it. Also check all handlers each timer 2310 interrupt. Intended to get systems with badly broken 2311 firmware running. 2312 2313 isapnp= [ISAPNP] 2314 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2315 2316 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2317 [Deprecated - use cpusets instead] 2318 Format: [flag-list,]<cpu-list> 2319 2320 Specify one or more CPUs to isolate from disturbances 2321 specified in the flag list (default: domain): 2322 2323 nohz 2324 Disable the tick when a single task runs. 2325 2326 A residual 1Hz tick is offloaded to workqueues, which you 2327 need to affine to housekeeping through the global 2328 workqueue's affinity configured via the 2329 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2330 by using the 'domain' flag described below. 2331 2332 NOTE: by default the global workqueue runs on all CPUs, 2333 so to protect individual CPUs the 'cpumask' file has to 2334 be configured manually after bootup. 2335 2336 domain 2337 Isolate from the general SMP balancing and scheduling 2338 algorithms. Note that performing domain isolation this way 2339 is irreversible: it's not possible to bring back a CPU to 2340 the domains once isolated through isolcpus. It's strongly 2341 advised to use cpusets instead to disable scheduler load 2342 balancing through the "cpuset.sched_load_balance" file. 2343 It offers a much more flexible interface where CPUs can 2344 move in and out of an isolated set anytime. 2345 2346 You can move a process onto or off an "isolated" CPU via 2347 the CPU affinity syscalls or cpuset. 2348 <cpu number> begins at 0 and the maximum value is 2349 "number of CPUs in system - 1". 2350 2351 managed_irq 2352 2353 Isolate from being targeted by managed interrupts 2354 which have an interrupt mask containing isolated 2355 CPUs. The affinity of managed interrupts is 2356 handled by the kernel and cannot be changed via 2357 the /proc/irq/* interfaces. 2358 2359 This isolation is best effort and only effective 2360 if the automatically assigned interrupt mask of a 2361 device queue contains isolated and housekeeping 2362 CPUs. If housekeeping CPUs are online then such 2363 interrupts are directed to the housekeeping CPU 2364 so that IO submitted on the housekeeping CPU 2365 cannot disturb the isolated CPU. 2366 2367 If a queue's affinity mask contains only isolated 2368 CPUs then this parameter has no effect on the 2369 interrupt routing decision, though interrupts are 2370 only delivered when tasks running on those 2371 isolated CPUs submit IO. IO submitted on 2372 housekeeping CPUs has no influence on those 2373 queues. 2374 2375 The format of <cpu-list> is described above. 2376 2377 iucv= [HW,NET] 2378 2379 ivrs_ioapic [HW,X86-64] 2380 Provide an override to the IOAPIC-ID<->DEVICE-ID 2381 mapping provided in the IVRS ACPI table. 2382 By default, PCI segment is 0, and can be omitted. 2383 2384 For example, to map IOAPIC-ID decimal 10 to 2385 PCI segment 0x1 and PCI device 00:14.0, 2386 write the parameter as: 2387 ivrs_ioapic=10@0001:00:14.0 2388 2389 Deprecated formats: 2390 * To map IOAPIC-ID decimal 10 to PCI device 00:14.0 2391 write the parameter as: 2392 ivrs_ioapic[10]=00:14.0 2393 * To map IOAPIC-ID decimal 10 to PCI segment 0x1 and 2394 PCI device 00:14.0 write the parameter as: 2395 ivrs_ioapic[10]=0001:00:14.0 2396 2397 ivrs_hpet [HW,X86-64] 2398 Provide an override to the HPET-ID<->DEVICE-ID 2399 mapping provided in the IVRS ACPI table. 2400 By default, PCI segment is 0, and can be omitted. 2401 2402 For example, to map HPET-ID decimal 10 to 2403 PCI segment 0x1 and PCI device 00:14.0, 2404 write the parameter as: 2405 ivrs_hpet=10@0001:00:14.0 2406 2407 Deprecated formats: 2408 * To map HPET-ID decimal 0 to PCI device 00:14.0 2409 write the parameter as: 2410 ivrs_hpet[0]=00:14.0 2411 * To map HPET-ID decimal 10 to PCI segment 0x1 and 2412 PCI device 00:14.0 write the parameter as: 2413 ivrs_ioapic[10]=0001:00:14.0 2414 2415 ivrs_acpihid [HW,X86-64] 2416 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2417 mapping provided in the IVRS ACPI table. 2418 By default, PCI segment is 0, and can be omitted. 2419 2420 For example, to map UART-HID:UID AMD0020:0 to 2421 PCI segment 0x1 and PCI device ID 00:14.5, 2422 write the parameter as: 2423 ivrs_acpihid=AMD0020:0@0001:00:14.5 2424 2425 Deprecated formats: 2426 * To map UART-HID:UID AMD0020:0 to PCI segment is 0, 2427 PCI device ID 00:14.5, write the parameter as: 2428 ivrs_acpihid[00:14.5]=AMD0020:0 2429 * To map UART-HID:UID AMD0020:0 to PCI segment 0x1 and 2430 PCI device ID 00:14.5, write the parameter as: 2431 ivrs_acpihid[0001:00:14.5]=AMD0020:0 2432 2433 js= [HW,JOY] Analog joystick 2434 See Documentation/input/joydev/joystick.rst. 2435 2436 kasan_multi_shot 2437 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2438 report on every invalid memory access. Without this 2439 parameter KASAN will print report only for the first 2440 invalid access. 2441 2442 keep_bootcon [KNL] 2443 Do not unregister boot console at start. This is only 2444 useful for debugging when something happens in the window 2445 between unregistering the boot console and initializing 2446 the real console. 2447 2448 keepinitrd [HW,ARM] 2449 2450 kernelcore= [KNL,X86,IA-64,PPC] 2451 Format: nn[KMGTPE] | nn% | "mirror" 2452 This parameter specifies the amount of memory usable by 2453 the kernel for non-movable allocations. The requested 2454 amount is spread evenly throughout all nodes in the 2455 system as ZONE_NORMAL. The remaining memory is used for 2456 movable memory in its own zone, ZONE_MOVABLE. In the 2457 event, a node is too small to have both ZONE_NORMAL and 2458 ZONE_MOVABLE, kernelcore memory will take priority and 2459 other nodes will have a larger ZONE_MOVABLE. 2460 2461 ZONE_MOVABLE is used for the allocation of pages that 2462 may be reclaimed or moved by the page migration 2463 subsystem. Note that allocations like PTEs-from-HighMem 2464 still use the HighMem zone if it exists, and the Normal 2465 zone if it does not. 2466 2467 It is possible to specify the exact amount of memory in 2468 the form of "nn[KMGTPE]", a percentage of total system 2469 memory in the form of "nn%", or "mirror". If "mirror" 2470 option is specified, mirrored (reliable) memory is used 2471 for non-movable allocations and remaining memory is used 2472 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2473 are exclusive, so you cannot specify multiple forms. 2474 2475 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2476 Format: <Controller#>[,poll interval] 2477 The controller # is the number of the ehci usb debug 2478 port as it is probed via PCI. The poll interval is 2479 optional and is the number seconds in between 2480 each poll cycle to the debug port in case you need 2481 the functionality for interrupting the kernel with 2482 gdb or control-c on the dbgp connection. When 2483 not using this parameter you use sysrq-g to break into 2484 the kernel debugger. 2485 2486 kgdboc= [KGDB,HW] kgdb over consoles. 2487 Requires a tty driver that supports console polling, 2488 or a supported polling keyboard driver (non-usb). 2489 Serial only format: <serial_device>[,baud] 2490 keyboard only format: kbd 2491 keyboard and serial format: kbd,<serial_device>[,baud] 2492 Optional Kernel mode setting: 2493 kms, kbd format: kms,kbd 2494 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2495 2496 kgdboc_earlycon= [KGDB,HW] 2497 If the boot console provides the ability to read 2498 characters and can work in polling mode, you can use 2499 this parameter to tell kgdb to use it as a backend 2500 until the normal console is registered. Intended to 2501 be used together with the kgdboc parameter which 2502 specifies the normal console to transition to. 2503 2504 The name of the early console should be specified 2505 as the value of this parameter. Note that the name of 2506 the early console might be different than the tty 2507 name passed to kgdboc. It's OK to leave the value 2508 blank and the first boot console that implements 2509 read() will be picked. 2510 2511 kgdbwait [KGDB] Stop kernel execution and enter the 2512 kernel debugger at the earliest opportunity. 2513 2514 kmac= [MIPS] Korina ethernet MAC address. 2515 Configure the RouterBoard 532 series on-chip 2516 Ethernet adapter MAC address. 2517 2518 kmemleak= [KNL] Boot-time kmemleak enable/disable 2519 Valid arguments: on, off 2520 Default: on 2521 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2522 the default is off. 2523 2524 kprobe_event=[probe-list] 2525 [FTRACE] Add kprobe events and enable at boot time. 2526 The probe-list is a semicolon delimited list of probe 2527 definitions. Each definition is same as kprobe_events 2528 interface, but the parameters are comma delimited. 2529 For example, to add a kprobe event on vfs_read with 2530 arg1 and arg2, add to the command line; 2531 2532 kprobe_event=p,vfs_read,$arg1,$arg2 2533 2534 See also Documentation/trace/kprobetrace.rst "Kernel 2535 Boot Parameter" section. 2536 2537 kpti= [ARM64] Control page table isolation of user 2538 and kernel address spaces. 2539 Default: enabled on cores which need mitigation. 2540 0: force disabled 2541 1: force enabled 2542 2543 kunit.enable= [KUNIT] Enable executing KUnit tests. Requires 2544 CONFIG_KUNIT to be set to be fully enabled. The 2545 default value can be overridden via 2546 KUNIT_DEFAULT_ENABLED. 2547 Default is 1 (enabled) 2548 2549 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2550 Default is 0 (don't ignore, but inject #GP) 2551 2552 kvm.eager_page_split= 2553 [KVM,X86] Controls whether or not KVM will try to 2554 proactively split all huge pages during dirty logging. 2555 Eager page splitting reduces interruptions to vCPU 2556 execution by eliminating the write-protection faults 2557 and MMU lock contention that would otherwise be 2558 required to split huge pages lazily. 2559 2560 VM workloads that rarely perform writes or that write 2561 only to a small region of VM memory may benefit from 2562 disabling eager page splitting to allow huge pages to 2563 still be used for reads. 2564 2565 The behavior of eager page splitting depends on whether 2566 KVM_DIRTY_LOG_INITIALLY_SET is enabled or disabled. If 2567 disabled, all huge pages in a memslot will be eagerly 2568 split when dirty logging is enabled on that memslot. If 2569 enabled, eager page splitting will be performed during 2570 the KVM_CLEAR_DIRTY ioctl, and only for the pages being 2571 cleared. 2572 2573 Eager page splitting is only supported when kvm.tdp_mmu=Y. 2574 2575 Default is Y (on). 2576 2577 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2578 Default is false (don't support). 2579 2580 kvm.nx_huge_pages= 2581 [KVM] Controls the software workaround for the 2582 X86_BUG_ITLB_MULTIHIT bug. 2583 force : Always deploy workaround. 2584 off : Never deploy workaround. 2585 auto : Deploy workaround based on the presence of 2586 X86_BUG_ITLB_MULTIHIT. 2587 2588 Default is 'auto'. 2589 2590 If the software workaround is enabled for the host, 2591 guests do need not to enable it for nested guests. 2592 2593 kvm.nx_huge_pages_recovery_ratio= 2594 [KVM] Controls how many 4KiB pages are periodically zapped 2595 back to huge pages. 0 disables the recovery, otherwise if 2596 the value is N KVM will zap 1/Nth of the 4KiB pages every 2597 period (see below). The default is 60. 2598 2599 kvm.nx_huge_pages_recovery_period_ms= 2600 [KVM] Controls the time period at which KVM zaps 4KiB pages 2601 back to huge pages. If the value is a non-zero N, KVM will 2602 zap a portion (see ratio above) of the pages every N msecs. 2603 If the value is 0 (the default), KVM will pick a period based 2604 on the ratio, such that a page is zapped after 1 hour on average. 2605 2606 kvm-amd.nested= [KVM,AMD] Control nested virtualization feature in 2607 KVM/SVM. Default is 1 (enabled). 2608 2609 kvm-amd.npt= [KVM,AMD] Control KVM's use of Nested Page Tables, 2610 a.k.a. Two-Dimensional Page Tables. Default is 1 2611 (enabled). Disable by KVM if hardware lacks support 2612 for NPT. 2613 2614 kvm-arm.mode= 2615 [KVM,ARM] Select one of KVM/arm64's modes of operation. 2616 2617 none: Forcefully disable KVM. 2618 2619 nvhe: Standard nVHE-based mode, without support for 2620 protected guests. 2621 2622 protected: nVHE-based mode with support for guests whose 2623 state is kept private from the host. 2624 2625 nested: VHE-based mode with support for nested 2626 virtualization. Requires at least ARMv8.3 2627 hardware. 2628 2629 Defaults to VHE/nVHE based on hardware support. Setting 2630 mode to "protected" will disable kexec and hibernation 2631 for the host. "nested" is experimental and should be 2632 used with extreme caution. 2633 2634 kvm-arm.vgic_v3_group0_trap= 2635 [KVM,ARM] Trap guest accesses to GICv3 group-0 2636 system registers 2637 2638 kvm-arm.vgic_v3_group1_trap= 2639 [KVM,ARM] Trap guest accesses to GICv3 group-1 2640 system registers 2641 2642 kvm-arm.vgic_v3_common_trap= 2643 [KVM,ARM] Trap guest accesses to GICv3 common 2644 system registers 2645 2646 kvm-arm.vgic_v4_enable= 2647 [KVM,ARM] Allow use of GICv4 for direct injection of 2648 LPIs. 2649 2650 kvm_cma_resv_ratio=n [PPC] 2651 Reserves given percentage from system memory area for 2652 contiguous memory allocation for KVM hash pagetable 2653 allocation. 2654 By default it reserves 5% of total system memory. 2655 Format: <integer> 2656 Default: 5 2657 2658 kvm-intel.ept= [KVM,Intel] Control KVM's use of Extended Page Tables, 2659 a.k.a. Two-Dimensional Page Tables. Default is 1 2660 (enabled). Disable by KVM if hardware lacks support 2661 for EPT. 2662 2663 kvm-intel.emulate_invalid_guest_state= 2664 [KVM,Intel] Control whether to emulate invalid guest 2665 state. Ignored if kvm-intel.enable_unrestricted_guest=1, 2666 as guest state is never invalid for unrestricted 2667 guests. This param doesn't apply to nested guests (L2), 2668 as KVM never emulates invalid L2 guest state. 2669 Default is 1 (enabled). 2670 2671 kvm-intel.flexpriority= 2672 [KVM,Intel] Control KVM's use of FlexPriority feature 2673 (TPR shadow). Default is 1 (enabled). Disable by KVM if 2674 hardware lacks support for it. 2675 2676 kvm-intel.nested= 2677 [KVM,Intel] Control nested virtualization feature in 2678 KVM/VMX. Default is 1 (enabled). 2679 2680 kvm-intel.unrestricted_guest= 2681 [KVM,Intel] Control KVM's use of unrestricted guest 2682 feature (virtualized real and unpaged mode). Default 2683 is 1 (enabled). Disable by KVM if EPT is disabled or 2684 hardware lacks support for it. 2685 2686 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2687 CVE-2018-3620. 2688 2689 Valid arguments: never, cond, always 2690 2691 always: L1D cache flush on every VMENTER. 2692 cond: Flush L1D on VMENTER only when the code between 2693 VMEXIT and VMENTER can leak host memory. 2694 never: Disables the mitigation 2695 2696 Default is cond (do L1 cache flush in specific instances) 2697 2698 kvm-intel.vpid= [KVM,Intel] Control KVM's use of Virtual Processor 2699 Identification feature (tagged TLBs). Default is 1 2700 (enabled). Disable by KVM if hardware lacks support 2701 for it. 2702 2703 l1d_flush= [X86,INTEL] 2704 Control mitigation for L1D based snooping vulnerability. 2705 2706 Certain CPUs are vulnerable to an exploit against CPU 2707 internal buffers which can forward information to a 2708 disclosure gadget under certain conditions. 2709 2710 In vulnerable processors, the speculatively 2711 forwarded data can be used in a cache side channel 2712 attack, to access data to which the attacker does 2713 not have direct access. 2714 2715 This parameter controls the mitigation. The 2716 options are: 2717 2718 on - enable the interface for the mitigation 2719 2720 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2721 affected CPUs 2722 2723 The kernel PTE inversion protection is unconditionally 2724 enabled and cannot be disabled. 2725 2726 full 2727 Provides all available mitigations for the 2728 L1TF vulnerability. Disables SMT and 2729 enables all mitigations in the 2730 hypervisors, i.e. unconditional L1D flush. 2731 2732 SMT control and L1D flush control via the 2733 sysfs interface is still possible after 2734 boot. Hypervisors will issue a warning 2735 when the first VM is started in a 2736 potentially insecure configuration, 2737 i.e. SMT enabled or L1D flush disabled. 2738 2739 full,force 2740 Same as 'full', but disables SMT and L1D 2741 flush runtime control. Implies the 2742 'nosmt=force' command line option. 2743 (i.e. sysfs control of SMT is disabled.) 2744 2745 flush 2746 Leaves SMT enabled and enables the default 2747 hypervisor mitigation, i.e. conditional 2748 L1D flush. 2749 2750 SMT control and L1D flush control via the 2751 sysfs interface is still possible after 2752 boot. Hypervisors will issue a warning 2753 when the first VM is started in a 2754 potentially insecure configuration, 2755 i.e. SMT enabled or L1D flush disabled. 2756 2757 flush,nosmt 2758 2759 Disables SMT and enables the default 2760 hypervisor mitigation. 2761 2762 SMT control and L1D flush control via the 2763 sysfs interface is still possible after 2764 boot. Hypervisors will issue a warning 2765 when the first VM is started in a 2766 potentially insecure configuration, 2767 i.e. SMT enabled or L1D flush disabled. 2768 2769 flush,nowarn 2770 Same as 'flush', but hypervisors will not 2771 warn when a VM is started in a potentially 2772 insecure configuration. 2773 2774 off 2775 Disables hypervisor mitigations and doesn't 2776 emit any warnings. 2777 It also drops the swap size and available 2778 RAM limit restriction on both hypervisor and 2779 bare metal. 2780 2781 Default is 'flush'. 2782 2783 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2784 2785 l2cr= [PPC] 2786 2787 l3cr= [PPC] 2788 2789 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2790 disabled it. 2791 2792 lapic= [X86,APIC] Do not use TSC deadline 2793 value for LAPIC timer one-shot implementation. Default 2794 back to the programmable timer unit in the LAPIC. 2795 Format: notscdeadline 2796 2797 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2798 in C2 power state. 2799 2800 libata.dma= [LIBATA] DMA control 2801 libata.dma=0 Disable all PATA and SATA DMA 2802 libata.dma=1 PATA and SATA Disk DMA only 2803 libata.dma=2 ATAPI (CDROM) DMA only 2804 libata.dma=4 Compact Flash DMA only 2805 Combinations also work, so libata.dma=3 enables DMA 2806 for disks and CDROMs, but not CFs. 2807 2808 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2809 libata.ignore_hpa=0 keep BIOS limits (default) 2810 libata.ignore_hpa=1 ignore limits, using full disk 2811 2812 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2813 when set. 2814 Format: <int> 2815 2816 libata.force= [LIBATA] Force configurations. The format is a comma- 2817 separated list of "[ID:]VAL" where ID is PORT[.DEVICE]. 2818 PORT and DEVICE are decimal numbers matching port, link 2819 or device. Basically, it matches the ATA ID string 2820 printed on console by libata. If the whole ID part is 2821 omitted, the last PORT and DEVICE values are used. If 2822 ID hasn't been specified yet, the configuration applies 2823 to all ports, links and devices. 2824 2825 If only DEVICE is omitted, the parameter applies to 2826 the port and all links and devices behind it. DEVICE 2827 number of 0 either selects the first device or the 2828 first fan-out link behind PMP device. It does not 2829 select the host link. DEVICE number of 15 selects the 2830 host link and device attached to it. 2831 2832 The VAL specifies the configuration to force. As long 2833 as there is no ambiguity, shortcut notation is allowed. 2834 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2835 The following configurations can be forced. 2836 2837 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2838 Any ID with matching PORT is used. 2839 2840 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2841 2842 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2843 udma[/][16,25,33,44,66,100,133] notation is also 2844 allowed. 2845 2846 * nohrst, nosrst, norst: suppress hard, soft and both 2847 resets. 2848 2849 * rstonce: only attempt one reset during hot-unplug 2850 link recovery. 2851 2852 * [no]dbdelay: Enable or disable the extra 200ms delay 2853 before debouncing a link PHY and device presence 2854 detection. 2855 2856 * [no]ncq: Turn on or off NCQ. 2857 2858 * [no]ncqtrim: Enable or disable queued DSM TRIM. 2859 2860 * [no]ncqati: Enable or disable NCQ trim on ATI chipset. 2861 2862 * [no]trim: Enable or disable (unqueued) TRIM. 2863 2864 * trim_zero: Indicate that TRIM command zeroes data. 2865 2866 * max_trim_128m: Set 128M maximum trim size limit. 2867 2868 * [no]dma: Turn on or off DMA transfers. 2869 2870 * atapi_dmadir: Enable ATAPI DMADIR bridge support. 2871 2872 * atapi_mod16_dma: Enable the use of ATAPI DMA for 2873 commands that are not a multiple of 16 bytes. 2874 2875 * [no]dmalog: Enable or disable the use of the 2876 READ LOG DMA EXT command to access logs. 2877 2878 * [no]iddevlog: Enable or disable access to the 2879 identify device data log. 2880 2881 * [no]logdir: Enable or disable access to the general 2882 purpose log directory. 2883 2884 * max_sec_128: Set transfer size limit to 128 sectors. 2885 2886 * max_sec_1024: Set or clear transfer size limit to 2887 1024 sectors. 2888 2889 * max_sec_lba48: Set or clear transfer size limit to 2890 65535 sectors. 2891 2892 * [no]lpm: Enable or disable link power management. 2893 2894 * [no]setxfer: Indicate if transfer speed mode setting 2895 should be skipped. 2896 2897 * [no]fua: Disable or enable FUA (Force Unit Access) 2898 support for devices supporting this feature. 2899 2900 * dump_id: Dump IDENTIFY data. 2901 2902 * disable: Disable this device. 2903 2904 If there are multiple matching configurations changing 2905 the same attribute, the last one is used. 2906 2907 load_ramdisk= [RAM] [Deprecated] 2908 2909 lockd.nlm_grace_period=P [NFS] Assign grace period. 2910 Format: <integer> 2911 2912 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2913 Format: <integer> 2914 2915 lockd.nlm_timeout=T [NFS] Assign timeout value. 2916 Format: <integer> 2917 2918 lockd.nlm_udpport=M [NFS] Assign UDP port. 2919 Format: <integer> 2920 2921 lockdown= [SECURITY] 2922 { integrity | confidentiality } 2923 Enable the kernel lockdown feature. If set to 2924 integrity, kernel features that allow userland to 2925 modify the running kernel are disabled. If set to 2926 confidentiality, kernel features that allow userland 2927 to extract confidential information from the kernel 2928 are also disabled. 2929 2930 locktorture.nreaders_stress= [KNL] 2931 Set the number of locking read-acquisition kthreads. 2932 Defaults to being automatically set based on the 2933 number of online CPUs. 2934 2935 locktorture.nwriters_stress= [KNL] 2936 Set the number of locking write-acquisition kthreads. 2937 2938 locktorture.onoff_holdoff= [KNL] 2939 Set time (s) after boot for CPU-hotplug testing. 2940 2941 locktorture.onoff_interval= [KNL] 2942 Set time (s) between CPU-hotplug operations, or 2943 zero to disable CPU-hotplug testing. 2944 2945 locktorture.shuffle_interval= [KNL] 2946 Set task-shuffle interval (jiffies). Shuffling 2947 tasks allows some CPUs to go into dyntick-idle 2948 mode during the locktorture test. 2949 2950 locktorture.shutdown_secs= [KNL] 2951 Set time (s) after boot system shutdown. This 2952 is useful for hands-off automated testing. 2953 2954 locktorture.stat_interval= [KNL] 2955 Time (s) between statistics printk()s. 2956 2957 locktorture.stutter= [KNL] 2958 Time (s) to stutter testing, for example, 2959 specifying five seconds causes the test to run for 2960 five seconds, wait for five seconds, and so on. 2961 This tests the locking primitive's ability to 2962 transition abruptly to and from idle. 2963 2964 locktorture.torture_type= [KNL] 2965 Specify the locking implementation to test. 2966 2967 locktorture.writer_fifo= [KNL] 2968 Run the write-side locktorture kthreads at 2969 sched_set_fifo() real-time priority. 2970 2971 locktorture.verbose= [KNL] 2972 Enable additional printk() statements. 2973 2974 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2975 Format: <irq> 2976 2977 loglevel= All Kernel Messages with a loglevel smaller than the 2978 console loglevel will be printed to the console. It can 2979 also be changed with klogd or other programs. The 2980 loglevels are defined as follows: 2981 2982 0 (KERN_EMERG) system is unusable 2983 1 (KERN_ALERT) action must be taken immediately 2984 2 (KERN_CRIT) critical conditions 2985 3 (KERN_ERR) error conditions 2986 4 (KERN_WARNING) warning conditions 2987 5 (KERN_NOTICE) normal but significant condition 2988 6 (KERN_INFO) informational 2989 7 (KERN_DEBUG) debug-level messages 2990 2991 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2992 in bytes. n must be a power of two and greater 2993 than the minimal size. The minimal size is defined 2994 by LOG_BUF_SHIFT kernel config parameter. There is 2995 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2996 that allows to increase the default size depending on 2997 the number of CPUs. See init/Kconfig for more details. 2998 2999 logo.nologo [FB] Disables display of the built-in Linux logo. 3000 This may be used to provide more screen space for 3001 kernel log messages and is useful when debugging 3002 kernel boot problems. 3003 3004 lp=0 [LP] Specify parallel ports to use, e.g, 3005 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 3006 lp=reset first parallel port). 'lp=0' disables the 3007 lp=auto printer driver. 'lp=reset' (which can be 3008 specified in addition to the ports) causes 3009 attached printers to be reset. Using 3010 lp=port1,port2,... specifies the parallel ports 3011 to associate lp devices with, starting with 3012 lp0. A port specification may be 'none' to skip 3013 that lp device, or a parport name such as 3014 'parport0'. Specifying 'lp=auto' instead of a 3015 port specification list means that device IDs 3016 from each port should be examined, to see if 3017 an IEEE 1284-compliant printer is attached; if 3018 so, the driver will manage that printer. 3019 See also header of drivers/char/lp.c. 3020 3021 lpj=n [KNL] 3022 Sets loops_per_jiffy to given constant, thus avoiding 3023 time-consuming boot-time autodetection (up to 250 ms per 3024 CPU). 0 enables autodetection (default). To determine 3025 the correct value for your kernel, boot with normal 3026 autodetection and see what value is printed. Note that 3027 on SMP systems the preset will be applied to all CPUs, 3028 which is likely to cause problems if your CPUs need 3029 significantly divergent settings. An incorrect value 3030 will cause delays in the kernel to be wrong, leading to 3031 unpredictable I/O errors and other breakage. Although 3032 unlikely, in the extreme case this might damage your 3033 hardware. 3034 3035 ltpc= [NET] 3036 Format: <io>,<irq>,<dma> 3037 3038 lsm.debug [SECURITY] Enable LSM initialization debugging output. 3039 3040 lsm=lsm1,...,lsmN 3041 [SECURITY] Choose order of LSM initialization. This 3042 overrides CONFIG_LSM, and the "security=" parameter. 3043 3044 machvec= [IA-64] Force the use of a particular machine-vector 3045 (machvec) in a generic kernel. 3046 Example: machvec=hpzx1 3047 3048 machtype= [Loongson] Share the same kernel image file between 3049 different yeeloong laptops. 3050 Example: machtype=lemote-yeeloong-2f-7inch 3051 3052 max_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory greater 3053 than or equal to this physical address is ignored. 3054 3055 maxcpus= [SMP] Maximum number of processors that an SMP kernel 3056 will bring up during bootup. maxcpus=n : n >= 0 limits 3057 the kernel to bring up 'n' processors. Surely after 3058 bootup you can bring up the other plugged cpu by executing 3059 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 3060 only takes effect during system bootup. 3061 While n=0 is a special case, it is equivalent to "nosmp", 3062 which also disables the IO APIC. 3063 3064 max_loop= [LOOP] The number of loop block devices that get 3065 (loop.max_loop) unconditionally pre-created at init time. The default 3066 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 3067 of statically allocating a predefined number, loop 3068 devices can be requested on-demand with the 3069 /dev/loop-control interface. 3070 3071 mce [X86-32] Machine Check Exception 3072 3073 mce=option [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 3074 3075 md= [HW] RAID subsystems devices and level 3076 See Documentation/admin-guide/md.rst. 3077 3078 mdacon= [MDA] 3079 Format: <first>,<last> 3080 Specifies range of consoles to be captured by the MDA. 3081 3082 mds= [X86,INTEL] 3083 Control mitigation for the Micro-architectural Data 3084 Sampling (MDS) vulnerability. 3085 3086 Certain CPUs are vulnerable to an exploit against CPU 3087 internal buffers which can forward information to a 3088 disclosure gadget under certain conditions. 3089 3090 In vulnerable processors, the speculatively 3091 forwarded data can be used in a cache side channel 3092 attack, to access data to which the attacker does 3093 not have direct access. 3094 3095 This parameter controls the MDS mitigation. The 3096 options are: 3097 3098 full - Enable MDS mitigation on vulnerable CPUs 3099 full,nosmt - Enable MDS mitigation and disable 3100 SMT on vulnerable CPUs 3101 off - Unconditionally disable MDS mitigation 3102 3103 On TAA-affected machines, mds=off can be prevented by 3104 an active TAA mitigation as both vulnerabilities are 3105 mitigated with the same mechanism so in order to disable 3106 this mitigation, you need to specify tsx_async_abort=off 3107 too. 3108 3109 Not specifying this option is equivalent to 3110 mds=full. 3111 3112 For details see: Documentation/admin-guide/hw-vuln/mds.rst 3113 3114 mem=nn[KMG] [HEXAGON] Set the memory size. 3115 Must be specified, otherwise memory size will be 0. 3116 3117 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 3118 Amount of memory to be used in cases as follows: 3119 3120 1 for test; 3121 2 when the kernel is not able to see the whole system memory; 3122 3 memory that lies after 'mem=' boundary is excluded from 3123 the hypervisor, then assigned to KVM guests. 3124 4 to limit the memory available for kdump kernel. 3125 3126 [ARC,MICROBLAZE] - the limit applies only to low memory, 3127 high memory is not affected. 3128 3129 [ARM64] - only limits memory covered by the linear 3130 mapping. The NOMAP regions are not affected. 3131 3132 [X86] Work as limiting max address. Use together 3133 with memmap= to avoid physical address space collisions. 3134 Without memmap= PCI devices could be placed at addresses 3135 belonging to unused RAM. 3136 3137 Note that this only takes effects during boot time since 3138 in above case 3, memory may need be hot added after boot 3139 if system memory of hypervisor is not sufficient. 3140 3141 mem=nn[KMG]@ss[KMG] 3142 [ARM,MIPS] - override the memory layout reported by 3143 firmware. 3144 Define a memory region of size nn[KMG] starting at 3145 ss[KMG]. 3146 Multiple different regions can be specified with 3147 multiple mem= parameters on the command line. 3148 3149 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 3150 memory. 3151 3152 memblock=debug [KNL] Enable memblock debug messages. 3153 3154 memchunk=nn[KMG] 3155 [KNL,SH] Allow user to override the default size for 3156 per-device physically contiguous DMA buffers. 3157 3158 memhp_default_state=online/offline/online_kernel/online_movable 3159 [KNL] Set the initial state for the memory hotplug 3160 onlining policy. If not specified, the default value is 3161 set according to the 3162 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 3163 option. 3164 See Documentation/admin-guide/mm/memory-hotplug.rst. 3165 3166 memmap=exactmap [KNL,X86] Enable setting of an exact 3167 E820 memory map, as specified by the user. 3168 Such memmap=exactmap lines can be constructed based on 3169 BIOS output or other requirements. See the memmap=nn@ss 3170 option description. 3171 3172 memmap=nn[KMG]@ss[KMG] 3173 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory. 3174 Region of memory to be used is from ss to ss+nn. 3175 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 3176 which limits max address to nn[KMG]. 3177 Multiple different regions can be specified, 3178 comma delimited. 3179 Example: 3180 memmap=100M@2G,100M#3G,1G!1024G 3181 3182 memmap=nn[KMG]#ss[KMG] 3183 [KNL,ACPI] Mark specific memory as ACPI data. 3184 Region of memory to be marked is from ss to ss+nn. 3185 3186 memmap=nn[KMG]$ss[KMG] 3187 [KNL,ACPI] Mark specific memory as reserved. 3188 Region of memory to be reserved is from ss to ss+nn. 3189 Example: Exclude memory from 0x18690000-0x1869ffff 3190 memmap=64K$0x18690000 3191 or 3192 memmap=0x10000$0x18690000 3193 Some bootloaders may need an escape character before '$', 3194 like Grub2, otherwise '$' and the following number 3195 will be eaten. 3196 3197 memmap=nn[KMG]!ss[KMG] 3198 [KNL,X86] Mark specific memory as protected. 3199 Region of memory to be used, from ss to ss+nn. 3200 The memory region may be marked as e820 type 12 (0xc) 3201 and is NVDIMM or ADR memory. 3202 3203 memmap=<size>%<offset>-<oldtype>+<newtype> 3204 [KNL,ACPI] Convert memory within the specified region 3205 from <oldtype> to <newtype>. If "-<oldtype>" is left 3206 out, the whole region will be marked as <newtype>, 3207 even if previously unavailable. If "+<newtype>" is left 3208 out, matching memory will be removed. Types are 3209 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 3210 3 = ACPI, 12 = PRAM. 3211 3212 memory_corruption_check=0/1 [X86] 3213 Some BIOSes seem to corrupt the first 64k of 3214 memory when doing things like suspend/resume. 3215 Setting this option will scan the memory 3216 looking for corruption. Enabling this will 3217 both detect corruption and prevent the kernel 3218 from using the memory being corrupted. 3219 However, its intended as a diagnostic tool; if 3220 repeatable BIOS-originated corruption always 3221 affects the same memory, you can use memmap= 3222 to prevent the kernel from using that memory. 3223 3224 memory_corruption_check_size=size [X86] 3225 By default it checks for corruption in the low 3226 64k, making this memory unavailable for normal 3227 use. Use this parameter to scan for 3228 corruption in more or less memory. 3229 3230 memory_corruption_check_period=seconds [X86] 3231 By default it checks for corruption every 60 3232 seconds. Use this parameter to check at some 3233 other rate. 0 disables periodic checking. 3234 3235 memory_hotplug.memmap_on_memory 3236 [KNL,X86,ARM] Boolean flag to enable this feature. 3237 Format: {on | off (default)} 3238 When enabled, runtime hotplugged memory will 3239 allocate its internal metadata (struct pages, 3240 those vmemmap pages cannot be optimized even 3241 if hugetlb_free_vmemmap is enabled) from the 3242 hotadded memory which will allow to hotadd a 3243 lot of memory without requiring additional 3244 memory to do so. 3245 This feature is disabled by default because it 3246 has some implication on large (e.g. GB) 3247 allocations in some configurations (e.g. small 3248 memory blocks). 3249 The state of the flag can be read in 3250 /sys/module/memory_hotplug/parameters/memmap_on_memory. 3251 Note that even when enabled, there are a few cases where 3252 the feature is not effective. 3253 3254 memtest= [KNL,X86,ARM,M68K,PPC,RISCV] Enable memtest 3255 Format: <integer> 3256 default : 0 <disable> 3257 Specifies the number of memtest passes to be 3258 performed. Each pass selects another test 3259 pattern from a given set of patterns. Memtest 3260 fills the memory with this pattern, validates 3261 memory contents and reserves bad memory 3262 regions that are detected. 3263 3264 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 3265 Valid arguments: on, off 3266 Default: off 3267 mem_encrypt=on: Activate SME 3268 mem_encrypt=off: Do not activate SME 3269 3270 Refer to Documentation/virt/kvm/x86/amd-memory-encryption.rst 3271 for details on when memory encryption can be activated. 3272 3273 mem_sleep_default= [SUSPEND] Default system suspend mode: 3274 s2idle - Suspend-To-Idle 3275 shallow - Power-On Suspend or equivalent (if supported) 3276 deep - Suspend-To-RAM or equivalent (if supported) 3277 See Documentation/admin-guide/pm/sleep-states.rst. 3278 3279 mfgpt_irq= [IA-32] Specify the IRQ to use for the 3280 Multi-Function General Purpose Timers on AMD Geode 3281 platforms. 3282 3283 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 3284 the BIOS has incorrectly applied a workaround. TinyBIOS 3285 version 0.98 is known to be affected, 0.99 fixes the 3286 problem by letting the user disable the workaround. 3287 3288 mga= [HW,DRM] 3289 3290 microcode.force_minrev= [X86] 3291 Format: <bool> 3292 Enable or disable the microcode minimal revision 3293 enforcement for the runtime microcode loader. 3294 3295 min_addr=nn[KMG] [KNL,BOOT,IA-64] All physical memory below this 3296 physical address is ignored. 3297 3298 mini2440= [ARM,HW,KNL] 3299 Format:[0..2][b][c][t] 3300 Default: "0tb" 3301 MINI2440 configuration specification: 3302 0 - The attached screen is the 3.5" TFT 3303 1 - The attached screen is the 7" TFT 3304 2 - The VGA Shield is attached (1024x768) 3305 Leaving out the screen size parameter will not load 3306 the TFT driver, and the framebuffer will be left 3307 unconfigured. 3308 b - Enable backlight. The TFT backlight pin will be 3309 linked to the kernel VESA blanking code and a GPIO 3310 LED. This parameter is not necessary when using the 3311 VGA shield. 3312 c - Enable the s3c camera interface. 3313 t - Reserved for enabling touchscreen support. The 3314 touchscreen support is not enabled in the mainstream 3315 kernel as of 2.6.30, a preliminary port can be found 3316 in the "bleeding edge" mini2440 support kernel at 3317 https://repo.or.cz/w/linux-2.6/mini2440.git 3318 3319 mitigations= 3320 [X86,PPC,S390,ARM64] Control optional mitigations for 3321 CPU vulnerabilities. This is a set of curated, 3322 arch-independent options, each of which is an 3323 aggregation of existing arch-specific options. 3324 3325 Note, "mitigations" is supported if and only if the 3326 kernel was built with CPU_MITIGATIONS=y. 3327 3328 off 3329 Disable all optional CPU mitigations. This 3330 improves system performance, but it may also 3331 expose users to several CPU vulnerabilities. 3332 Equivalent to: if nokaslr then kpti=0 [ARM64] 3333 gather_data_sampling=off [X86] 3334 kvm.nx_huge_pages=off [X86] 3335 l1tf=off [X86] 3336 mds=off [X86] 3337 mmio_stale_data=off [X86] 3338 no_entry_flush [PPC] 3339 no_uaccess_flush [PPC] 3340 nobp=0 [S390] 3341 nopti [X86,PPC] 3342 nospectre_bhb [ARM64] 3343 nospectre_v1 [X86,PPC] 3344 nospectre_v2 [X86,PPC,S390,ARM64] 3345 reg_file_data_sampling=off [X86] 3346 retbleed=off [X86] 3347 spec_store_bypass_disable=off [X86,PPC] 3348 spectre_bhi=off [X86] 3349 spectre_v2_user=off [X86] 3350 srbds=off [X86,INTEL] 3351 ssbd=force-off [ARM64] 3352 tsx_async_abort=off [X86] 3353 3354 Exceptions: 3355 This does not have any effect on 3356 kvm.nx_huge_pages when 3357 kvm.nx_huge_pages=force. 3358 3359 auto (default) 3360 Mitigate all CPU vulnerabilities, but leave SMT 3361 enabled, even if it's vulnerable. This is for 3362 users who don't want to be surprised by SMT 3363 getting disabled across kernel upgrades, or who 3364 have other ways of avoiding SMT-based attacks. 3365 Equivalent to: (default behavior) 3366 3367 auto,nosmt 3368 Mitigate all CPU vulnerabilities, disabling SMT 3369 if needed. This is for users who always want to 3370 be fully mitigated, even if it means losing SMT. 3371 Equivalent to: l1tf=flush,nosmt [X86] 3372 mds=full,nosmt [X86] 3373 tsx_async_abort=full,nosmt [X86] 3374 mmio_stale_data=full,nosmt [X86] 3375 retbleed=auto,nosmt [X86] 3376 3377 mminit_loglevel= 3378 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 3379 parameter allows control of the logging verbosity for 3380 the additional memory initialisation checks. A value 3381 of 0 disables mminit logging and a level of 4 will 3382 log everything. Information is printed at KERN_DEBUG 3383 so loglevel=8 may also need to be specified. 3384 3385 mmio_stale_data= 3386 [X86,INTEL] Control mitigation for the Processor 3387 MMIO Stale Data vulnerabilities. 3388 3389 Processor MMIO Stale Data is a class of 3390 vulnerabilities that may expose data after an MMIO 3391 operation. Exposed data could originate or end in 3392 the same CPU buffers as affected by MDS and TAA. 3393 Therefore, similar to MDS and TAA, the mitigation 3394 is to clear the affected CPU buffers. 3395 3396 This parameter controls the mitigation. The 3397 options are: 3398 3399 full - Enable mitigation on vulnerable CPUs 3400 3401 full,nosmt - Enable mitigation and disable SMT on 3402 vulnerable CPUs. 3403 3404 off - Unconditionally disable mitigation 3405 3406 On MDS or TAA affected machines, 3407 mmio_stale_data=off can be prevented by an active 3408 MDS or TAA mitigation as these vulnerabilities are 3409 mitigated with the same mechanism so in order to 3410 disable this mitigation, you need to specify 3411 mds=off and tsx_async_abort=off too. 3412 3413 Not specifying this option is equivalent to 3414 mmio_stale_data=full. 3415 3416 For details see: 3417 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3418 3419 <module>.async_probe[=<bool>] [KNL] 3420 If no <bool> value is specified or if the value 3421 specified is not a valid <bool>, enable asynchronous 3422 probe on this module. Otherwise, enable/disable 3423 asynchronous probe on this module as indicated by the 3424 <bool> value. See also: module.async_probe 3425 3426 module.async_probe=<bool> 3427 [KNL] When set to true, modules will use async probing 3428 by default. To enable/disable async probing for a 3429 specific module, use the module specific control that 3430 is documented under <module>.async_probe. When both 3431 module.async_probe and <module>.async_probe are 3432 specified, <module>.async_probe takes precedence for 3433 the specific module. 3434 3435 module.enable_dups_trace 3436 [KNL] When CONFIG_MODULE_DEBUG_AUTOLOAD_DUPS is set, 3437 this means that duplicate request_module() calls will 3438 trigger a WARN_ON() instead of a pr_warn(). Note that 3439 if MODULE_DEBUG_AUTOLOAD_DUPS_TRACE is set, WARN_ON()s 3440 will always be issued and this option does nothing. 3441 module.sig_enforce 3442 [KNL] When CONFIG_MODULE_SIG is set, this means that 3443 modules without (valid) signatures will fail to load. 3444 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3445 is always true, so this option does nothing. 3446 3447 module_blacklist= [KNL] Do not load a comma-separated list of 3448 modules. Useful for debugging problem modules. 3449 3450 mousedev.tap_time= 3451 [MOUSE] Maximum time between finger touching and 3452 leaving touchpad surface for touch to be considered 3453 a tap and be reported as a left button click (for 3454 touchpads working in absolute mode only). 3455 Format: <msecs> 3456 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 3457 reporting absolute coordinates, such as tablets 3458 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 3459 reporting absolute coordinates, such as tablets 3460 3461 movablecore= [KNL,X86,IA-64,PPC] 3462 Format: nn[KMGTPE] | nn% 3463 This parameter is the complement to kernelcore=, it 3464 specifies the amount of memory used for migratable 3465 allocations. If both kernelcore and movablecore is 3466 specified, then kernelcore will be at *least* the 3467 specified value but may be more. If movablecore on its 3468 own is specified, the administrator must be careful 3469 that the amount of memory usable for all allocations 3470 is not too small. 3471 3472 movable_node [KNL] Boot-time switch to make hotplugable memory 3473 NUMA nodes to be movable. This means that the memory 3474 of such nodes will be usable only for movable 3475 allocations which rules out almost all kernel 3476 allocations. Use with caution! 3477 3478 MTD_Partition= [MTD] 3479 Format: <name>,<region-number>,<size>,<offset> 3480 3481 MTD_Region= [MTD] Format: 3482 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 3483 3484 mtdparts= [MTD] 3485 See drivers/mtd/parsers/cmdlinepart.c 3486 3487 mtdset= [ARM] 3488 ARM/S3C2412 JIVE boot control 3489 3490 See arch/arm/mach-s3c/mach-jive.c 3491 3492 mtouchusb.raw_coordinates= 3493 [HW] Make the MicroTouch USB driver use raw coordinates 3494 ('y', default) or cooked coordinates ('n') 3495 3496 mtrr=debug [X86] 3497 Enable printing debug information related to MTRR 3498 registers at boot time. 3499 3500 mtrr_chunk_size=nn[KMG] [X86] 3501 used for mtrr cleanup. It is largest continuous chunk 3502 that could hold holes aka. UC entries. 3503 3504 mtrr_gran_size=nn[KMG] [X86] 3505 Used for mtrr cleanup. It is granularity of mtrr block. 3506 Default is 1. 3507 Large value could prevent small alignment from 3508 using up MTRRs. 3509 3510 mtrr_spare_reg_nr=n [X86] 3511 Format: <integer> 3512 Range: 0,7 : spare reg number 3513 Default : 1 3514 Used for mtrr cleanup. It is spare mtrr entries number. 3515 Set to 2 or more if your graphical card needs more. 3516 3517 multitce=off [PPC] This parameter disables the use of the pSeries 3518 firmware feature for updating multiple TCE entries 3519 at a time. 3520 3521 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 3522 3523 netdev= [NET] Network devices parameters 3524 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 3525 Note that mem_start is often overloaded to mean 3526 something different and driver-specific. 3527 This usage is only documented in each driver source 3528 file if at all. 3529 3530 netpoll.carrier_timeout= 3531 [NET] Specifies amount of time (in seconds) that 3532 netpoll should wait for a carrier. By default netpoll 3533 waits 4 seconds. 3534 3535 nf_conntrack.acct= 3536 [NETFILTER] Enable connection tracking flow accounting 3537 0 to disable accounting 3538 1 to enable accounting 3539 Default value is 0. 3540 3541 nfs.cache_getent= 3542 [NFS] sets the pathname to the program which is used 3543 to update the NFS client cache entries. 3544 3545 nfs.cache_getent_timeout= 3546 [NFS] sets the timeout after which an attempt to 3547 update a cache entry is deemed to have failed. 3548 3549 nfs.callback_nr_threads= 3550 [NFSv4] set the total number of threads that the 3551 NFS client will assign to service NFSv4 callback 3552 requests. 3553 3554 nfs.callback_tcpport= 3555 [NFS] set the TCP port on which the NFSv4 callback 3556 channel should listen. 3557 3558 nfs.enable_ino64= 3559 [NFS] enable 64-bit inode numbers. 3560 If zero, the NFS client will fake up a 32-bit inode 3561 number for the readdir() and stat() syscalls instead 3562 of returning the full 64-bit number. 3563 The default is to return 64-bit inode numbers. 3564 3565 nfs.idmap_cache_timeout= 3566 [NFS] set the maximum lifetime for idmapper cache 3567 entries. 3568 3569 nfs.max_session_cb_slots= 3570 [NFSv4.1] Sets the maximum number of session 3571 slots the client will assign to the callback 3572 channel. This determines the maximum number of 3573 callbacks the client will process in parallel for 3574 a particular server. 3575 3576 nfs.max_session_slots= 3577 [NFSv4.1] Sets the maximum number of session slots 3578 the client will attempt to negotiate with the server. 3579 This limits the number of simultaneous RPC requests 3580 that the client can send to the NFSv4.1 server. 3581 Note that there is little point in setting this 3582 value higher than the max_tcp_slot_table_limit. 3583 3584 nfs.nfs4_disable_idmapping= 3585 [NFSv4] When set to the default of '1', this option 3586 ensures that both the RPC level authentication 3587 scheme and the NFS level operations agree to use 3588 numeric uids/gids if the mount is using the 3589 'sec=sys' security flavour. In effect it is 3590 disabling idmapping, which can make migration from 3591 legacy NFSv2/v3 systems to NFSv4 easier. 3592 Servers that do not support this mode of operation 3593 will be autodetected by the client, and it will fall 3594 back to using the idmapper. 3595 To turn off this behaviour, set the value to '0'. 3596 3597 nfs.nfs4_unique_id= 3598 [NFS4] Specify an additional fixed unique ident- 3599 ification string that NFSv4 clients can insert into 3600 their nfs_client_id4 string. This is typically a 3601 UUID that is generated at system install time. 3602 3603 nfs.recover_lost_locks= 3604 [NFSv4] Attempt to recover locks that were lost due 3605 to a lease timeout on the server. Please note that 3606 doing this risks data corruption, since there are 3607 no guarantees that the file will remain unchanged 3608 after the locks are lost. 3609 If you want to enable the kernel legacy behaviour of 3610 attempting to recover these locks, then set this 3611 parameter to '1'. 3612 The default parameter value of '0' causes the kernel 3613 not to attempt recovery of lost locks. 3614 3615 nfs.send_implementation_id= 3616 [NFSv4.1] Send client implementation identification 3617 information in exchange_id requests. 3618 If zero, no implementation identification information 3619 will be sent. 3620 The default is to send the implementation identification 3621 information. 3622 3623 nfs4.layoutstats_timer= 3624 [NFSv4.2] Change the rate at which the kernel sends 3625 layoutstats to the pNFS metadata server. 3626 3627 Setting this to value to 0 causes the kernel to use 3628 whatever value is the default set by the layout 3629 driver. A non-zero value sets the minimum interval 3630 in seconds between layoutstats transmissions. 3631 3632 nfsd.inter_copy_offload_enable= 3633 [NFSv4.2] When set to 1, the server will support 3634 server-to-server copies for which this server is 3635 the destination of the copy. 3636 3637 nfsd.nfs4_disable_idmapping= 3638 [NFSv4] When set to the default of '1', the NFSv4 3639 server will return only numeric uids and gids to 3640 clients using auth_sys, and will accept numeric uids 3641 and gids from such clients. This is intended to ease 3642 migration from NFSv2/v3. 3643 3644 nfsd.nfsd4_ssc_umount_timeout= 3645 [NFSv4.2] When used as the destination of a 3646 server-to-server copy, knfsd temporarily mounts 3647 the source server. It caches the mount in case 3648 it will be needed again, and discards it if not 3649 used for the number of milliseconds specified by 3650 this parameter. 3651 3652 nfsaddrs= [NFS] Deprecated. Use ip= instead. 3653 See Documentation/admin-guide/nfs/nfsroot.rst. 3654 3655 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 3656 See Documentation/admin-guide/nfs/nfsroot.rst. 3657 3658 nfsrootdebug [NFS] enable nfsroot debugging messages. 3659 See Documentation/admin-guide/nfs/nfsroot.rst. 3660 3661 nmi_backtrace.backtrace_idle [KNL] 3662 Dump stacks even of idle CPUs in response to an 3663 NMI stack-backtrace request. 3664 3665 nmi_debug= [KNL,SH] Specify one or more actions to take 3666 when a NMI is triggered. 3667 Format: [state][,regs][,debounce][,die] 3668 3669 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3670 Format: [panic,][nopanic,][num] 3671 Valid num: 0 or 1 3672 0 - turn hardlockup detector in nmi_watchdog off 3673 1 - turn hardlockup detector in nmi_watchdog on 3674 When panic is specified, panic when an NMI watchdog 3675 timeout occurs (or 'nopanic' to not panic on an NMI 3676 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3677 To disable both hard and soft lockup detectors, 3678 please see 'nowatchdog'. 3679 This is useful when you use a panic=... timeout and 3680 need the box quickly up again. 3681 3682 These settings can be accessed at runtime via 3683 the nmi_watchdog and hardlockup_panic sysctls. 3684 3685 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3686 emulation library even if a 387 maths coprocessor 3687 is present. 3688 3689 no4lvl [RISCV] Disable 4-level and 5-level paging modes. Forces 3690 kernel to use 3-level paging instead. 3691 3692 no5lvl [X86-64,RISCV] Disable 5-level paging mode. Forces 3693 kernel to use 4-level paging instead. 3694 3695 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3696 caches in the slab allocator. Saves per-node memory, 3697 but will impact performance. 3698 3699 noalign [KNL,ARM] 3700 3701 noaltinstr [S390] Disables alternative instructions patching 3702 (CPU alternatives feature). 3703 3704 noapic [SMP,APIC] Tells the kernel to not make use of any 3705 IOAPICs that may be present in the system. 3706 3707 noautogroup Disable scheduler automatic task group creation. 3708 3709 nocache [ARM] 3710 3711 no_console_suspend 3712 [HW] Never suspend the console 3713 Disable suspending of consoles during suspend and 3714 hibernate operations. Once disabled, debugging 3715 messages can reach various consoles while the rest 3716 of the system is being put to sleep (ie, while 3717 debugging driver suspend/resume hooks). This may 3718 not work reliably with all consoles, but is known 3719 to work with serial and VGA consoles. 3720 To facilitate more flexible debugging, we also add 3721 console_suspend, a printk module parameter to control 3722 it. Users could use console_suspend (usually 3723 /sys/module/printk/parameters/console_suspend) to 3724 turn on/off it dynamically. 3725 3726 no_debug_objects 3727 [KNL] Disable object debugging 3728 3729 nodsp [SH] Disable hardware DSP at boot time. 3730 3731 noefi Disable EFI runtime services support. 3732 3733 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel. 3734 3735 noexec [IA-64] 3736 3737 noexec32 [X86-64] 3738 This affects only 32-bit executables. 3739 noexec32=on: enable non-executable mappings (default) 3740 read doesn't imply executable mappings 3741 noexec32=off: disable non-executable mappings 3742 read implies executable mappings 3743 3744 no_file_caps Tells the kernel not to honor file capabilities. The 3745 only way then for a file to be executed with privilege 3746 is to be setuid root or executed by root. 3747 3748 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3749 3750 nofsgsbase [X86] Disables FSGSBASE instructions. 3751 3752 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3753 register save and restore. The kernel will only save 3754 legacy floating-point registers on task switch. 3755 3756 nohalt [IA-64] Tells the kernel not to use the power saving 3757 function PAL_HALT_LIGHT when idle. This increases 3758 power-consumption. On the positive side, it reduces 3759 interrupt wake-up latency, which may improve performance 3760 in certain environments such as networked servers or 3761 real-time systems. 3762 3763 no_hash_pointers 3764 Force pointers printed to the console or buffers to be 3765 unhashed. By default, when a pointer is printed via %p 3766 format string, that pointer is "hashed", i.e. obscured 3767 by hashing the pointer value. This is a security feature 3768 that hides actual kernel addresses from unprivileged 3769 users, but it also makes debugging the kernel more 3770 difficult since unequal pointers can no longer be 3771 compared. However, if this command-line option is 3772 specified, then all normal pointers will have their true 3773 value printed. This option should only be specified when 3774 debugging the kernel. Please do not use on production 3775 kernels. 3776 3777 nohibernate [HIBERNATION] Disable hibernation and resume. 3778 3779 nohlt [ARM,ARM64,MICROBLAZE,MIPS,PPC,SH] Forces the kernel to 3780 busy wait in do_idle() and not use the arch_cpu_idle() 3781 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 3782 to be effective. This is useful on platforms where the 3783 sleep(SH) or wfi(ARM,ARM64) instructions do not work 3784 correctly or when doing power measurements to evaluate 3785 the impact of the sleep instructions. This is also 3786 useful when using JTAG debugger. 3787 3788 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings. 3789 3790 nohugevmalloc [KNL,X86,PPC,ARM64] Disable kernel huge vmalloc mappings. 3791 3792 nohz= [KNL] Boottime enable/disable dynamic ticks 3793 Valid arguments: on, off 3794 Default: on 3795 3796 nohz_full= [KNL,BOOT,SMP,ISOL] 3797 The argument is a cpu list, as described above. 3798 In kernels built with CONFIG_NO_HZ_FULL=y, set 3799 the specified list of CPUs whose tick will be stopped 3800 whenever possible. The boot CPU will be forced outside 3801 the range to maintain the timekeeping. Any CPUs 3802 in this list will have their RCU callbacks offloaded, 3803 just as if they had also been called out in the 3804 rcu_nocbs= boot parameter. 3805 3806 Note that this argument takes precedence over 3807 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 3808 3809 noinitrd [RAM] Tells the kernel not to load any configured 3810 initial RAM disk. 3811 3812 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3813 remapping. 3814 [Deprecated - use intremap=off] 3815 3816 nointroute [IA-64] 3817 3818 noinvpcid [X86] Disable the INVPCID cpu feature. 3819 3820 noiotrap [SH] Disables trapped I/O port accesses. 3821 3822 noirqdebug [X86-32] Disables the code which attempts to detect and 3823 disable unhandled interrupt sources. 3824 3825 noisapnp [ISAPNP] Disables ISA PnP code. 3826 3827 nojitter [IA-64] Disables jitter checking for ITC timers. 3828 3829 nokaslr [KNL] 3830 When CONFIG_RANDOMIZE_BASE is set, this disables 3831 kernel and module base offset ASLR (Address Space 3832 Layout Randomization). 3833 3834 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3835 fault handling. 3836 3837 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3838 3839 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3840 3841 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3842 3843 nomca [IA-64] Disable machine check abort handling 3844 3845 nomce [X86-32] Disable Machine Check Exception 3846 3847 nomfgpt [X86-32] Disable Multi-Function General Purpose 3848 Timer usage (for AMD Geode machines). 3849 3850 nomodeset Disable kernel modesetting. Most systems' firmware 3851 sets up a display mode and provides framebuffer memory 3852 for output. With nomodeset, DRM and fbdev drivers will 3853 not load if they could possibly displace the pre- 3854 initialized output. Only the system framebuffer will 3855 be available for use. The respective drivers will not 3856 perform display-mode changes or accelerated rendering. 3857 3858 Useful as error fallback, or for testing and debugging. 3859 3860 nomodule Disable module load 3861 3862 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3863 shutdown the other cpus. Instead use the REBOOT_VECTOR 3864 irq. 3865 3866 nopat [X86] Disable PAT (page attribute table extension of 3867 pagetables) support. 3868 3869 nopcid [X86-64] Disable the PCID cpu feature. 3870 3871 nopku [X86] Disable Memory Protection Keys CPU feature found 3872 in some Intel CPUs. 3873 3874 nopti [X86-64] 3875 Equivalent to pti=off 3876 3877 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 3878 Disables the PV optimizations forcing the guest to run 3879 as generic guest with no PV drivers. Currently support 3880 XEN HVM, KVM, HYPER_V and VMWARE guest. 3881 3882 nopvspin [X86,XEN,KVM] 3883 Disables the qspinlock slow path using PV optimizations 3884 which allow the hypervisor to 'idle' the guest on lock 3885 contention. 3886 3887 norandmaps Don't use address space randomization. Equivalent to 3888 echo 0 > /proc/sys/kernel/randomize_va_space 3889 3890 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3891 with UP alternatives 3892 3893 noresume [SWSUSP] Disables resume and restores original swap 3894 space. 3895 3896 nosbagart [IA-64] 3897 3898 no-scroll [VGA] Disables scrollback. 3899 This is required for the Braillex ib80-piezo Braille 3900 reader made by F.H. Papenmeier (Germany). 3901 3902 nosgx [X86-64,SGX] Disables Intel SGX kernel support. 3903 3904 nosmap [PPC] 3905 Disable SMAP (Supervisor Mode Access Prevention) 3906 even if it is supported by processor. 3907 3908 nosmep [PPC64s] 3909 Disable SMEP (Supervisor Mode Execution Prevention) 3910 even if it is supported by processor. 3911 3912 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3913 and disable the IO APIC. legacy for "maxcpus=0". 3914 3915 nosmt [KNL,MIPS,PPC,S390] Disable symmetric multithreading (SMT). 3916 Equivalent to smt=1. 3917 3918 [KNL,X86,PPC] Disable symmetric multithreading (SMT). 3919 nosmt=force: Force disable SMT, cannot be undone 3920 via the sysfs control file. 3921 3922 nosoftlockup [KNL] Disable the soft-lockup detector. 3923 3924 nospec_store_bypass_disable 3925 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3926 3927 nospectre_bhb [ARM64] Disable all mitigations for Spectre-BHB (branch 3928 history injection) vulnerability. System may allow data leaks 3929 with this option. 3930 3931 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3932 (bounds check bypass). With this option data leaks are 3933 possible in the system. 3934 3935 nospectre_v2 [X86,PPC_E500,ARM64] Disable all mitigations for 3936 the Spectre variant 2 (indirect branch prediction) 3937 vulnerability. System may allow data leaks with this 3938 option. 3939 3940 no-steal-acc [X86,PV_OPS,ARM64,PPC/PSERIES] Disable paravirtualized 3941 steal time accounting. steal time is computed, but 3942 won't influence scheduler behaviour 3943 3944 nosync [HW,M68K] Disables sync negotiation for all devices. 3945 3946 no_timer_check [X86,APIC] Disables the code which tests for 3947 broken timer IRQ sources. 3948 3949 no_uaccess_flush 3950 [PPC] Don't flush the L1-D cache after accessing user data. 3951 3952 novmcoredd [KNL,KDUMP] 3953 Disable device dump. Device dump allows drivers to 3954 append dump data to vmcore so you can collect driver 3955 specified debug info. Drivers can append the data 3956 without any limit and this data is stored in memory, 3957 so this may cause significant memory stress. Disabling 3958 device dump can help save memory but the driver debug 3959 data will be no longer available. This parameter 3960 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3961 is set. 3962 3963 no-vmw-sched-clock 3964 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3965 clock and use the default one. 3966 3967 nowatchdog [KNL] Disable both lockup detectors, i.e. 3968 soft-lockup and NMI watchdog (hard-lockup). 3969 3970 nowb [ARM] 3971 3972 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3973 3974 NOTE: this parameter will be ignored on systems with the 3975 LEGACY_XAPIC_DISABLED bit set in the 3976 IA32_XAPIC_DISABLE_STATUS MSR. 3977 3978 noxsave [BUGS=X86] Disables x86 extended register state save 3979 and restore using xsave. The kernel will fallback to 3980 enabling legacy floating-point and sse state. 3981 3982 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3983 register states. The kernel will fall back to use 3984 xsave to save the states. By using this parameter, 3985 performance of saving the states is degraded because 3986 xsave doesn't support modified optimization while 3987 xsaveopt supports it on xsaveopt enabled systems. 3988 3989 noxsaves [X86] Disables xsaves and xrstors used in saving and 3990 restoring x86 extended register state in compacted 3991 form of xsave area. The kernel will fall back to use 3992 xsaveopt and xrstor to save and restore the states 3993 in standard form of xsave area. By using this 3994 parameter, xsave area per process might occupy more 3995 memory on xsaves enabled systems. 3996 3997 nps_mtm_hs_ctr= [KNL,ARC] 3998 This parameter sets the maximum duration, in 3999 cycles, each HW thread of the CTOP can run 4000 without interruptions, before HW switches it. 4001 The actual maximum duration is 16 times this 4002 parameter's value. 4003 Format: integer between 1 and 255 4004 Default: 255 4005 4006 nptcg= [IA-64] Override max number of concurrent global TLB 4007 purges which is reported from either PAL_VM_SUMMARY or 4008 SAL PALO. 4009 4010 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 4011 could support. nr_cpus=n : n >= 1 limits the kernel to 4012 support 'n' processors. It could be larger than the 4013 number of already plugged CPU during bootup, later in 4014 runtime you can physically add extra cpu until it reaches 4015 n. So during boot up some boot time memory for per-cpu 4016 variables need be pre-allocated for later physical cpu 4017 hot plugging. 4018 4019 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 4020 4021 numa=off [KNL, ARM64, PPC, RISCV, SPARC, X86] Disable NUMA, Only 4022 set up a single NUMA node spanning all memory. 4023 4024 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 4025 NUMA balancing. 4026 Allowed values are enable and disable 4027 4028 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 4029 'node', 'default' can be specified 4030 This can be set from sysctl after boot. 4031 See Documentation/admin-guide/sysctl/vm.rst for details. 4032 4033 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 4034 See Documentation/core-api/debugging-via-ohci1394.rst for more 4035 info. 4036 4037 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 4038 Rather than timing out after 20 ms if an EC 4039 command is not properly ACKed, override the length 4040 of the timeout. We have interrupts disabled while 4041 waiting for the ACK, so if this is set too high 4042 interrupts *may* be lost! 4043 4044 omap_mux= [OMAP] Override bootloader pin multiplexing. 4045 Format: <mux_mode0.mode_name=value>... 4046 For example, to override I2C bus2: 4047 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 4048 4049 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 4050 4051 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 4052 4053 boundary - index of last SLC block on Flex-OneNAND. 4054 The remaining blocks are configured as MLC blocks. 4055 lock - Configure if Flex-OneNAND boundary should be locked. 4056 Once locked, the boundary cannot be changed. 4057 1 indicates lock status, 0 indicates unlock status. 4058 4059 oops=panic Always panic on oopses. Default is to just kill the 4060 process, but there is a small probability of 4061 deadlocking the machine. 4062 This will also cause panics on machine check exceptions. 4063 Useful together with panic=30 to trigger a reboot. 4064 4065 page_alloc.shuffle= 4066 [KNL] Boolean flag to control whether the page allocator 4067 should randomize its free lists. The randomization may 4068 be automatically enabled if the kernel detects it is 4069 running on a platform with a direct-mapped memory-side 4070 cache, and this parameter can be used to 4071 override/disable that behavior. The state of the flag 4072 can be read from sysfs at: 4073 /sys/module/page_alloc/parameters/shuffle. 4074 4075 page_owner= [KNL] Boot-time page_owner enabling option. 4076 Storage of the information about who allocated 4077 each page is disabled in default. With this switch, 4078 we can turn it on. 4079 on: enable the feature 4080 4081 page_poison= [KNL] Boot-time parameter changing the state of 4082 poisoning on the buddy allocator, available with 4083 CONFIG_PAGE_POISONING=y. 4084 off: turn off poisoning (default) 4085 on: turn on poisoning 4086 4087 page_reporting.page_reporting_order= 4088 [KNL] Minimal page reporting order 4089 Format: <integer> 4090 Adjust the minimal page reporting order. The page 4091 reporting is disabled when it exceeds MAX_ORDER. 4092 4093 panic= [KNL] Kernel behaviour on panic: delay <timeout> 4094 timeout > 0: seconds before rebooting 4095 timeout = 0: wait forever 4096 timeout < 0: reboot immediately 4097 Format: <timeout> 4098 4099 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 4100 Format: <hex>[,nousertaint] 4101 Hexadecimal bitmask representing the set of TAINT flags 4102 that will cause the kernel to panic when add_taint() is 4103 called with any of the flags in this set. 4104 The optional switch "nousertaint" can be utilized to 4105 prevent userspace forced crashes by writing to sysctl 4106 /proc/sys/kernel/tainted any flagset matching with the 4107 bitmask set on panic_on_taint. 4108 See Documentation/admin-guide/tainted-kernels.rst for 4109 extra details on the taint flags that users can pick 4110 to compose the bitmask to assign to panic_on_taint. 4111 4112 panic_on_warn=1 panic() instead of WARN(). Useful to cause kdump 4113 on a WARN(). 4114 4115 panic_print= Bitmask for printing system info when panic happens. 4116 User can chose combination of the following bits: 4117 bit 0: print all tasks info 4118 bit 1: print system memory info 4119 bit 2: print timer info 4120 bit 3: print locks info if CONFIG_LOCKDEP is on 4121 bit 4: print ftrace buffer 4122 bit 5: print all printk messages in buffer 4123 bit 6: print all CPUs backtrace (if available in the arch) 4124 *Be aware* that this option may print a _lot_ of lines, 4125 so there are risks of losing older messages in the log. 4126 Use this option carefully, maybe worth to setup a 4127 bigger log buffer with "log_buf_len" along with this. 4128 4129 parkbd.port= [HW] Parallel port number the keyboard adapter is 4130 connected to, default is 0. 4131 Format: <parport#> 4132 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 4133 0 for XT, 1 for AT (default is AT). 4134 Format: <mode> 4135 4136 parport= [HW,PPT] Specify parallel ports. 0 disables. 4137 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 4138 Use 'auto' to force the driver to use any 4139 IRQ/DMA settings detected (the default is to 4140 ignore detected IRQ/DMA settings because of 4141 possible conflicts). You can specify the base 4142 address, IRQ, and DMA settings; IRQ and DMA 4143 should be numbers, or 'auto' (for using detected 4144 settings on that particular port), or 'nofifo' 4145 (to avoid using a FIFO even if it is detected). 4146 Parallel ports are assigned in the order they 4147 are specified on the command line, starting 4148 with parport0. 4149 4150 parport_init_mode= [HW,PPT] 4151 Configure VIA parallel port to operate in 4152 a specific mode. This is necessary on Pegasos 4153 computer where firmware has no options for setting 4154 up parallel port mode and sets it to spp. 4155 Currently this function knows 686a and 8231 chips. 4156 Format: [spp|ps2|epp|ecp|ecpepp] 4157 4158 pata_legacy.all= [HW,LIBATA] 4159 Format: <int> 4160 Set to non-zero to probe primary and secondary ISA 4161 port ranges on PCI systems where no PCI PATA device 4162 has been found at either range. Disabled by default. 4163 4164 pata_legacy.autospeed= [HW,LIBATA] 4165 Format: <int> 4166 Set to non-zero if a chip is present that snoops speed 4167 changes. Disabled by default. 4168 4169 pata_legacy.ht6560a= [HW,LIBATA] 4170 Format: <int> 4171 Set to 1, 2, or 3 for HT 6560A on the primary channel, 4172 the secondary channel, or both channels respectively. 4173 Disabled by default. 4174 4175 pata_legacy.ht6560b= [HW,LIBATA] 4176 Format: <int> 4177 Set to 1, 2, or 3 for HT 6560B on the primary channel, 4178 the secondary channel, or both channels respectively. 4179 Disabled by default. 4180 4181 pata_legacy.iordy_mask= [HW,LIBATA] 4182 Format: <int> 4183 IORDY enable mask. Set individual bits to allow IORDY 4184 for the respective channel. Bit 0 is for the first 4185 legacy channel handled by this driver, bit 1 is for 4186 the second channel, and so on. The sequence will often 4187 correspond to the primary legacy channel, the secondary 4188 legacy channel, and so on, but the handling of a PCI 4189 bus and the use of other driver options may interfere 4190 with the sequence. By default IORDY is allowed across 4191 all channels. 4192 4193 pata_legacy.opti82c46x= [HW,LIBATA] 4194 Format: <int> 4195 Set to 1, 2, or 3 for Opti 82c611A on the primary 4196 channel, the secondary channel, or both channels 4197 respectively. Disabled by default. 4198 4199 pata_legacy.opti82c611a= [HW,LIBATA] 4200 Format: <int> 4201 Set to 1, 2, or 3 for Opti 82c465MV on the primary 4202 channel, the secondary channel, or both channels 4203 respectively. Disabled by default. 4204 4205 pata_legacy.pio_mask= [HW,LIBATA] 4206 Format: <int> 4207 PIO mode mask for autospeed devices. Set individual 4208 bits to allow the use of the respective PIO modes. 4209 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 4210 All modes allowed by default. 4211 4212 pata_legacy.probe_all= [HW,LIBATA] 4213 Format: <int> 4214 Set to non-zero to probe tertiary and further ISA 4215 port ranges on PCI systems. Disabled by default. 4216 4217 pata_legacy.probe_mask= [HW,LIBATA] 4218 Format: <int> 4219 Probe mask for legacy ISA PATA ports. Depending on 4220 platform configuration and the use of other driver 4221 options up to 6 legacy ports are supported: 0x1f0, 4222 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 4223 of individual ports can be disabled by setting the 4224 corresponding bits in the mask to 1. Bit 0 is for 4225 the first port in the list above (0x1f0), and so on. 4226 By default all supported ports are probed. 4227 4228 pata_legacy.qdi= [HW,LIBATA] 4229 Format: <int> 4230 Set to non-zero to probe QDI controllers. By default 4231 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 4232 4233 pata_legacy.winbond= [HW,LIBATA] 4234 Format: <int> 4235 Set to non-zero to probe Winbond controllers. Use 4236 the standard I/O port (0x130) if 1, otherwise the 4237 value given is the I/O port to use (typically 0x1b0). 4238 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 4239 0 otherwise. 4240 4241 pata_platform.pio_mask= [HW,LIBATA] 4242 Format: <int> 4243 Supported PIO mode mask. Set individual bits to allow 4244 the use of the respective PIO modes. Bit 0 is for 4245 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 4246 allowed by default. 4247 4248 pause_on_oops=<int> 4249 Halt all CPUs after the first oops has been printed for 4250 the specified number of seconds. This is to be used if 4251 your oopses keep scrolling off the screen. 4252 4253 pcbit= [HW,ISDN] 4254 4255 pci=option[,option...] [PCI] various PCI subsystem options. 4256 4257 Some options herein operate on a specific device 4258 or a set of devices (<pci_dev>). These are 4259 specified in one of the following formats: 4260 4261 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 4262 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 4263 4264 Note: the first format specifies a PCI 4265 bus/device/function address which may change 4266 if new hardware is inserted, if motherboard 4267 firmware changes, or due to changes caused 4268 by other kernel parameters. If the 4269 domain is left unspecified, it is 4270 taken to be zero. Optionally, a path 4271 to a device through multiple device/function 4272 addresses can be specified after the base 4273 address (this is more robust against 4274 renumbering issues). The second format 4275 selects devices using IDs from the 4276 configuration space which may match multiple 4277 devices in the system. 4278 4279 earlydump dump PCI config space before the kernel 4280 changes anything 4281 off [X86] don't probe for the PCI bus 4282 bios [X86-32] force use of PCI BIOS, don't access 4283 the hardware directly. Use this if your machine 4284 has a non-standard PCI host bridge. 4285 nobios [X86-32] disallow use of PCI BIOS, only direct 4286 hardware access methods are allowed. Use this 4287 if you experience crashes upon bootup and you 4288 suspect they are caused by the BIOS. 4289 conf1 [X86] Force use of PCI Configuration Access 4290 Mechanism 1 (config address in IO port 0xCF8, 4291 data in IO port 0xCFC, both 32-bit). 4292 conf2 [X86] Force use of PCI Configuration Access 4293 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 4294 the function, IO port 0xCFA, also 8-bit, sets 4295 bus number. The config space is then accessed 4296 through ports 0xC000-0xCFFF). 4297 See http://wiki.osdev.org/PCI for more info 4298 on the configuration access mechanisms. 4299 noaer [PCIE] If the PCIEAER kernel config parameter is 4300 enabled, this kernel boot option can be used to 4301 disable the use of PCIE advanced error reporting. 4302 nodomains [PCI] Disable support for multiple PCI 4303 root domains (aka PCI segments, in ACPI-speak). 4304 nommconf [X86] Disable use of MMCONFIG for PCI 4305 Configuration 4306 check_enable_amd_mmconf [X86] check for and enable 4307 properly configured MMIO access to PCI 4308 config space on AMD family 10h CPU 4309 nomsi [MSI] If the PCI_MSI kernel config parameter is 4310 enabled, this kernel boot option can be used to 4311 disable the use of MSI interrupts system-wide. 4312 noioapicquirk [APIC] Disable all boot interrupt quirks. 4313 Safety option to keep boot IRQs enabled. This 4314 should never be necessary. 4315 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 4316 primary IO-APIC for bridges that cannot disable 4317 boot IRQs. This fixes a source of spurious IRQs 4318 when the system masks IRQs. 4319 noioapicreroute [APIC] Disable workaround that uses the 4320 boot IRQ equivalent of an IRQ that connects to 4321 a chipset where boot IRQs cannot be disabled. 4322 The opposite of ioapicreroute. 4323 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 4324 routing table. These calls are known to be buggy 4325 on several machines and they hang the machine 4326 when used, but on other computers it's the only 4327 way to get the interrupt routing table. Try 4328 this option if the kernel is unable to allocate 4329 IRQs or discover secondary PCI buses on your 4330 motherboard. 4331 rom [X86] Assign address space to expansion ROMs. 4332 Use with caution as certain devices share 4333 address decoders between ROMs and other 4334 resources. 4335 norom [X86] Do not assign address space to 4336 expansion ROMs that do not already have 4337 BIOS assigned address ranges. 4338 nobar [X86] Do not assign address space to the 4339 BARs that weren't assigned by the BIOS. 4340 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 4341 assigned automatically to PCI devices. You can 4342 make the kernel exclude IRQs of your ISA cards 4343 this way. 4344 pirqaddr=0xAAAAA [X86] Specify the physical address 4345 of the PIRQ table (normally generated 4346 by the BIOS) if it is outside the 4347 F0000h-100000h range. 4348 lastbus=N [X86] Scan all buses thru bus #N. Can be 4349 useful if the kernel is unable to find your 4350 secondary buses and you want to tell it 4351 explicitly which ones they are. 4352 assign-busses [X86] Always assign all PCI bus 4353 numbers ourselves, overriding 4354 whatever the firmware may have done. 4355 usepirqmask [X86] Honor the possible IRQ mask stored 4356 in the BIOS $PIR table. This is needed on 4357 some systems with broken BIOSes, notably 4358 some HP Pavilion N5400 and Omnibook XE3 4359 notebooks. This will have no effect if ACPI 4360 IRQ routing is enabled. 4361 noacpi [X86] Do not use ACPI for IRQ routing 4362 or for PCI scanning. 4363 use_crs [X86] Use PCI host bridge window information 4364 from ACPI. On BIOSes from 2008 or later, this 4365 is enabled by default. If you need to use this, 4366 please report a bug. 4367 nocrs [X86] Ignore PCI host bridge windows from ACPI. 4368 If you need to use this, please report a bug. 4369 use_e820 [X86] Use E820 reservations to exclude parts of 4370 PCI host bridge windows. This is a workaround 4371 for BIOS defects in host bridge _CRS methods. 4372 If you need to use this, please report a bug to 4373 <linux-pci@vger.kernel.org>. 4374 no_e820 [X86] Ignore E820 reservations for PCI host 4375 bridge windows. This is the default on modern 4376 hardware. If you need to use this, please report 4377 a bug to <linux-pci@vger.kernel.org>. 4378 routeirq Do IRQ routing for all PCI devices. 4379 This is normally done in pci_enable_device(), 4380 so this option is a temporary workaround 4381 for broken drivers that don't call it. 4382 skip_isa_align [X86] do not align io start addr, so can 4383 handle more pci cards 4384 noearly [X86] Don't do any early type 1 scanning. 4385 This might help on some broken boards which 4386 machine check when some devices' config space 4387 is read. But various workarounds are disabled 4388 and some IOMMU drivers will not work. 4389 bfsort Sort PCI devices into breadth-first order. 4390 This sorting is done to get a device 4391 order compatible with older (<= 2.4) kernels. 4392 nobfsort Don't sort PCI devices into breadth-first order. 4393 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 4394 tuning and use the BIOS-configured MPS defaults. 4395 pcie_bus_safe Set every device's MPS to the largest value 4396 supported by all devices below the root complex. 4397 pcie_bus_perf Set device MPS to the largest allowable MPS 4398 based on its parent bus. Also set MRRS (Max 4399 Read Request Size) to the largest supported 4400 value (no larger than the MPS that the device 4401 or bus can support) for best performance. 4402 pcie_bus_peer2peer Set every device's MPS to 128B, which 4403 every device is guaranteed to support. This 4404 configuration allows peer-to-peer DMA between 4405 any pair of devices, possibly at the cost of 4406 reduced performance. This also guarantees 4407 that hot-added devices will work. 4408 cbiosize=nn[KMG] The fixed amount of bus space which is 4409 reserved for the CardBus bridge's IO window. 4410 The default value is 256 bytes. 4411 cbmemsize=nn[KMG] The fixed amount of bus space which is 4412 reserved for the CardBus bridge's memory 4413 window. The default value is 64 megabytes. 4414 resource_alignment= 4415 Format: 4416 [<order of align>@]<pci_dev>[; ...] 4417 Specifies alignment and device to reassign 4418 aligned memory resources. How to 4419 specify the device is described above. 4420 If <order of align> is not specified, 4421 PAGE_SIZE is used as alignment. 4422 A PCI-PCI bridge can be specified if resource 4423 windows need to be expanded. 4424 To specify the alignment for several 4425 instances of a device, the PCI vendor, 4426 device, subvendor, and subdevice may be 4427 specified, e.g., 12@pci:8086:9c22:103c:198f 4428 for 4096-byte alignment. 4429 ecrc= Enable/disable PCIe ECRC (transaction layer 4430 end-to-end CRC checking). Only effective if 4431 OS has native AER control (either granted by 4432 ACPI _OSC or forced via "pcie_ports=native") 4433 bios: Use BIOS/firmware settings. This is the 4434 the default. 4435 off: Turn ECRC off 4436 on: Turn ECRC on. 4437 hpiosize=nn[KMG] The fixed amount of bus space which is 4438 reserved for hotplug bridge's IO window. 4439 Default size is 256 bytes. 4440 hpmmiosize=nn[KMG] The fixed amount of bus space which is 4441 reserved for hotplug bridge's MMIO window. 4442 Default size is 2 megabytes. 4443 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 4444 reserved for hotplug bridge's MMIO_PREF window. 4445 Default size is 2 megabytes. 4446 hpmemsize=nn[KMG] The fixed amount of bus space which is 4447 reserved for hotplug bridge's MMIO and 4448 MMIO_PREF window. 4449 Default size is 2 megabytes. 4450 hpbussize=nn The minimum amount of additional bus numbers 4451 reserved for buses below a hotplug bridge. 4452 Default is 1. 4453 realloc= Enable/disable reallocating PCI bridge resources 4454 if allocations done by BIOS are too small to 4455 accommodate resources required by all child 4456 devices. 4457 off: Turn realloc off 4458 on: Turn realloc on 4459 realloc same as realloc=on 4460 noari do not use PCIe ARI. 4461 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 4462 do not use PCIe ATS (and IOMMU device IOTLB). 4463 pcie_scan_all Scan all possible PCIe devices. Otherwise we 4464 only look for one device below a PCIe downstream 4465 port. 4466 big_root_window Try to add a big 64bit memory window to the PCIe 4467 root complex on AMD CPUs. Some GFX hardware 4468 can resize a BAR to allow access to all VRAM. 4469 Adding the window is slightly risky (it may 4470 conflict with unreported devices), so this 4471 taints the kernel. 4472 disable_acs_redir=<pci_dev>[; ...] 4473 Specify one or more PCI devices (in the format 4474 specified above) separated by semicolons. 4475 Each device specified will have the PCI ACS 4476 redirect capabilities forced off which will 4477 allow P2P traffic between devices through 4478 bridges without forcing it upstream. Note: 4479 this removes isolation between devices and 4480 may put more devices in an IOMMU group. 4481 force_floating [S390] Force usage of floating interrupts. 4482 nomio [S390] Do not use MIO instructions. 4483 norid [S390] ignore the RID field and force use of 4484 one PCI domain per PCI function 4485 4486 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 4487 Management. 4488 off Disable ASPM. 4489 force Enable ASPM even on devices that claim not to support it. 4490 WARNING: Forcing ASPM on may cause system lockups. 4491 4492 pcie_ports= [PCIE] PCIe port services handling: 4493 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 4494 even if the platform doesn't give the OS permission to 4495 use them. This may cause conflicts if the platform 4496 also tries to use these services. 4497 dpc-native Use native PCIe service for DPC only. May 4498 cause conflicts if firmware uses AER or DPC. 4499 compat Disable native PCIe services (PME, AER, DPC, PCIe 4500 hotplug). 4501 4502 pcie_port_pm= [PCIE] PCIe port power management handling: 4503 off Disable power management of all PCIe ports 4504 force Forcibly enable power management of all PCIe ports 4505 4506 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 4507 nomsi Do not use MSI for native PCIe PME signaling (this makes 4508 all PCIe root ports use INTx for all services). 4509 4510 pcmv= [HW,PCMCIA] BadgePAD 4 4511 4512 pd_ignore_unused 4513 [PM] 4514 Keep all power-domains already enabled by bootloader on, 4515 even if no driver has claimed them. This is useful 4516 for debug and development, but should not be 4517 needed on a platform with proper driver support. 4518 4519 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 4520 boot time. 4521 Format: { 0 | 1 } 4522 See arch/parisc/kernel/pdc_chassis.c 4523 4524 percpu_alloc= Select which percpu first chunk allocator to use. 4525 Currently supported values are "embed" and "page". 4526 Archs may support subset or none of the selections. 4527 See comments in mm/percpu.c for details on each 4528 allocator. This parameter is primarily for debugging 4529 and performance comparison. 4530 4531 pirq= [SMP,APIC] Manual mp-table setup 4532 See Documentation/arch/x86/i386/IO-APIC.rst. 4533 4534 plip= [PPT,NET] Parallel port network link 4535 Format: { parport<nr> | timid | 0 } 4536 See also Documentation/admin-guide/parport.rst. 4537 4538 pmtmr= [X86] Manual setup of pmtmr I/O Port. 4539 Override pmtimer IOPort with a hex value. 4540 e.g. pmtmr=0x508 4541 4542 pmu_override= [PPC] Override the PMU. 4543 This option takes over the PMU facility, so it is no 4544 longer usable by perf. Setting this option starts the 4545 PMU counters by setting MMCR0 to 0 (the FC bit is 4546 cleared). If a number is given, then MMCR1 is set to 4547 that number, otherwise (e.g., 'pmu_override=on'), MMCR1 4548 remains 0. 4549 4550 pm_debug_messages [SUSPEND,KNL] 4551 Enable suspend/resume debug messages during boot up. 4552 4553 pnp.debug=1 [PNP] 4554 Enable PNP debug messages (depends on the 4555 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 4556 via /sys/module/pnp/parameters/debug. We always show 4557 current resource usage; turning this on also shows 4558 possible settings and some assignment information. 4559 4560 pnpacpi= [ACPI] 4561 { off } 4562 4563 pnpbios= [ISAPNP] 4564 { on | off | curr | res | no-curr | no-res } 4565 4566 pnp_reserve_irq= 4567 [ISAPNP] Exclude IRQs for the autoconfiguration 4568 4569 pnp_reserve_dma= 4570 [ISAPNP] Exclude DMAs for the autoconfiguration 4571 4572 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 4573 Ranges are in pairs (I/O port base and size). 4574 4575 pnp_reserve_mem= 4576 [ISAPNP] Exclude memory regions for the 4577 autoconfiguration. 4578 Ranges are in pairs (memory base and size). 4579 4580 ports= [IP_VS_FTP] IPVS ftp helper module 4581 Default is 21. 4582 Up to 8 (IP_VS_APP_MAX_PORTS) ports 4583 may be specified. 4584 Format: <port>,<port>.... 4585 4586 powersave=off [PPC] This option disables power saving features. 4587 It specifically disables cpuidle and sets the 4588 platform machine description specific power_save 4589 function to NULL. On Idle the CPU just reduces 4590 execution priority. 4591 4592 ppc_strict_facility_enable 4593 [PPC] This option catches any kernel floating point, 4594 Altivec, VSX and SPE outside of regions specifically 4595 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 4596 There is some performance impact when enabling this. 4597 4598 ppc_tm= [PPC] 4599 Format: {"off"} 4600 Disable Hardware Transactional Memory 4601 4602 preempt= [KNL] 4603 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 4604 none - Limited to cond_resched() calls 4605 voluntary - Limited to cond_resched() and might_sleep() calls 4606 full - Any section that isn't explicitly preempt disabled 4607 can be preempted anytime. 4608 4609 print-fatal-signals= 4610 [KNL] debug: print fatal signals 4611 4612 If enabled, warn about various signal handling 4613 related application anomalies: too many signals, 4614 too many POSIX.1 timers, fatal signals causing a 4615 coredump - etc. 4616 4617 If you hit the warning due to signal overflow, 4618 you might want to try "ulimit -i unlimited". 4619 4620 default: off. 4621 4622 printk.always_kmsg_dump= 4623 Trigger kmsg_dump for cases other than kernel oops or 4624 panics 4625 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4626 default: disabled 4627 4628 printk.console_no_auto_verbose= 4629 Disable console loglevel raise on oops, panic 4630 or lockdep-detected issues (only if lock debug is on). 4631 With an exception to setups with low baudrate on 4632 serial console, keeping this 0 is a good choice 4633 in order to provide more debug information. 4634 Format: <bool> 4635 default: 0 (auto_verbose is enabled) 4636 4637 printk.devkmsg={on,off,ratelimit} 4638 Control writing to /dev/kmsg. 4639 on - unlimited logging to /dev/kmsg from userspace 4640 off - logging to /dev/kmsg disabled 4641 ratelimit - ratelimit the logging 4642 Default: ratelimit 4643 4644 printk.time= Show timing data prefixed to each printk message line 4645 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4646 4647 proc_mem.force_override= [KNL] 4648 Format: {always | ptrace | never} 4649 Traditionally /proc/pid/mem allows memory permissions to be 4650 overridden without restrictions. This option may be set to 4651 restrict that. Can be one of: 4652 - 'always': traditional behavior always allows mem overrides. 4653 - 'ptrace': only allow mem overrides for active ptracers. 4654 - 'never': never allow mem overrides. 4655 If not specified, default is the CONFIG_PROC_MEM_* choice. 4656 4657 processor.max_cstate= [HW,ACPI] 4658 Limit processor to maximum C-state 4659 max_cstate=9 overrides any DMI blacklist limit. 4660 4661 processor.nocst [HW,ACPI] 4662 Ignore the _CST method to determine C-states, 4663 instead using the legacy FADT method 4664 4665 profile= [KNL] Enable kernel profiling via /proc/profile 4666 Format: [<profiletype>,]<number> 4667 Param: <profiletype>: "schedule" or "kvm" 4668 [defaults to kernel profiling] 4669 Param: "schedule" - profile schedule points. 4670 Param: "kvm" - profile VM exits. 4671 Param: <number> - step/bucket size as a power of 2 for 4672 statistical time based profiling. 4673 4674 prompt_ramdisk= [RAM] [Deprecated] 4675 4676 prot_virt= [S390] enable hosting protected virtual machines 4677 isolated from the hypervisor (if hardware supports 4678 that). 4679 Format: <bool> 4680 4681 psi= [KNL] Enable or disable pressure stall information 4682 tracking. 4683 Format: <bool> 4684 4685 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 4686 probe for; one of (bare|imps|exps|lifebook|any). 4687 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 4688 per second. 4689 psmouse.resetafter= [HW,MOUSE] 4690 Try to reset the device after so many bad packets 4691 (0 = never). 4692 psmouse.resolution= 4693 [HW,MOUSE] Set desired mouse resolution, in dpi. 4694 psmouse.smartscroll= 4695 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 4696 0 = disabled, 1 = enabled (default). 4697 4698 pstore.backend= Specify the name of the pstore backend to use 4699 4700 pti= [X86-64] Control Page Table Isolation of user and 4701 kernel address spaces. Disabling this feature 4702 removes hardening, but improves performance of 4703 system calls and interrupts. 4704 4705 on - unconditionally enable 4706 off - unconditionally disable 4707 auto - kernel detects whether your CPU model is 4708 vulnerable to issues that PTI mitigates 4709 4710 Not specifying this option is equivalent to pti=auto. 4711 4712 pty.legacy_count= 4713 [KNL] Number of legacy pty's. Overwrites compiled-in 4714 default number. 4715 4716 quiet [KNL] Disable most log messages 4717 4718 r128= [HW,DRM] 4719 4720 radix_hcall_invalidate=on [PPC/PSERIES] 4721 Disable RADIX GTSE feature and use hcall for TLB 4722 invalidate. 4723 4724 raid= [HW,RAID] 4725 See Documentation/admin-guide/md.rst. 4726 4727 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 4728 See Documentation/admin-guide/blockdev/ramdisk.rst. 4729 4730 ramdisk_start= [RAM] RAM disk image start address 4731 4732 random.trust_cpu=off 4733 [KNL] Disable trusting the use of the CPU's 4734 random number generator (if available) to 4735 initialize the kernel's RNG. 4736 4737 random.trust_bootloader=off 4738 [KNL] Disable trusting the use of the a seed 4739 passed by the bootloader (if available) to 4740 initialize the kernel's RNG. 4741 4742 randomize_kstack_offset= 4743 [KNL] Enable or disable kernel stack offset 4744 randomization, which provides roughly 5 bits of 4745 entropy, frustrating memory corruption attacks 4746 that depend on stack address determinism or 4747 cross-syscall address exposures. This is only 4748 available on architectures that have defined 4749 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 4750 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4751 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 4752 4753 ras=option[,option,...] [KNL] RAS-specific options 4754 4755 cec_disable [X86] 4756 Disable the Correctable Errors Collector, 4757 see CONFIG_RAS_CEC help text. 4758 4759 rcu_nocbs[=cpu-list] 4760 [KNL] The optional argument is a cpu list, 4761 as described above. 4762 4763 In kernels built with CONFIG_RCU_NOCB_CPU=y, 4764 enable the no-callback CPU mode, which prevents 4765 such CPUs' callbacks from being invoked in 4766 softirq context. Invocation of such CPUs' RCU 4767 callbacks will instead be offloaded to "rcuox/N" 4768 kthreads created for that purpose, where "x" is 4769 "p" for RCU-preempt, "s" for RCU-sched, and "g" 4770 for the kthreads that mediate grace periods; and 4771 "N" is the CPU number. This reduces OS jitter on 4772 the offloaded CPUs, which can be useful for HPC 4773 and real-time workloads. It can also improve 4774 energy efficiency for asymmetric multiprocessors. 4775 4776 If a cpulist is passed as an argument, the specified 4777 list of CPUs is set to no-callback mode from boot. 4778 4779 Otherwise, if the '=' sign and the cpulist 4780 arguments are omitted, no CPU will be set to 4781 no-callback mode from boot but the mode may be 4782 toggled at runtime via cpusets. 4783 4784 Note that this argument takes precedence over 4785 the CONFIG_RCU_NOCB_CPU_DEFAULT_ALL option. 4786 4787 rcu_nocb_poll [KNL] 4788 Rather than requiring that offloaded CPUs 4789 (specified by rcu_nocbs= above) explicitly 4790 awaken the corresponding "rcuoN" kthreads, 4791 make these kthreads poll for callbacks. 4792 This improves the real-time response for the 4793 offloaded CPUs by relieving them of the need to 4794 wake up the corresponding kthread, but degrades 4795 energy efficiency by requiring that the kthreads 4796 periodically wake up to do the polling. 4797 4798 rcutree.blimit= [KNL] 4799 Set maximum number of finished RCU callbacks to 4800 process in one batch. 4801 4802 rcutree.dump_tree= [KNL] 4803 Dump the structure of the rcu_node combining tree 4804 out at early boot. This is used for diagnostic 4805 purposes, to verify correct tree setup. 4806 4807 rcutree.gp_cleanup_delay= [KNL] 4808 Set the number of jiffies to delay each step of 4809 RCU grace-period cleanup. 4810 4811 rcutree.gp_init_delay= [KNL] 4812 Set the number of jiffies to delay each step of 4813 RCU grace-period initialization. 4814 4815 rcutree.gp_preinit_delay= [KNL] 4816 Set the number of jiffies to delay each step of 4817 RCU grace-period pre-initialization, that is, 4818 the propagation of recent CPU-hotplug changes up 4819 the rcu_node combining tree. 4820 4821 rcutree.jiffies_till_first_fqs= [KNL] 4822 Set delay from grace-period initialization to 4823 first attempt to force quiescent states. 4824 Units are jiffies, minimum value is zero, 4825 and maximum value is HZ. 4826 4827 rcutree.jiffies_till_next_fqs= [KNL] 4828 Set delay between subsequent attempts to force 4829 quiescent states. Units are jiffies, minimum 4830 value is one, and maximum value is HZ. 4831 4832 rcutree.jiffies_till_sched_qs= [KNL] 4833 Set required age in jiffies for a 4834 given grace period before RCU starts 4835 soliciting quiescent-state help from 4836 rcu_note_context_switch() and cond_resched(). 4837 If not specified, the kernel will calculate 4838 a value based on the most recent settings 4839 of rcutree.jiffies_till_first_fqs 4840 and rcutree.jiffies_till_next_fqs. 4841 This calculated value may be viewed in 4842 rcutree.jiffies_to_sched_qs. Any attempt to set 4843 rcutree.jiffies_to_sched_qs will be cheerfully 4844 overwritten. 4845 4846 rcutree.kthread_prio= [KNL,BOOT] 4847 Set the SCHED_FIFO priority of the RCU per-CPU 4848 kthreads (rcuc/N). This value is also used for 4849 the priority of the RCU boost threads (rcub/N) 4850 and for the RCU grace-period kthreads (rcu_bh, 4851 rcu_preempt, and rcu_sched). If RCU_BOOST is 4852 set, valid values are 1-99 and the default is 1 4853 (the least-favored priority). Otherwise, when 4854 RCU_BOOST is not set, valid values are 0-99 and 4855 the default is zero (non-realtime operation). 4856 When RCU_NOCB_CPU is set, also adjust the 4857 priority of NOCB callback kthreads. 4858 4859 rcutree.nocb_nobypass_lim_per_jiffy= [KNL] 4860 On callback-offloaded (rcu_nocbs) CPUs, 4861 RCU reduces the lock contention that would 4862 otherwise be caused by callback floods through 4863 use of the ->nocb_bypass list. However, in the 4864 common non-flooded case, RCU queues directly to 4865 the main ->cblist in order to avoid the extra 4866 overhead of the ->nocb_bypass list and its lock. 4867 But if there are too many callbacks queued during 4868 a single jiffy, RCU pre-queues the callbacks into 4869 the ->nocb_bypass queue. The definition of "too 4870 many" is supplied by this kernel boot parameter. 4871 4872 rcutree.qhimark= [KNL] 4873 Set threshold of queued RCU callbacks beyond which 4874 batch limiting is disabled. 4875 4876 rcutree.qlowmark= [KNL] 4877 Set threshold of queued RCU callbacks below which 4878 batch limiting is re-enabled. 4879 4880 rcutree.qovld= [KNL] 4881 Set threshold of queued RCU callbacks beyond which 4882 RCU's force-quiescent-state scan will aggressively 4883 enlist help from cond_resched() and sched IPIs to 4884 help CPUs more quickly reach quiescent states. 4885 Set to less than zero to make this be set based 4886 on rcutree.qhimark at boot time and to zero to 4887 disable more aggressive help enlistment. 4888 4889 rcutree.rcu_delay_page_cache_fill_msec= [KNL] 4890 Set the page-cache refill delay (in milliseconds) 4891 in response to low-memory conditions. The range 4892 of permitted values is in the range 0:100000. 4893 4894 rcutree.rcu_divisor= [KNL] 4895 Set the shift-right count to use to compute 4896 the callback-invocation batch limit bl from 4897 the number of callbacks queued on this CPU. 4898 The result will be bounded below by the value of 4899 the rcutree.blimit kernel parameter. Every bl 4900 callbacks, the softirq handler will exit in 4901 order to allow the CPU to do other work. 4902 4903 Please note that this callback-invocation batch 4904 limit applies only to non-offloaded callback 4905 invocation. Offloaded callbacks are instead 4906 invoked in the context of an rcuoc kthread, which 4907 scheduler will preempt as it does any other task. 4908 4909 rcutree.rcu_fanout_exact= [KNL] 4910 Disable autobalancing of the rcu_node combining 4911 tree. This is used by rcutorture, and might 4912 possibly be useful for architectures having high 4913 cache-to-cache transfer latencies. 4914 4915 rcutree.rcu_fanout_leaf= [KNL] 4916 Change the number of CPUs assigned to each 4917 leaf rcu_node structure. Useful for very 4918 large systems, which will choose the value 64, 4919 and for NUMA systems with large remote-access 4920 latencies, which will choose a value aligned 4921 with the appropriate hardware boundaries. 4922 4923 rcutree.rcu_min_cached_objs= [KNL] 4924 Minimum number of objects which are cached and 4925 maintained per one CPU. Object size is equal 4926 to PAGE_SIZE. The cache allows to reduce the 4927 pressure to page allocator, also it makes the 4928 whole algorithm to behave better in low memory 4929 condition. 4930 4931 rcutree.rcu_nocb_gp_stride= [KNL] 4932 Set the number of NOCB callback kthreads in 4933 each group, which defaults to the square root 4934 of the number of CPUs. Larger numbers reduce 4935 the wakeup overhead on the global grace-period 4936 kthread, but increases that same overhead on 4937 each group's NOCB grace-period kthread. 4938 4939 rcutree.rcu_kick_kthreads= [KNL] 4940 Cause the grace-period kthread to get an extra 4941 wake_up() if it sleeps three times longer than 4942 it should at force-quiescent-state time. 4943 This wake_up() will be accompanied by a 4944 WARN_ONCE() splat and an ftrace_dump(). 4945 4946 rcutree.rcu_resched_ns= [KNL] 4947 Limit the time spend invoking a batch of RCU 4948 callbacks to the specified number of nanoseconds. 4949 By default, this limit is checked only once 4950 every 32 callbacks in order to limit the pain 4951 inflicted by local_clock() overhead. 4952 4953 rcutree.rcu_unlock_delay= [KNL] 4954 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 4955 this specifies an rcu_read_unlock()-time delay 4956 in microseconds. This defaults to zero. 4957 Larger delays increase the probability of 4958 catching RCU pointer leaks, that is, buggy use 4959 of RCU-protected pointers after the relevant 4960 rcu_read_unlock() has completed. 4961 4962 rcutree.sysrq_rcu= [KNL] 4963 Commandeer a sysrq key to dump out Tree RCU's 4964 rcu_node tree with an eye towards determining 4965 why a new grace period has not yet started. 4966 4967 rcutree.use_softirq= [KNL] 4968 If set to zero, move all RCU_SOFTIRQ processing to 4969 per-CPU rcuc kthreads. Defaults to a non-zero 4970 value, meaning that RCU_SOFTIRQ is used by default. 4971 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4972 4973 But note that CONFIG_PREEMPT_RT=y kernels disable 4974 this kernel boot parameter, forcibly setting it 4975 to zero. 4976 4977 rcuscale.gp_async= [KNL] 4978 Measure performance of asynchronous 4979 grace-period primitives such as call_rcu(). 4980 4981 rcuscale.gp_async_max= [KNL] 4982 Specify the maximum number of outstanding 4983 callbacks per writer thread. When a writer 4984 thread exceeds this limit, it invokes the 4985 corresponding flavor of rcu_barrier() to allow 4986 previously posted callbacks to drain. 4987 4988 rcuscale.gp_exp= [KNL] 4989 Measure performance of expedited synchronous 4990 grace-period primitives. 4991 4992 rcuscale.holdoff= [KNL] 4993 Set test-start holdoff period. The purpose of 4994 this parameter is to delay the start of the 4995 test until boot completes in order to avoid 4996 interference. 4997 4998 rcuscale.kfree_by_call_rcu= [KNL] 4999 In kernels built with CONFIG_RCU_LAZY=y, test 5000 call_rcu() instead of kfree_rcu(). 5001 5002 rcuscale.kfree_mult= [KNL] 5003 Instead of allocating an object of size kfree_obj, 5004 allocate one of kfree_mult * sizeof(kfree_obj). 5005 Defaults to 1. 5006 5007 rcuscale.kfree_rcu_test= [KNL] 5008 Set to measure performance of kfree_rcu() flooding. 5009 5010 rcuscale.kfree_rcu_test_double= [KNL] 5011 Test the double-argument variant of kfree_rcu(). 5012 If this parameter has the same value as 5013 rcuscale.kfree_rcu_test_single, both the single- 5014 and double-argument variants are tested. 5015 5016 rcuscale.kfree_rcu_test_single= [KNL] 5017 Test the single-argument variant of kfree_rcu(). 5018 If this parameter has the same value as 5019 rcuscale.kfree_rcu_test_double, both the single- 5020 and double-argument variants are tested. 5021 5022 rcuscale.kfree_nthreads= [KNL] 5023 The number of threads running loops of kfree_rcu(). 5024 5025 rcuscale.kfree_alloc_num= [KNL] 5026 Number of allocations and frees done in an iteration. 5027 5028 rcuscale.kfree_loops= [KNL] 5029 Number of loops doing rcuscale.kfree_alloc_num number 5030 of allocations and frees. 5031 5032 rcuscale.minruntime= [KNL] 5033 Set the minimum test run time in seconds. This 5034 does not affect the data-collection interval, 5035 but instead allows better measurement of things 5036 like CPU consumption. 5037 5038 rcuscale.nreaders= [KNL] 5039 Set number of RCU readers. The value -1 selects 5040 N, where N is the number of CPUs. A value 5041 "n" less than -1 selects N-n+1, where N is again 5042 the number of CPUs. For example, -2 selects N 5043 (the number of CPUs), -3 selects N+1, and so on. 5044 A value of "n" less than or equal to -N selects 5045 a single reader. 5046 5047 rcuscale.nwriters= [KNL] 5048 Set number of RCU writers. The values operate 5049 the same as for rcuscale.nreaders. 5050 N, where N is the number of CPUs 5051 5052 rcuscale.scale_type= [KNL] 5053 Specify the RCU implementation to test. 5054 5055 rcuscale.shutdown= [KNL] 5056 Shut the system down after performance tests 5057 complete. This is useful for hands-off automated 5058 testing. 5059 5060 rcuscale.verbose= [KNL] 5061 Enable additional printk() statements. 5062 5063 rcuscale.writer_holdoff= [KNL] 5064 Write-side holdoff between grace periods, 5065 in microseconds. The default of zero says 5066 no holdoff. 5067 5068 rcuscale.writer_holdoff_jiffies= [KNL] 5069 Additional write-side holdoff between grace 5070 periods, but in jiffies. The default of zero 5071 says no holdoff. 5072 5073 rcutorture.fqs_duration= [KNL] 5074 Set duration of force_quiescent_state bursts 5075 in microseconds. 5076 5077 rcutorture.fqs_holdoff= [KNL] 5078 Set holdoff time within force_quiescent_state bursts 5079 in microseconds. 5080 5081 rcutorture.fqs_stutter= [KNL] 5082 Set wait time between force_quiescent_state bursts 5083 in seconds. 5084 5085 rcutorture.fwd_progress= [KNL] 5086 Specifies the number of kthreads to be used 5087 for RCU grace-period forward-progress testing 5088 for the types of RCU supporting this notion. 5089 Defaults to 1 kthread, values less than zero or 5090 greater than the number of CPUs cause the number 5091 of CPUs to be used. 5092 5093 rcutorture.fwd_progress_div= [KNL] 5094 Specify the fraction of a CPU-stall-warning 5095 period to do tight-loop forward-progress testing. 5096 5097 rcutorture.fwd_progress_holdoff= [KNL] 5098 Number of seconds to wait between successive 5099 forward-progress tests. 5100 5101 rcutorture.fwd_progress_need_resched= [KNL] 5102 Enclose cond_resched() calls within checks for 5103 need_resched() during tight-loop forward-progress 5104 testing. 5105 5106 rcutorture.gp_cond= [KNL] 5107 Use conditional/asynchronous update-side 5108 primitives, if available. 5109 5110 rcutorture.gp_exp= [KNL] 5111 Use expedited update-side primitives, if available. 5112 5113 rcutorture.gp_normal= [KNL] 5114 Use normal (non-expedited) asynchronous 5115 update-side primitives, if available. 5116 5117 rcutorture.gp_sync= [KNL] 5118 Use normal (non-expedited) synchronous 5119 update-side primitives, if available. If all 5120 of rcutorture.gp_cond=, rcutorture.gp_exp=, 5121 rcutorture.gp_normal=, and rcutorture.gp_sync= 5122 are zero, rcutorture acts as if is interpreted 5123 they are all non-zero. 5124 5125 rcutorture.irqreader= [KNL] 5126 Run RCU readers from irq handlers, or, more 5127 accurately, from a timer handler. Not all RCU 5128 flavors take kindly to this sort of thing. 5129 5130 rcutorture.leakpointer= [KNL] 5131 Leak an RCU-protected pointer out of the reader. 5132 This can of course result in splats, and is 5133 intended to test the ability of things like 5134 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 5135 such leaks. 5136 5137 rcutorture.n_barrier_cbs= [KNL] 5138 Set callbacks/threads for rcu_barrier() testing. 5139 5140 rcutorture.nfakewriters= [KNL] 5141 Set number of concurrent RCU writers. These just 5142 stress RCU, they don't participate in the actual 5143 test, hence the "fake". 5144 5145 rcutorture.nocbs_nthreads= [KNL] 5146 Set number of RCU callback-offload togglers. 5147 Zero (the default) disables toggling. 5148 5149 rcutorture.nocbs_toggle= [KNL] 5150 Set the delay in milliseconds between successive 5151 callback-offload toggling attempts. 5152 5153 rcutorture.nreaders= [KNL] 5154 Set number of RCU readers. The value -1 selects 5155 N-1, where N is the number of CPUs. A value 5156 "n" less than -1 selects N-n-2, where N is again 5157 the number of CPUs. For example, -2 selects N 5158 (the number of CPUs), -3 selects N+1, and so on. 5159 5160 rcutorture.object_debug= [KNL] 5161 Enable debug-object double-call_rcu() testing. 5162 5163 rcutorture.onoff_holdoff= [KNL] 5164 Set time (s) after boot for CPU-hotplug testing. 5165 5166 rcutorture.onoff_interval= [KNL] 5167 Set time (jiffies) between CPU-hotplug operations, 5168 or zero to disable CPU-hotplug testing. 5169 5170 rcutorture.read_exit= [KNL] 5171 Set the number of read-then-exit kthreads used 5172 to test the interaction of RCU updaters and 5173 task-exit processing. 5174 5175 rcutorture.read_exit_burst= [KNL] 5176 The number of times in a given read-then-exit 5177 episode that a set of read-then-exit kthreads 5178 is spawned. 5179 5180 rcutorture.read_exit_delay= [KNL] 5181 The delay, in seconds, between successive 5182 read-then-exit testing episodes. 5183 5184 rcutorture.shuffle_interval= [KNL] 5185 Set task-shuffle interval (s). Shuffling tasks 5186 allows some CPUs to go into dyntick-idle mode 5187 during the rcutorture test. 5188 5189 rcutorture.shutdown_secs= [KNL] 5190 Set time (s) after boot system shutdown. This 5191 is useful for hands-off automated testing. 5192 5193 rcutorture.stall_cpu= [KNL] 5194 Duration of CPU stall (s) to test RCU CPU stall 5195 warnings, zero to disable. 5196 5197 rcutorture.stall_cpu_block= [KNL] 5198 Sleep while stalling if set. This will result 5199 in warnings from preemptible RCU in addition to 5200 any other stall-related activity. Note that 5201 in kernels built with CONFIG_PREEMPTION=n and 5202 CONFIG_PREEMPT_COUNT=y, this parameter will 5203 cause the CPU to pass through a quiescent state. 5204 Given CONFIG_PREEMPTION=n, this will suppress 5205 RCU CPU stall warnings, but will instead result 5206 in scheduling-while-atomic splats. 5207 5208 Use of this module parameter results in splats. 5209 5210 5211 rcutorture.stall_cpu_holdoff= [KNL] 5212 Time to wait (s) after boot before inducing stall. 5213 5214 rcutorture.stall_cpu_irqsoff= [KNL] 5215 Disable interrupts while stalling if set. 5216 5217 rcutorture.stall_gp_kthread= [KNL] 5218 Duration (s) of forced sleep within RCU 5219 grace-period kthread to test RCU CPU stall 5220 warnings, zero to disable. If both stall_cpu 5221 and stall_gp_kthread are specified, the 5222 kthread is starved first, then the CPU. 5223 5224 rcutorture.stat_interval= [KNL] 5225 Time (s) between statistics printk()s. 5226 5227 rcutorture.stutter= [KNL] 5228 Time (s) to stutter testing, for example, specifying 5229 five seconds causes the test to run for five seconds, 5230 wait for five seconds, and so on. This tests RCU's 5231 ability to transition abruptly to and from idle. 5232 5233 rcutorture.test_boost= [KNL] 5234 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 5235 "Maybe" means test if the RCU implementation 5236 under test support RCU priority boosting. 5237 5238 rcutorture.test_boost_duration= [KNL] 5239 Duration (s) of each individual boost test. 5240 5241 rcutorture.test_boost_interval= [KNL] 5242 Interval (s) between each boost test. 5243 5244 rcutorture.test_no_idle_hz= [KNL] 5245 Test RCU's dyntick-idle handling. See also the 5246 rcutorture.shuffle_interval parameter. 5247 5248 rcutorture.torture_type= [KNL] 5249 Specify the RCU implementation to test. 5250 5251 rcutorture.verbose= [KNL] 5252 Enable additional printk() statements. 5253 5254 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 5255 Dump ftrace buffer after reporting RCU CPU 5256 stall warning. 5257 5258 rcupdate.rcu_cpu_stall_suppress= [KNL] 5259 Suppress RCU CPU stall warning messages. 5260 5261 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 5262 Suppress RCU CPU stall warning messages and 5263 rcutorture writer stall warnings that occur 5264 during early boot, that is, during the time 5265 before the init task is spawned. 5266 5267 rcupdate.rcu_cpu_stall_timeout= [KNL] 5268 Set timeout for RCU CPU stall warning messages. 5269 The value is in seconds and the maximum allowed 5270 value is 300 seconds. 5271 5272 rcupdate.rcu_exp_cpu_stall_timeout= [KNL] 5273 Set timeout for expedited RCU CPU stall warning 5274 messages. The value is in milliseconds 5275 and the maximum allowed value is 21000 5276 milliseconds. Please note that this value is 5277 adjusted to an arch timer tick resolution. 5278 Setting this to zero causes the value from 5279 rcupdate.rcu_cpu_stall_timeout to be used (after 5280 conversion from seconds to milliseconds). 5281 5282 rcupdate.rcu_cpu_stall_cputime= [KNL] 5283 Provide statistics on the cputime and count of 5284 interrupts and tasks during the sampling period. For 5285 multiple continuous RCU stalls, all sampling periods 5286 begin at half of the first RCU stall timeout. 5287 5288 rcupdate.rcu_exp_stall_task_details= [KNL] 5289 Print stack dumps of any tasks blocking the 5290 current expedited RCU grace period during an 5291 expedited RCU CPU stall warning. 5292 5293 rcupdate.rcu_expedited= [KNL] 5294 Use expedited grace-period primitives, for 5295 example, synchronize_rcu_expedited() instead 5296 of synchronize_rcu(). This reduces latency, 5297 but can increase CPU utilization, degrade 5298 real-time latency, and degrade energy efficiency. 5299 No effect on CONFIG_TINY_RCU kernels. 5300 5301 rcupdate.rcu_normal= [KNL] 5302 Use only normal grace-period primitives, 5303 for example, synchronize_rcu() instead of 5304 synchronize_rcu_expedited(). This improves 5305 real-time latency, CPU utilization, and 5306 energy efficiency, but can expose users to 5307 increased grace-period latency. This parameter 5308 overrides rcupdate.rcu_expedited. No effect on 5309 CONFIG_TINY_RCU kernels. 5310 5311 rcupdate.rcu_normal_after_boot= [KNL] 5312 Once boot has completed (that is, after 5313 rcu_end_inkernel_boot() has been invoked), use 5314 only normal grace-period primitives. No effect 5315 on CONFIG_TINY_RCU kernels. 5316 5317 But note that CONFIG_PREEMPT_RT=y kernels enables 5318 this kernel boot parameter, forcibly setting 5319 it to the value one, that is, converting any 5320 post-boot attempt at an expedited RCU grace 5321 period to instead use normal non-expedited 5322 grace-period processing. 5323 5324 rcupdate.rcu_task_collapse_lim= [KNL] 5325 Set the maximum number of callbacks present 5326 at the beginning of a grace period that allows 5327 the RCU Tasks flavors to collapse back to using 5328 a single callback queue. This switching only 5329 occurs when rcupdate.rcu_task_enqueue_lim is 5330 set to the default value of -1. 5331 5332 rcupdate.rcu_task_contend_lim= [KNL] 5333 Set the minimum number of callback-queuing-time 5334 lock-contention events per jiffy required to 5335 cause the RCU Tasks flavors to switch to per-CPU 5336 callback queuing. This switching only occurs 5337 when rcupdate.rcu_task_enqueue_lim is set to 5338 the default value of -1. 5339 5340 rcupdate.rcu_task_enqueue_lim= [KNL] 5341 Set the number of callback queues to use for the 5342 RCU Tasks family of RCU flavors. The default 5343 of -1 allows this to be automatically (and 5344 dynamically) adjusted. This parameter is intended 5345 for use in testing. 5346 5347 rcupdate.rcu_task_ipi_delay= [KNL] 5348 Set time in jiffies during which RCU tasks will 5349 avoid sending IPIs, starting with the beginning 5350 of a given grace period. Setting a large 5351 number avoids disturbing real-time workloads, 5352 but lengthens grace periods. 5353 5354 rcupdate.rcu_task_lazy_lim= [KNL] 5355 Number of callbacks on a given CPU that will 5356 cancel laziness on that CPU. Use -1 to disable 5357 cancellation of laziness, but be advised that 5358 doing so increases the danger of OOM due to 5359 callback flooding. 5360 5361 rcupdate.rcu_task_stall_info= [KNL] 5362 Set initial timeout in jiffies for RCU task stall 5363 informational messages, which give some indication 5364 of the problem for those not patient enough to 5365 wait for ten minutes. Informational messages are 5366 only printed prior to the stall-warning message 5367 for a given grace period. Disable with a value 5368 less than or equal to zero. Defaults to ten 5369 seconds. A change in value does not take effect 5370 until the beginning of the next grace period. 5371 5372 rcupdate.rcu_task_stall_info_mult= [KNL] 5373 Multiplier for time interval between successive 5374 RCU task stall informational messages for a given 5375 RCU tasks grace period. This value is clamped 5376 to one through ten, inclusive. It defaults to 5377 the value three, so that the first informational 5378 message is printed 10 seconds into the grace 5379 period, the second at 40 seconds, the third at 5380 160 seconds, and then the stall warning at 600 5381 seconds would prevent a fourth at 640 seconds. 5382 5383 rcupdate.rcu_task_stall_timeout= [KNL] 5384 Set timeout in jiffies for RCU task stall 5385 warning messages. Disable with a value less 5386 than or equal to zero. Defaults to ten minutes. 5387 A change in value does not take effect until 5388 the beginning of the next grace period. 5389 5390 rcupdate.rcu_tasks_lazy_ms= [KNL] 5391 Set timeout in milliseconds RCU Tasks asynchronous 5392 callback batching for call_rcu_tasks(). 5393 A negative value will take the default. A value 5394 of zero will disable batching. Batching is 5395 always disabled for synchronize_rcu_tasks(). 5396 5397 rcupdate.rcu_tasks_rude_lazy_ms= [KNL] 5398 Set timeout in milliseconds RCU Tasks 5399 Rude asynchronous callback batching for 5400 call_rcu_tasks_rude(). A negative value 5401 will take the default. A value of zero will 5402 disable batching. Batching is always disabled 5403 for synchronize_rcu_tasks_rude(). 5404 5405 rcupdate.rcu_tasks_trace_lazy_ms= [KNL] 5406 Set timeout in milliseconds RCU Tasks 5407 Trace asynchronous callback batching for 5408 call_rcu_tasks_trace(). A negative value 5409 will take the default. A value of zero will 5410 disable batching. Batching is always disabled 5411 for synchronize_rcu_tasks_trace(). 5412 5413 rcupdate.rcu_self_test= [KNL] 5414 Run the RCU early boot self tests 5415 5416 rdinit= [KNL] 5417 Format: <full_path> 5418 Run specified binary instead of /init from the ramdisk, 5419 used for early userspace startup. See initrd. 5420 5421 rdrand= [X86] 5422 force - Override the decision by the kernel to hide the 5423 advertisement of RDRAND support (this affects 5424 certain AMD processors because of buggy BIOS 5425 support, specifically around the suspend/resume 5426 path). 5427 5428 rdt= [HW,X86,RDT] 5429 Turn on/off individual RDT features. List is: 5430 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 5431 mba, smba, bmec. 5432 E.g. to turn on cmt and turn off mba use: 5433 rdt=cmt,!mba 5434 5435 reboot= [KNL] 5436 Format (x86 or x86_64): 5437 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] | d[efault] \ 5438 [[,]s[mp]#### \ 5439 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 5440 [[,]f[orce] 5441 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 5442 (prefix with 'panic_' to set mode for panic 5443 reboot only), 5444 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 5445 reboot_force is either force or not specified, 5446 reboot_cpu is s[mp]#### with #### being the processor 5447 to be used for rebooting. 5448 5449 refscale.holdoff= [KNL] 5450 Set test-start holdoff period. The purpose of 5451 this parameter is to delay the start of the 5452 test until boot completes in order to avoid 5453 interference. 5454 5455 refscale.loops= [KNL] 5456 Set the number of loops over the synchronization 5457 primitive under test. Increasing this number 5458 reduces noise due to loop start/end overhead, 5459 but the default has already reduced the per-pass 5460 noise to a handful of picoseconds on ca. 2020 5461 x86 laptops. 5462 5463 refscale.nreaders= [KNL] 5464 Set number of readers. The default value of -1 5465 selects N, where N is roughly 75% of the number 5466 of CPUs. A value of zero is an interesting choice. 5467 5468 refscale.nruns= [KNL] 5469 Set number of runs, each of which is dumped onto 5470 the console log. 5471 5472 refscale.readdelay= [KNL] 5473 Set the read-side critical-section duration, 5474 measured in microseconds. 5475 5476 refscale.scale_type= [KNL] 5477 Specify the read-protection implementation to test. 5478 5479 refscale.shutdown= [KNL] 5480 Shut down the system at the end of the performance 5481 test. This defaults to 1 (shut it down) when 5482 refscale is built into the kernel and to 0 (leave 5483 it running) when refscale is built as a module. 5484 5485 refscale.verbose= [KNL] 5486 Enable additional printk() statements. 5487 5488 refscale.verbose_batched= [KNL] 5489 Batch the additional printk() statements. If zero 5490 (the default) or negative, print everything. Otherwise, 5491 print every Nth verbose statement, where N is the value 5492 specified. 5493 5494 relax_domain_level= 5495 [KNL, SMP] Set scheduler's default relax_domain_level. 5496 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 5497 5498 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 5499 Format: <base1>,<size1>[,<base2>,<size2>,...] 5500 Reserve I/O ports or memory so the kernel won't use 5501 them. If <base> is less than 0x10000, the region 5502 is assumed to be I/O ports; otherwise it is memory. 5503 5504 reservetop= [X86-32] 5505 Format: nn[KMG] 5506 Reserves a hole at the top of the kernel virtual 5507 address space. 5508 5509 reset_devices [KNL] Force drivers to reset the underlying device 5510 during initialization. 5511 5512 resume= [SWSUSP] 5513 Specify the partition device for software suspend 5514 Format: 5515 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 5516 5517 resume_offset= [SWSUSP] 5518 Specify the offset from the beginning of the partition 5519 given by "resume=" at which the swap header is located, 5520 in <PAGE_SIZE> units (needed only for swap files). 5521 See Documentation/power/swsusp-and-swap-files.rst 5522 5523 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 5524 read the resume files 5525 5526 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 5527 Useful for devices that are detected asynchronously 5528 (e.g. USB and MMC devices). 5529 5530 retain_initrd [RAM] Keep initrd memory after extraction 5531 5532 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 5533 Speculative Code Execution with Return Instructions) 5534 vulnerability. 5535 5536 AMD-based UNRET and IBPB mitigations alone do not stop 5537 sibling threads from influencing the predictions of other 5538 sibling threads. For that reason, STIBP is used on pro- 5539 cessors that support it, and mitigate SMT on processors 5540 that don't. 5541 5542 off - no mitigation 5543 auto - automatically select a migitation 5544 auto,nosmt - automatically select a mitigation, 5545 disabling SMT if necessary for 5546 the full mitigation (only on Zen1 5547 and older without STIBP). 5548 ibpb - On AMD, mitigate short speculation 5549 windows on basic block boundaries too. 5550 Safe, highest perf impact. It also 5551 enables STIBP if present. Not suitable 5552 on Intel. 5553 ibpb,nosmt - Like "ibpb" above but will disable SMT 5554 when STIBP is not available. This is 5555 the alternative for systems which do not 5556 have STIBP. 5557 unret - Force enable untrained return thunks, 5558 only effective on AMD f15h-f17h based 5559 systems. 5560 unret,nosmt - Like unret, but will disable SMT when STIBP 5561 is not available. This is the alternative for 5562 systems which do not have STIBP. 5563 5564 Selecting 'auto' will choose a mitigation method at run 5565 time according to the CPU. 5566 5567 Not specifying this option is equivalent to retbleed=auto. 5568 5569 rfkill.default_state= 5570 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 5571 etc. communication is blocked by default. 5572 1 Unblocked. 5573 5574 rfkill.master_switch_mode= 5575 0 The "airplane mode" button does nothing. 5576 1 The "airplane mode" button toggles between everything 5577 blocked and the previous configuration. 5578 2 The "airplane mode" button toggles between everything 5579 blocked and everything unblocked. 5580 5581 rhash_entries= [KNL,NET] 5582 Set number of hash buckets for route cache 5583 5584 ring3mwait=disable 5585 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 5586 CPUs. 5587 5588 riscv_isa_fallback [RISCV] 5589 When CONFIG_RISCV_ISA_FALLBACK is not enabled, permit 5590 falling back to detecting extension support by parsing 5591 "riscv,isa" property on devicetree systems when the 5592 replacement properties are not found. See the Kconfig 5593 entry for RISCV_ISA_FALLBACK. 5594 5595 ro [KNL] Mount root device read-only on boot 5596 5597 rodata= [KNL] 5598 on Mark read-only kernel memory as read-only (default). 5599 off Leave read-only kernel memory writable for debugging. 5600 full Mark read-only kernel memory and aliases as read-only 5601 [arm64] 5602 5603 rockchip.usb_uart 5604 Enable the uart passthrough on the designated usb port 5605 on Rockchip SoCs. When active, the signals of the 5606 debug-uart get routed to the D+ and D- pins of the usb 5607 port and the regular usb controller gets disabled. 5608 5609 root= [KNL] Root filesystem 5610 Usually this a a block device specifier of some kind, 5611 see the early_lookup_bdev comment in 5612 block/early-lookup.c for details. 5613 Alternatively this can be "ram" for the legacy initial 5614 ramdisk, "nfs" and "cifs" for root on a network file 5615 system, or "mtd" and "ubi" for mounting from raw flash. 5616 5617 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 5618 mount the root filesystem 5619 5620 rootflags= [KNL] Set root filesystem mount option string 5621 5622 rootfstype= [KNL] Set root filesystem type 5623 5624 rootwait [KNL] Wait (indefinitely) for root device to show up. 5625 Useful for devices that are detected asynchronously 5626 (e.g. USB and MMC devices). 5627 5628 rootwait= [KNL] Maximum time (in seconds) to wait for root device 5629 to show up before attempting to mount the root 5630 filesystem. 5631 5632 rproc_mem=nn[KMG][@address] 5633 [KNL,ARM,CMA] Remoteproc physical memory block. 5634 Memory area to be used by remote processor image, 5635 managed by CMA. 5636 5637 rw [KNL] Mount root device read-write on boot 5638 5639 S [KNL] Run init in single mode 5640 5641 s390_iommu= [HW,S390] 5642 Set s390 IOTLB flushing mode 5643 strict 5644 With strict flushing every unmap operation will result in 5645 an IOTLB flush. Default is lazy flushing before reuse, 5646 which is faster. 5647 5648 s390_iommu_aperture= [KNL,S390] 5649 Specifies the size of the per device DMA address space 5650 accessible through the DMA and IOMMU APIs as a decimal 5651 factor of the size of main memory. 5652 The default is 1 meaning that one can concurrently use 5653 as many DMA addresses as physical memory is installed, 5654 if supported by hardware, and thus map all of memory 5655 once. With a value of 2 one can map all of memory twice 5656 and so on. As a special case a factor of 0 imposes no 5657 restrictions other than those given by hardware at the 5658 cost of significant additional memory use for tables. 5659 5660 sa1100ir [NET] 5661 See drivers/net/irda/sa1100_ir.c. 5662 5663 sched_verbose [KNL] Enables verbose scheduler debug messages. 5664 5665 schedstats= [KNL,X86] Enable or disable scheduled statistics. 5666 Allowed values are enable and disable. This feature 5667 incurs a small amount of overhead in the scheduler 5668 but is useful for debugging and performance tuning. 5669 5670 sched_thermal_decay_shift= 5671 [KNL, SMP] Set a decay shift for scheduler thermal 5672 pressure signal. Thermal pressure signal follows the 5673 default decay period of other scheduler pelt 5674 signals(usually 32 ms but configurable). Setting 5675 sched_thermal_decay_shift will left shift the decay 5676 period for the thermal pressure signal by the shift 5677 value. 5678 i.e. with the default pelt decay period of 32 ms 5679 sched_thermal_decay_shift thermal pressure decay pr 5680 1 64 ms 5681 2 128 ms 5682 and so on. 5683 Format: integer between 0 and 10 5684 Default is 0. 5685 5686 scftorture.holdoff= [KNL] 5687 Number of seconds to hold off before starting 5688 test. Defaults to zero for module insertion and 5689 to 10 seconds for built-in smp_call_function() 5690 tests. 5691 5692 scftorture.longwait= [KNL] 5693 Request ridiculously long waits randomly selected 5694 up to the chosen limit in seconds. Zero (the 5695 default) disables this feature. Please note 5696 that requesting even small non-zero numbers of 5697 seconds can result in RCU CPU stall warnings, 5698 softlockup complaints, and so on. 5699 5700 scftorture.nthreads= [KNL] 5701 Number of kthreads to spawn to invoke the 5702 smp_call_function() family of functions. 5703 The default of -1 specifies a number of kthreads 5704 equal to the number of CPUs. 5705 5706 scftorture.onoff_holdoff= [KNL] 5707 Number seconds to wait after the start of the 5708 test before initiating CPU-hotplug operations. 5709 5710 scftorture.onoff_interval= [KNL] 5711 Number seconds to wait between successive 5712 CPU-hotplug operations. Specifying zero (which 5713 is the default) disables CPU-hotplug operations. 5714 5715 scftorture.shutdown_secs= [KNL] 5716 The number of seconds following the start of the 5717 test after which to shut down the system. The 5718 default of zero avoids shutting down the system. 5719 Non-zero values are useful for automated tests. 5720 5721 scftorture.stat_interval= [KNL] 5722 The number of seconds between outputting the 5723 current test statistics to the console. A value 5724 of zero disables statistics output. 5725 5726 scftorture.stutter_cpus= [KNL] 5727 The number of jiffies to wait between each change 5728 to the set of CPUs under test. 5729 5730 scftorture.use_cpus_read_lock= [KNL] 5731 Use use_cpus_read_lock() instead of the default 5732 preempt_disable() to disable CPU hotplug 5733 while invoking one of the smp_call_function*() 5734 functions. 5735 5736 scftorture.verbose= [KNL] 5737 Enable additional printk() statements. 5738 5739 scftorture.weight_single= [KNL] 5740 The probability weighting to use for the 5741 smp_call_function_single() function with a zero 5742 "wait" parameter. A value of -1 selects the 5743 default if all other weights are -1. However, 5744 if at least one weight has some other value, a 5745 value of -1 will instead select a weight of zero. 5746 5747 scftorture.weight_single_wait= [KNL] 5748 The probability weighting to use for the 5749 smp_call_function_single() function with a 5750 non-zero "wait" parameter. See weight_single. 5751 5752 scftorture.weight_many= [KNL] 5753 The probability weighting to use for the 5754 smp_call_function_many() function with a zero 5755 "wait" parameter. See weight_single. 5756 Note well that setting a high probability for 5757 this weighting can place serious IPI load 5758 on the system. 5759 5760 scftorture.weight_many_wait= [KNL] 5761 The probability weighting to use for the 5762 smp_call_function_many() function with a 5763 non-zero "wait" parameter. See weight_single 5764 and weight_many. 5765 5766 scftorture.weight_all= [KNL] 5767 The probability weighting to use for the 5768 smp_call_function_all() function with a zero 5769 "wait" parameter. See weight_single and 5770 weight_many. 5771 5772 scftorture.weight_all_wait= [KNL] 5773 The probability weighting to use for the 5774 smp_call_function_all() function with a 5775 non-zero "wait" parameter. See weight_single 5776 and weight_many. 5777 5778 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 5779 xtime_lock contention on larger systems, and/or RCU lock 5780 contention on all systems with CONFIG_MAXSMP set. 5781 Format: { "0" | "1" } 5782 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 5783 1 -- enable. 5784 Note: increases power consumption, thus should only be 5785 enabled if running jitter sensitive (HPC/RT) workloads. 5786 5787 security= [SECURITY] Choose a legacy "major" security module to 5788 enable at boot. This has been deprecated by the 5789 "lsm=" parameter. 5790 5791 selinux= [SELINUX] Disable or enable SELinux at boot time. 5792 Format: { "0" | "1" } 5793 See security/selinux/Kconfig help text. 5794 0 -- disable. 5795 1 -- enable. 5796 Default value is 1. 5797 5798 serialnumber [BUGS=X86-32] 5799 5800 sev=option[,option...] [X86-64] See Documentation/arch/x86/x86_64/boot-options.rst 5801 5802 shapers= [NET] 5803 Maximal number of shapers. 5804 5805 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 5806 Limit apic dumping. The parameter defines the maximal 5807 number of local apics being dumped. Also it is possible 5808 to set it to "all" by meaning -- no limit here. 5809 Format: { 1 (default) | 2 | ... | all }. 5810 The parameter valid if only apic=debug or 5811 apic=verbose is specified. 5812 Example: apic=debug show_lapic=all 5813 5814 simeth= [IA-64] 5815 simscsi= 5816 5817 slram= [HW,MTD] 5818 5819 slab_merge [MM] 5820 Enable merging of slabs with similar size when the 5821 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 5822 5823 slab_nomerge [MM] 5824 Disable merging of slabs with similar size. May be 5825 necessary if there is some reason to distinguish 5826 allocs to different slabs, especially in hardened 5827 environments where the risk of heap overflows and 5828 layout control by attackers can usually be 5829 frustrated by disabling merging. This will reduce 5830 most of the exposure of a heap attack to a single 5831 cache (risks via metadata attacks are mostly 5832 unchanged). Debug options disable merging on their 5833 own. 5834 For more information see Documentation/mm/slub.rst. 5835 5836 slab_max_order= [MM, SLAB] 5837 Determines the maximum allowed order for slabs. 5838 A high setting may cause OOMs due to memory 5839 fragmentation. Defaults to 1 for systems with 5840 more than 32MB of RAM, 0 otherwise. 5841 5842 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB] 5843 Enabling slub_debug allows one to determine the 5844 culprit if slab objects become corrupted. Enabling 5845 slub_debug can create guard zones around objects and 5846 may poison objects when not in use. Also tracks the 5847 last alloc / free. For more information see 5848 Documentation/mm/slub.rst. 5849 5850 slub_max_order= [MM, SLUB] 5851 Determines the maximum allowed order for slabs. 5852 A high setting may cause OOMs due to memory 5853 fragmentation. For more information see 5854 Documentation/mm/slub.rst. 5855 5856 slub_min_objects= [MM, SLUB] 5857 The minimum number of objects per slab. SLUB will 5858 increase the slab order up to slub_max_order to 5859 generate a sufficiently large slab able to contain 5860 the number of objects indicated. The higher the number 5861 of objects the smaller the overhead of tracking slabs 5862 and the less frequently locks need to be acquired. 5863 For more information see Documentation/mm/slub.rst. 5864 5865 slub_min_order= [MM, SLUB] 5866 Determines the minimum page order for slabs. Must be 5867 lower than slub_max_order. 5868 For more information see Documentation/mm/slub.rst. 5869 5870 slub_merge [MM, SLUB] 5871 Same with slab_merge. 5872 5873 slub_nomerge [MM, SLUB] 5874 Same with slab_nomerge. This is supported for legacy. 5875 See slab_nomerge for more information. 5876 5877 smart2= [HW] 5878 Format: <io1>[,<io2>[,...,<io8>]] 5879 5880 smp.csd_lock_timeout= [KNL] 5881 Specify the period of time in milliseconds 5882 that smp_call_function() and friends will wait 5883 for a CPU to release the CSD lock. This is 5884 useful when diagnosing bugs involving CPUs 5885 disabling interrupts for extended periods 5886 of time. Defaults to 5,000 milliseconds, and 5887 setting a value of zero disables this feature. 5888 This feature may be more efficiently disabled 5889 using the csdlock_debug- kernel parameter. 5890 5891 smp.panic_on_ipistall= [KNL] 5892 If a csd_lock_timeout extends for more than 5893 the specified number of milliseconds, panic the 5894 system. By default, let CSD-lock acquisition 5895 take as long as they take. Specifying 300,000 5896 for this value provides a 5-minute timeout. 5897 5898 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 5899 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 5900 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 5901 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 5902 smsc-ircc2.ircc_irq= [HW] IRQ line 5903 smsc-ircc2.ircc_dma= [HW] DMA channel 5904 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 5905 0: Toshiba Satellite 1800 (GP data pin select) 5906 1: Fast pin select (default) 5907 2: ATC IRMode 5908 5909 smt= [KNL,MIPS,S390] Set the maximum number of threads (logical 5910 CPUs) to use per physical CPU on systems capable of 5911 symmetric multithreading (SMT). Will be capped to the 5912 actual hardware limit. 5913 Format: <integer> 5914 Default: -1 (no limit) 5915 5916 softlockup_panic= 5917 [KNL] Should the soft-lockup detector generate panics. 5918 Format: 0 | 1 5919 5920 A value of 1 instructs the soft-lockup detector 5921 to panic the machine when a soft-lockup occurs. It is 5922 also controlled by the kernel.softlockup_panic sysctl 5923 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 5924 respective build-time switch to that functionality. 5925 5926 softlockup_all_cpu_backtrace= 5927 [KNL] Should the soft-lockup detector generate 5928 backtraces on all cpus. 5929 Format: 0 | 1 5930 5931 sonypi.*= [HW] Sony Programmable I/O Control Device driver 5932 See Documentation/admin-guide/laptops/sonypi.rst 5933 5934 spectre_bhi= [X86] Control mitigation of Branch History Injection 5935 (BHI) vulnerability. This setting affects the 5936 deployment of the HW BHI control and the SW BHB 5937 clearing sequence. 5938 5939 on - (default) Enable the HW or SW mitigation 5940 as needed. 5941 off - Disable the mitigation. 5942 5943 spectre_v2= [X86] Control mitigation of Spectre variant 2 5944 (indirect branch speculation) vulnerability. 5945 The default operation protects the kernel from 5946 user space attacks. 5947 5948 on - unconditionally enable, implies 5949 spectre_v2_user=on 5950 off - unconditionally disable, implies 5951 spectre_v2_user=off 5952 auto - kernel detects whether your CPU model is 5953 vulnerable 5954 5955 Selecting 'on' will, and 'auto' may, choose a 5956 mitigation method at run time according to the 5957 CPU, the available microcode, the setting of the 5958 CONFIG_RETPOLINE configuration option, and the 5959 compiler with which the kernel was built. 5960 5961 Selecting 'on' will also enable the mitigation 5962 against user space to user space task attacks. 5963 5964 Selecting 'off' will disable both the kernel and 5965 the user space protections. 5966 5967 Specific mitigations can also be selected manually: 5968 5969 retpoline - replace indirect branches 5970 retpoline,generic - Retpolines 5971 retpoline,lfence - LFENCE; indirect branch 5972 retpoline,amd - alias for retpoline,lfence 5973 eibrs - Enhanced/Auto IBRS 5974 eibrs,retpoline - Enhanced/Auto IBRS + Retpolines 5975 eibrs,lfence - Enhanced/Auto IBRS + LFENCE 5976 ibrs - use IBRS to protect kernel 5977 5978 Not specifying this option is equivalent to 5979 spectre_v2=auto. 5980 5981 spectre_v2_user= 5982 [X86] Control mitigation of Spectre variant 2 5983 (indirect branch speculation) vulnerability between 5984 user space tasks 5985 5986 on - Unconditionally enable mitigations. Is 5987 enforced by spectre_v2=on 5988 5989 off - Unconditionally disable mitigations. Is 5990 enforced by spectre_v2=off 5991 5992 prctl - Indirect branch speculation is enabled, 5993 but mitigation can be enabled via prctl 5994 per thread. The mitigation control state 5995 is inherited on fork. 5996 5997 prctl,ibpb 5998 - Like "prctl" above, but only STIBP is 5999 controlled per thread. IBPB is issued 6000 always when switching between different user 6001 space processes. 6002 6003 seccomp 6004 - Same as "prctl" above, but all seccomp 6005 threads will enable the mitigation unless 6006 they explicitly opt out. 6007 6008 seccomp,ibpb 6009 - Like "seccomp" above, but only STIBP is 6010 controlled per thread. IBPB is issued 6011 always when switching between different 6012 user space processes. 6013 6014 auto - Kernel selects the mitigation depending on 6015 the available CPU features and vulnerability. 6016 6017 Default mitigation: "prctl" 6018 6019 Not specifying this option is equivalent to 6020 spectre_v2_user=auto. 6021 6022 spec_rstack_overflow= 6023 [X86] Control RAS overflow mitigation on AMD Zen CPUs 6024 6025 off - Disable mitigation 6026 microcode - Enable microcode mitigation only 6027 safe-ret - Enable sw-only safe RET mitigation (default) 6028 ibpb - Enable mitigation by issuing IBPB on 6029 kernel entry 6030 ibpb-vmexit - Issue IBPB only on VMEXIT 6031 (cloud-specific mitigation) 6032 6033 spec_store_bypass_disable= 6034 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 6035 (Speculative Store Bypass vulnerability) 6036 6037 Certain CPUs are vulnerable to an exploit against a 6038 a common industry wide performance optimization known 6039 as "Speculative Store Bypass" in which recent stores 6040 to the same memory location may not be observed by 6041 later loads during speculative execution. The idea 6042 is that such stores are unlikely and that they can 6043 be detected prior to instruction retirement at the 6044 end of a particular speculation execution window. 6045 6046 In vulnerable processors, the speculatively forwarded 6047 store can be used in a cache side channel attack, for 6048 example to read memory to which the attacker does not 6049 directly have access (e.g. inside sandboxed code). 6050 6051 This parameter controls whether the Speculative Store 6052 Bypass optimization is used. 6053 6054 On x86 the options are: 6055 6056 on - Unconditionally disable Speculative Store Bypass 6057 off - Unconditionally enable Speculative Store Bypass 6058 auto - Kernel detects whether the CPU model contains an 6059 implementation of Speculative Store Bypass and 6060 picks the most appropriate mitigation. If the 6061 CPU is not vulnerable, "off" is selected. If the 6062 CPU is vulnerable the default mitigation is 6063 architecture and Kconfig dependent. See below. 6064 prctl - Control Speculative Store Bypass per thread 6065 via prctl. Speculative Store Bypass is enabled 6066 for a process by default. The state of the control 6067 is inherited on fork. 6068 seccomp - Same as "prctl" above, but all seccomp threads 6069 will disable SSB unless they explicitly opt out. 6070 6071 Default mitigations: 6072 X86: "prctl" 6073 6074 On powerpc the options are: 6075 6076 on,auto - On Power8 and Power9 insert a store-forwarding 6077 barrier on kernel entry and exit. On Power7 6078 perform a software flush on kernel entry and 6079 exit. 6080 off - No action. 6081 6082 Not specifying this option is equivalent to 6083 spec_store_bypass_disable=auto. 6084 6085 spia_io_base= [HW,MTD] 6086 spia_fio_base= 6087 spia_pedr= 6088 spia_peddr= 6089 6090 split_lock_detect= 6091 [X86] Enable split lock detection or bus lock detection 6092 6093 When enabled (and if hardware support is present), atomic 6094 instructions that access data across cache line 6095 boundaries will result in an alignment check exception 6096 for split lock detection or a debug exception for 6097 bus lock detection. 6098 6099 off - not enabled 6100 6101 warn - the kernel will emit rate-limited warnings 6102 about applications triggering the #AC 6103 exception or the #DB exception. This mode is 6104 the default on CPUs that support split lock 6105 detection or bus lock detection. Default 6106 behavior is by #AC if both features are 6107 enabled in hardware. 6108 6109 fatal - the kernel will send SIGBUS to applications 6110 that trigger the #AC exception or the #DB 6111 exception. Default behavior is by #AC if 6112 both features are enabled in hardware. 6113 6114 ratelimit:N - 6115 Set system wide rate limit to N bus locks 6116 per second for bus lock detection. 6117 0 < N <= 1000. 6118 6119 N/A for split lock detection. 6120 6121 6122 If an #AC exception is hit in the kernel or in 6123 firmware (i.e. not while executing in user mode) 6124 the kernel will oops in either "warn" or "fatal" 6125 mode. 6126 6127 #DB exception for bus lock is triggered only when 6128 CPL > 0. 6129 6130 srbds= [X86,INTEL] 6131 Control the Special Register Buffer Data Sampling 6132 (SRBDS) mitigation. 6133 6134 Certain CPUs are vulnerable to an MDS-like 6135 exploit which can leak bits from the random 6136 number generator. 6137 6138 By default, this issue is mitigated by 6139 microcode. However, the microcode fix can cause 6140 the RDRAND and RDSEED instructions to become 6141 much slower. Among other effects, this will 6142 result in reduced throughput from /dev/urandom. 6143 6144 The microcode mitigation can be disabled with 6145 the following option: 6146 6147 off: Disable mitigation and remove 6148 performance impact to RDRAND and RDSEED 6149 6150 srcutree.big_cpu_lim [KNL] 6151 Specifies the number of CPUs constituting a 6152 large system, such that srcu_struct structures 6153 should immediately allocate an srcu_node array. 6154 This kernel-boot parameter defaults to 128, 6155 but takes effect only when the low-order four 6156 bits of srcutree.convert_to_big is equal to 3 6157 (decide at boot). 6158 6159 srcutree.convert_to_big [KNL] 6160 Specifies under what conditions an SRCU tree 6161 srcu_struct structure will be converted to big 6162 form, that is, with an rcu_node tree: 6163 6164 0: Never. 6165 1: At init_srcu_struct() time. 6166 2: When rcutorture decides to. 6167 3: Decide at boot time (default). 6168 0x1X: Above plus if high contention. 6169 6170 Either way, the srcu_node tree will be sized based 6171 on the actual runtime number of CPUs (nr_cpu_ids) 6172 instead of the compile-time CONFIG_NR_CPUS. 6173 6174 srcutree.counter_wrap_check [KNL] 6175 Specifies how frequently to check for 6176 grace-period sequence counter wrap for the 6177 srcu_data structure's ->srcu_gp_seq_needed field. 6178 The greater the number of bits set in this kernel 6179 parameter, the less frequently counter wrap will 6180 be checked for. Note that the bottom two bits 6181 are ignored. 6182 6183 srcutree.exp_holdoff [KNL] 6184 Specifies how many nanoseconds must elapse 6185 since the end of the last SRCU grace period for 6186 a given srcu_struct until the next normal SRCU 6187 grace period will be considered for automatic 6188 expediting. Set to zero to disable automatic 6189 expediting. 6190 6191 srcutree.srcu_max_nodelay [KNL] 6192 Specifies the number of no-delay instances 6193 per jiffy for which the SRCU grace period 6194 worker thread will be rescheduled with zero 6195 delay. Beyond this limit, worker thread will 6196 be rescheduled with a sleep delay of one jiffy. 6197 6198 srcutree.srcu_max_nodelay_phase [KNL] 6199 Specifies the per-grace-period phase, number of 6200 non-sleeping polls of readers. Beyond this limit, 6201 grace period worker thread will be rescheduled 6202 with a sleep delay of one jiffy, between each 6203 rescan of the readers, for a grace period phase. 6204 6205 srcutree.srcu_retry_check_delay [KNL] 6206 Specifies number of microseconds of non-sleeping 6207 delay between each non-sleeping poll of readers. 6208 6209 srcutree.small_contention_lim [KNL] 6210 Specifies the number of update-side contention 6211 events per jiffy will be tolerated before 6212 initiating a conversion of an srcu_struct 6213 structure to big form. Note that the value of 6214 srcutree.convert_to_big must have the 0x10 bit 6215 set for contention-based conversions to occur. 6216 6217 ssbd= [ARM64,HW] 6218 Speculative Store Bypass Disable control 6219 6220 On CPUs that are vulnerable to the Speculative 6221 Store Bypass vulnerability and offer a 6222 firmware based mitigation, this parameter 6223 indicates how the mitigation should be used: 6224 6225 force-on: Unconditionally enable mitigation for 6226 for both kernel and userspace 6227 force-off: Unconditionally disable mitigation for 6228 for both kernel and userspace 6229 kernel: Always enable mitigation in the 6230 kernel, and offer a prctl interface 6231 to allow userspace to register its 6232 interest in being mitigated too. 6233 6234 stack_guard_gap= [MM] 6235 override the default stack gap protection. The value 6236 is in page units and it defines how many pages prior 6237 to (for stacks growing down) resp. after (for stacks 6238 growing up) the main stack are reserved for no other 6239 mapping. Default value is 256 pages. 6240 6241 stack_depot_disable= [KNL] 6242 Setting this to true through kernel command line will 6243 disable the stack depot thereby saving the static memory 6244 consumed by the stack hash table. By default this is set 6245 to false. 6246 6247 stacktrace [FTRACE] 6248 Enabled the stack tracer on boot up. 6249 6250 stacktrace_filter=[function-list] 6251 [FTRACE] Limit the functions that the stack tracer 6252 will trace at boot up. function-list is a comma-separated 6253 list of functions. This list can be changed at run 6254 time by the stack_trace_filter file in the debugfs 6255 tracing directory. Note, this enables stack tracing 6256 and the stacktrace above is not needed. 6257 6258 sti= [PARISC,HW] 6259 Format: <num> 6260 Set the STI (builtin display/keyboard on the HP-PARISC 6261 machines) console (graphic card) which should be used 6262 as the initial boot-console. 6263 See also comment in drivers/video/console/sticore.c. 6264 6265 sti_font= [HW] 6266 See comment in drivers/video/console/sticore.c. 6267 6268 stifb= [HW] 6269 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 6270 6271 strict_sas_size= 6272 [X86] 6273 Format: <bool> 6274 Enable or disable strict sigaltstack size checks 6275 against the required signal frame size which 6276 depends on the supported FPU features. This can 6277 be used to filter out binaries which have 6278 not yet been made aware of AT_MINSIGSTKSZ. 6279 6280 stress_hpt [PPC] 6281 Limits the number of kernel HPT entries in the hash 6282 page table to increase the rate of hash page table 6283 faults on kernel addresses. 6284 6285 stress_slb [PPC] 6286 Limits the number of kernel SLB entries, and flushes 6287 them frequently to increase the rate of SLB faults 6288 on kernel addresses. 6289 6290 sunrpc.min_resvport= 6291 sunrpc.max_resvport= 6292 [NFS,SUNRPC] 6293 SunRPC servers often require that client requests 6294 originate from a privileged port (i.e. a port in the 6295 range 0 < portnr < 1024). 6296 An administrator who wishes to reserve some of these 6297 ports for other uses may adjust the range that the 6298 kernel's sunrpc client considers to be privileged 6299 using these two parameters to set the minimum and 6300 maximum port values. 6301 6302 sunrpc.svc_rpc_per_connection_limit= 6303 [NFS,SUNRPC] 6304 Limit the number of requests that the server will 6305 process in parallel from a single connection. 6306 The default value is 0 (no limit). 6307 6308 sunrpc.pool_mode= 6309 [NFS] 6310 Control how the NFS server code allocates CPUs to 6311 service thread pools. Depending on how many NICs 6312 you have and where their interrupts are bound, this 6313 option will affect which CPUs will do NFS serving. 6314 Note: this parameter cannot be changed while the 6315 NFS server is running. 6316 6317 auto the server chooses an appropriate mode 6318 automatically using heuristics 6319 global a single global pool contains all CPUs 6320 percpu one pool for each CPU 6321 pernode one pool for each NUMA node (equivalent 6322 to global on non-NUMA machines) 6323 6324 sunrpc.tcp_slot_table_entries= 6325 sunrpc.udp_slot_table_entries= 6326 [NFS,SUNRPC] 6327 Sets the upper limit on the number of simultaneous 6328 RPC calls that can be sent from the client to a 6329 server. Increasing these values may allow you to 6330 improve throughput, but will also increase the 6331 amount of memory reserved for use by the client. 6332 6333 suspend.pm_test_delay= 6334 [SUSPEND] 6335 Sets the number of seconds to remain in a suspend test 6336 mode before resuming the system (see 6337 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 6338 is set. Default value is 5. 6339 6340 svm= [PPC] 6341 Format: { on | off | y | n | 1 | 0 } 6342 This parameter controls use of the Protected 6343 Execution Facility on pSeries. 6344 6345 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 6346 Format: { <int> [,<int>] | force | noforce } 6347 <int> -- Number of I/O TLB slabs 6348 <int> -- Second integer after comma. Number of swiotlb 6349 areas with their own lock. Will be rounded up 6350 to a power of 2. 6351 force -- force using of bounce buffers even if they 6352 wouldn't be automatically used by the kernel 6353 noforce -- Never use bounce buffers (for debugging) 6354 6355 switches= [HW,M68k] 6356 6357 sysctl.*= [KNL] 6358 Set a sysctl parameter, right before loading the init 6359 process, as if the value was written to the respective 6360 /proc/sys/... file. Both '.' and '/' are recognized as 6361 separators. Unrecognized parameters and invalid values 6362 are reported in the kernel log. Sysctls registered 6363 later by a loaded module cannot be set this way. 6364 Example: sysctl.vm.swappiness=40 6365 6366 sysrq_always_enabled 6367 [KNL] 6368 Ignore sysrq setting - this boot parameter will 6369 neutralize any effect of /proc/sys/kernel/sysrq. 6370 Useful for debugging. 6371 6372 tcpmhash_entries= [KNL,NET] 6373 Set the number of tcp_metrics_hash slots. 6374 Default value is 8192 or 16384 depending on total 6375 ram pages. This is used to specify the TCP metrics 6376 cache size. See Documentation/networking/ip-sysctl.rst 6377 "tcp_no_metrics_save" section for more details. 6378 6379 tdfx= [HW,DRM] 6380 6381 test_suspend= [SUSPEND] 6382 Format: { "mem" | "standby" | "freeze" }[,N] 6383 Specify "mem" (for Suspend-to-RAM) or "standby" (for 6384 standby suspend) or "freeze" (for suspend type freeze) 6385 as the system sleep state during system startup with 6386 the optional capability to repeat N number of times. 6387 The system is woken from this state using a 6388 wakeup-capable RTC alarm. 6389 6390 thash_entries= [KNL,NET] 6391 Set number of hash buckets for TCP connection 6392 6393 thermal.act= [HW,ACPI] 6394 -1: disable all active trip points in all thermal zones 6395 <degrees C>: override all lowest active trip points 6396 6397 thermal.crt= [HW,ACPI] 6398 -1: disable all critical trip points in all thermal zones 6399 <degrees C>: override all critical trip points 6400 6401 thermal.off= [HW,ACPI] 6402 1: disable ACPI thermal control 6403 6404 thermal.psv= [HW,ACPI] 6405 -1: disable all passive trip points 6406 <degrees C>: override all passive trip points to this 6407 value 6408 6409 thermal.tzp= [HW,ACPI] 6410 Specify global default ACPI thermal zone polling rate 6411 <deci-seconds>: poll all this frequency 6412 0: no polling (default) 6413 6414 threadirqs [KNL] 6415 Force threading of all interrupt handlers except those 6416 marked explicitly IRQF_NO_THREAD. 6417 6418 topology= [S390] 6419 Format: {off | on} 6420 Specify if the kernel should make use of the cpu 6421 topology information if the hardware supports this. 6422 The scheduler will make use of this information and 6423 e.g. base its process migration decisions on it. 6424 Default is on. 6425 6426 topology_updates= [KNL, PPC, NUMA] 6427 Format: {off} 6428 Specify if the kernel should ignore (off) 6429 topology updates sent by the hypervisor to this 6430 LPAR. 6431 6432 torture.disable_onoff_at_boot= [KNL] 6433 Prevent the CPU-hotplug component of torturing 6434 until after init has spawned. 6435 6436 torture.ftrace_dump_at_shutdown= [KNL] 6437 Dump the ftrace buffer at torture-test shutdown, 6438 even if there were no errors. This can be a 6439 very costly operation when many torture tests 6440 are running concurrently, especially on systems 6441 with rotating-rust storage. 6442 6443 torture.verbose_sleep_frequency= [KNL] 6444 Specifies how many verbose printk()s should be 6445 emitted between each sleep. The default of zero 6446 disables verbose-printk() sleeping. 6447 6448 torture.verbose_sleep_duration= [KNL] 6449 Duration of each verbose-printk() sleep in jiffies. 6450 6451 tp720= [HW,PS2] 6452 6453 tpm_suspend_pcr=[HW,TPM] 6454 Format: integer pcr id 6455 Specify that at suspend time, the tpm driver 6456 should extend the specified pcr with zeros, 6457 as a workaround for some chips which fail to 6458 flush the last written pcr on TPM_SaveState. 6459 This will guarantee that all the other pcrs 6460 are saved. 6461 6462 tpm_tis.interrupts= [HW,TPM] 6463 Enable interrupts for the MMIO based physical layer 6464 for the FIFO interface. By default it is set to false 6465 (0). For more information about TPM hardware interfaces 6466 defined by Trusted Computing Group (TCG) see 6467 https://trustedcomputinggroup.org/resource/pc-client-platform-tpm-profile-ptp-specification/ 6468 6469 tp_printk [FTRACE] 6470 Have the tracepoints sent to printk as well as the 6471 tracing ring buffer. This is useful for early boot up 6472 where the system hangs or reboots and does not give the 6473 option for reading the tracing buffer or performing a 6474 ftrace_dump_on_oops. 6475 6476 To turn off having tracepoints sent to printk, 6477 echo 0 > /proc/sys/kernel/tracepoint_printk 6478 Note, echoing 1 into this file without the 6479 tracepoint_printk kernel cmdline option has no effect. 6480 6481 The tp_printk_stop_on_boot (see below) can also be used 6482 to stop the printing of events to console at 6483 late_initcall_sync. 6484 6485 ** CAUTION ** 6486 6487 Having tracepoints sent to printk() and activating high 6488 frequency tracepoints such as irq or sched, can cause 6489 the system to live lock. 6490 6491 tp_printk_stop_on_boot [FTRACE] 6492 When tp_printk (above) is set, it can cause a lot of noise 6493 on the console. It may be useful to only include the 6494 printing of events during boot up, as user space may 6495 make the system inoperable. 6496 6497 This command line option will stop the printing of events 6498 to console at the late_initcall_sync() time frame. 6499 6500 trace_buf_size=nn[KMG] 6501 [FTRACE] will set tracing buffer size on each cpu. 6502 6503 trace_clock= [FTRACE] Set the clock used for tracing events 6504 at boot up. 6505 local - Use the per CPU time stamp counter 6506 (converted into nanoseconds). Fast, but 6507 depending on the architecture, may not be 6508 in sync between CPUs. 6509 global - Event time stamps are synchronize across 6510 CPUs. May be slower than the local clock, 6511 but better for some race conditions. 6512 counter - Simple counting of events (1, 2, ..) 6513 note, some counts may be skipped due to the 6514 infrastructure grabbing the clock more than 6515 once per event. 6516 uptime - Use jiffies as the time stamp. 6517 perf - Use the same clock that perf uses. 6518 mono - Use ktime_get_mono_fast_ns() for time stamps. 6519 mono_raw - Use ktime_get_raw_fast_ns() for time 6520 stamps. 6521 boot - Use ktime_get_boot_fast_ns() for time stamps. 6522 Architectures may add more clocks. See 6523 Documentation/trace/ftrace.rst for more details. 6524 6525 trace_event=[event-list] 6526 [FTRACE] Set and start specified trace events in order 6527 to facilitate early boot debugging. The event-list is a 6528 comma-separated list of trace events to enable. See 6529 also Documentation/trace/events.rst 6530 6531 trace_instance=[instance-info] 6532 [FTRACE] Create a ring buffer instance early in boot up. 6533 This will be listed in: 6534 6535 /sys/kernel/tracing/instances 6536 6537 Events can be enabled at the time the instance is created 6538 via: 6539 6540 trace_instance=<name>,<system1>:<event1>,<system2>:<event2> 6541 6542 Note, the "<system*>:" portion is optional if the event is 6543 unique. 6544 6545 trace_instance=foo,sched:sched_switch,irq_handler_entry,initcall 6546 6547 will enable the "sched_switch" event (note, the "sched:" is optional, and 6548 the same thing would happen if it was left off). The irq_handler_entry 6549 event, and all events under the "initcall" system. 6550 6551 trace_options=[option-list] 6552 [FTRACE] Enable or disable tracer options at boot. 6553 The option-list is a comma delimited list of options 6554 that can be enabled or disabled just as if you were 6555 to echo the option name into 6556 6557 /sys/kernel/tracing/trace_options 6558 6559 For example, to enable stacktrace option (to dump the 6560 stack trace of each event), add to the command line: 6561 6562 trace_options=stacktrace 6563 6564 See also Documentation/trace/ftrace.rst "trace options" 6565 section. 6566 6567 trace_trigger=[trigger-list] 6568 [FTRACE] Add a event trigger on specific events. 6569 Set a trigger on top of a specific event, with an optional 6570 filter. 6571 6572 The format is is "trace_trigger=<event>.<trigger>[ if <filter>],..." 6573 Where more than one trigger may be specified that are comma deliminated. 6574 6575 For example: 6576 6577 trace_trigger="sched_switch.stacktrace if prev_state == 2" 6578 6579 The above will enable the "stacktrace" trigger on the "sched_switch" 6580 event but only trigger it if the "prev_state" of the "sched_switch" 6581 event is "2" (TASK_UNINTERUPTIBLE). 6582 6583 See also "Event triggers" in Documentation/trace/events.rst 6584 6585 6586 traceoff_on_warning 6587 [FTRACE] enable this option to disable tracing when a 6588 warning is hit. This turns off "tracing_on". Tracing can 6589 be enabled again by echoing '1' into the "tracing_on" 6590 file located in /sys/kernel/tracing/ 6591 6592 This option is useful, as it disables the trace before 6593 the WARNING dump is called, which prevents the trace to 6594 be filled with content caused by the warning output. 6595 6596 This option can also be set at run time via the sysctl 6597 option: kernel/traceoff_on_warning 6598 6599 transparent_hugepage= 6600 [KNL] 6601 Format: [always|madvise|never] 6602 Can be used to control the default behavior of the system 6603 with respect to transparent hugepages. 6604 See Documentation/admin-guide/mm/transhuge.rst 6605 for more details. 6606 6607 trusted.source= [KEYS] 6608 Format: <string> 6609 This parameter identifies the trust source as a backend 6610 for trusted keys implementation. Supported trust 6611 sources: 6612 - "tpm" 6613 - "tee" 6614 - "caam" 6615 If not specified then it defaults to iterating through 6616 the trust source list starting with TPM and assigns the 6617 first trust source as a backend which is initialized 6618 successfully during iteration. 6619 6620 trusted.rng= [KEYS] 6621 Format: <string> 6622 The RNG used to generate key material for trusted keys. 6623 Can be one of: 6624 - "kernel" 6625 - the same value as trusted.source: "tpm" or "tee" 6626 - "default" 6627 If not specified, "default" is used. In this case, 6628 the RNG's choice is left to each individual trust source. 6629 6630 tsc= Disable clocksource stability checks for TSC. 6631 Format: <string> 6632 [x86] reliable: mark tsc clocksource as reliable, this 6633 disables clocksource verification at runtime, as well 6634 as the stability checks done at bootup. Used to enable 6635 high-resolution timer mode on older hardware, and in 6636 virtualized environment. 6637 [x86] noirqtime: Do not use TSC to do irq accounting. 6638 Used to run time disable IRQ_TIME_ACCOUNTING on any 6639 platforms where RDTSC is slow and this accounting 6640 can add overhead. 6641 [x86] unstable: mark the TSC clocksource as unstable, this 6642 marks the TSC unconditionally unstable at bootup and 6643 avoids any further wobbles once the TSC watchdog notices. 6644 [x86] nowatchdog: disable clocksource watchdog. Used 6645 in situations with strict latency requirements (where 6646 interruptions from clocksource watchdog are not 6647 acceptable). 6648 [x86] recalibrate: force recalibration against a HW timer 6649 (HPET or PM timer) on systems whose TSC frequency was 6650 obtained from HW or FW using either an MSR or CPUID(0x15). 6651 Warn if the difference is more than 500 ppm. 6652 [x86] watchdog: Use TSC as the watchdog clocksource with 6653 which to check other HW timers (HPET or PM timer), but 6654 only on systems where TSC has been deemed trustworthy. 6655 This will be suppressed by an earlier tsc=nowatchdog and 6656 can be overridden by a later tsc=nowatchdog. A console 6657 message will flag any such suppression or overriding. 6658 6659 tsc_early_khz= [X86] Skip early TSC calibration and use the given 6660 value instead. Useful when the early TSC frequency discovery 6661 procedure is not reliable, such as on overclocked systems 6662 with CPUID.16h support and partial CPUID.15h support. 6663 Format: <unsigned int> 6664 6665 tsx= [X86] Control Transactional Synchronization 6666 Extensions (TSX) feature in Intel processors that 6667 support TSX control. 6668 6669 This parameter controls the TSX feature. The options are: 6670 6671 on - Enable TSX on the system. Although there are 6672 mitigations for all known security vulnerabilities, 6673 TSX has been known to be an accelerator for 6674 several previous speculation-related CVEs, and 6675 so there may be unknown security risks associated 6676 with leaving it enabled. 6677 6678 off - Disable TSX on the system. (Note that this 6679 option takes effect only on newer CPUs which are 6680 not vulnerable to MDS, i.e., have 6681 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 6682 the new IA32_TSX_CTRL MSR through a microcode 6683 update. This new MSR allows for the reliable 6684 deactivation of the TSX functionality.) 6685 6686 auto - Disable TSX if X86_BUG_TAA is present, 6687 otherwise enable TSX on the system. 6688 6689 Not specifying this option is equivalent to tsx=off. 6690 6691 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6692 for more details. 6693 6694 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 6695 Abort (TAA) vulnerability. 6696 6697 Similar to Micro-architectural Data Sampling (MDS) 6698 certain CPUs that support Transactional 6699 Synchronization Extensions (TSX) are vulnerable to an 6700 exploit against CPU internal buffers which can forward 6701 information to a disclosure gadget under certain 6702 conditions. 6703 6704 In vulnerable processors, the speculatively forwarded 6705 data can be used in a cache side channel attack, to 6706 access data to which the attacker does not have direct 6707 access. 6708 6709 This parameter controls the TAA mitigation. The 6710 options are: 6711 6712 full - Enable TAA mitigation on vulnerable CPUs 6713 if TSX is enabled. 6714 6715 full,nosmt - Enable TAA mitigation and disable SMT on 6716 vulnerable CPUs. If TSX is disabled, SMT 6717 is not disabled because CPU is not 6718 vulnerable to cross-thread TAA attacks. 6719 off - Unconditionally disable TAA mitigation 6720 6721 On MDS-affected machines, tsx_async_abort=off can be 6722 prevented by an active MDS mitigation as both vulnerabilities 6723 are mitigated with the same mechanism so in order to disable 6724 this mitigation, you need to specify mds=off too. 6725 6726 Not specifying this option is equivalent to 6727 tsx_async_abort=full. On CPUs which are MDS affected 6728 and deploy MDS mitigation, TAA mitigation is not 6729 required and doesn't provide any additional 6730 mitigation. 6731 6732 For details see: 6733 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 6734 6735 turbografx.map[2|3]= [HW,JOY] 6736 TurboGraFX parallel port interface 6737 Format: 6738 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 6739 See also Documentation/input/devices/joystick-parport.rst 6740 6741 udbg-immortal [PPC] When debugging early kernel crashes that 6742 happen after console_init() and before a proper 6743 console driver takes over, this boot options might 6744 help "seeing" what's going on. 6745 6746 uhash_entries= [KNL,NET] 6747 Set number of hash buckets for UDP/UDP-Lite connections 6748 6749 uhci-hcd.ignore_oc= 6750 [USB] Ignore overcurrent events (default N). 6751 Some badly-designed motherboards generate lots of 6752 bogus events, for ports that aren't wired to 6753 anything. Set this parameter to avoid log spamming. 6754 Note that genuine overcurrent events won't be 6755 reported either. 6756 6757 unknown_nmi_panic 6758 [X86] Cause panic on unknown NMI. 6759 6760 unwind_debug [X86-64] 6761 Enable unwinder debug output. This can be 6762 useful for debugging certain unwinder error 6763 conditions, including corrupt stacks and 6764 bad/missing unwinder metadata. 6765 6766 usbcore.authorized_default= 6767 [USB] Default USB device authorization: 6768 (default -1 = authorized (same as 1), 6769 0 = not authorized, 1 = authorized, 2 = authorized 6770 if device connected to internal port) 6771 6772 usbcore.autosuspend= 6773 [USB] The autosuspend time delay (in seconds) used 6774 for newly-detected USB devices (default 2). This 6775 is the time required before an idle device will be 6776 autosuspended. Devices for which the delay is set 6777 to a negative value won't be autosuspended at all. 6778 6779 usbcore.usbfs_snoop= 6780 [USB] Set to log all usbfs traffic (default 0 = off). 6781 6782 usbcore.usbfs_snoop_max= 6783 [USB] Maximum number of bytes to snoop in each URB 6784 (default = 65536). 6785 6786 usbcore.blinkenlights= 6787 [USB] Set to cycle leds on hubs (default 0 = off). 6788 6789 usbcore.old_scheme_first= 6790 [USB] Start with the old device initialization 6791 scheme (default 0 = off). 6792 6793 usbcore.usbfs_memory_mb= 6794 [USB] Memory limit (in MB) for buffers allocated by 6795 usbfs (default = 16, 0 = max = 2047). 6796 6797 usbcore.use_both_schemes= 6798 [USB] Try the other device initialization scheme 6799 if the first one fails (default 1 = enabled). 6800 6801 usbcore.initial_descriptor_timeout= 6802 [USB] Specifies timeout for the initial 64-byte 6803 USB_REQ_GET_DESCRIPTOR request in milliseconds 6804 (default 5000 = 5.0 seconds). 6805 6806 usbcore.nousb [USB] Disable the USB subsystem 6807 6808 usbcore.quirks= 6809 [USB] A list of quirk entries to augment the built-in 6810 usb core quirk list. List entries are separated by 6811 commas. Each entry has the form 6812 VendorID:ProductID:Flags. The IDs are 4-digit hex 6813 numbers and Flags is a set of letters. Each letter 6814 will change the built-in quirk; setting it if it is 6815 clear and clearing it if it is set. The letters have 6816 the following meanings: 6817 a = USB_QUIRK_STRING_FETCH_255 (string 6818 descriptors must not be fetched using 6819 a 255-byte read); 6820 b = USB_QUIRK_RESET_RESUME (device can't resume 6821 correctly so reset it instead); 6822 c = USB_QUIRK_NO_SET_INTF (device can't handle 6823 Set-Interface requests); 6824 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 6825 handle its Configuration or Interface 6826 strings); 6827 e = USB_QUIRK_RESET (device can't be reset 6828 (e.g morph devices), don't use reset); 6829 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 6830 more interface descriptions than the 6831 bNumInterfaces count, and can't handle 6832 talking to these interfaces); 6833 g = USB_QUIRK_DELAY_INIT (device needs a pause 6834 during initialization, after we read 6835 the device descriptor); 6836 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 6837 high speed and super speed interrupt 6838 endpoints, the USB 2.0 and USB 3.0 spec 6839 require the interval in microframes (1 6840 microframe = 125 microseconds) to be 6841 calculated as interval = 2 ^ 6842 (bInterval-1). 6843 Devices with this quirk report their 6844 bInterval as the result of this 6845 calculation instead of the exponent 6846 variable used in the calculation); 6847 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 6848 handle device_qualifier descriptor 6849 requests); 6850 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 6851 generates spurious wakeup, ignore 6852 remote wakeup capability); 6853 k = USB_QUIRK_NO_LPM (device can't handle Link 6854 Power Management); 6855 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 6856 (Device reports its bInterval as linear 6857 frames instead of the USB 2.0 6858 calculation); 6859 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 6860 to be disconnected before suspend to 6861 prevent spurious wakeup); 6862 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 6863 pause after every control message); 6864 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 6865 delay after resetting its port); 6866 p = USB_QUIRK_SHORT_SET_ADDRESS_REQ_TIMEOUT 6867 (Reduce timeout of the SET_ADDRESS 6868 request from 5000 ms to 500 ms); 6869 Example: quirks=0781:5580:bk,0a5c:5834:gij 6870 6871 usbhid.mousepoll= 6872 [USBHID] The interval which mice are to be polled at. 6873 6874 usbhid.jspoll= 6875 [USBHID] The interval which joysticks are to be polled at. 6876 6877 usbhid.kbpoll= 6878 [USBHID] The interval which keyboards are to be polled at. 6879 6880 usb-storage.delay_use= 6881 [UMS] The delay in seconds before a new device is 6882 scanned for Logical Units (default 1). 6883 6884 usb-storage.quirks= 6885 [UMS] A list of quirks entries to supplement or 6886 override the built-in unusual_devs list. List 6887 entries are separated by commas. Each entry has 6888 the form VID:PID:Flags where VID and PID are Vendor 6889 and Product ID values (4-digit hex numbers) and 6890 Flags is a set of characters, each corresponding 6891 to a common usb-storage quirk flag as follows: 6892 a = SANE_SENSE (collect more than 18 bytes 6893 of sense data, not on uas); 6894 b = BAD_SENSE (don't collect more than 18 6895 bytes of sense data, not on uas); 6896 c = FIX_CAPACITY (decrease the reported 6897 device capacity by one sector); 6898 d = NO_READ_DISC_INFO (don't use 6899 READ_DISC_INFO command, not on uas); 6900 e = NO_READ_CAPACITY_16 (don't use 6901 READ_CAPACITY_16 command); 6902 f = NO_REPORT_OPCODES (don't use report opcodes 6903 command, uas only); 6904 g = MAX_SECTORS_240 (don't transfer more than 6905 240 sectors at a time, uas only); 6906 h = CAPACITY_HEURISTICS (decrease the 6907 reported device capacity by one 6908 sector if the number is odd); 6909 i = IGNORE_DEVICE (don't bind to this 6910 device); 6911 j = NO_REPORT_LUNS (don't use report luns 6912 command, uas only); 6913 k = NO_SAME (do not use WRITE_SAME, uas only) 6914 l = NOT_LOCKABLE (don't try to lock and 6915 unlock ejectable media, not on uas); 6916 m = MAX_SECTORS_64 (don't transfer more 6917 than 64 sectors = 32 KB at a time, 6918 not on uas); 6919 n = INITIAL_READ10 (force a retry of the 6920 initial READ(10) command, not on uas); 6921 o = CAPACITY_OK (accept the capacity 6922 reported by the device, not on uas); 6923 p = WRITE_CACHE (the device cache is ON 6924 by default, not on uas); 6925 r = IGNORE_RESIDUE (the device reports 6926 bogus residue values, not on uas); 6927 s = SINGLE_LUN (the device has only one 6928 Logical Unit); 6929 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 6930 commands, uas only); 6931 u = IGNORE_UAS (don't bind to the uas driver); 6932 w = NO_WP_DETECT (don't test whether the 6933 medium is write-protected). 6934 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 6935 even if the device claims no cache, 6936 not on uas) 6937 Example: quirks=0419:aaf5:rl,0421:0433:rc 6938 6939 user_debug= [KNL,ARM] 6940 Format: <int> 6941 See arch/arm/Kconfig.debug help text. 6942 1 - undefined instruction events 6943 2 - system calls 6944 4 - invalid data aborts 6945 8 - SIGSEGV faults 6946 16 - SIGBUS faults 6947 Example: user_debug=31 6948 6949 userpte= 6950 [X86] Flags controlling user PTE allocations. 6951 6952 nohigh = do not allocate PTE pages in 6953 HIGHMEM regardless of setting 6954 of CONFIG_HIGHPTE. 6955 6956 vdso= [X86,SH,SPARC] 6957 On X86_32, this is an alias for vdso32=. Otherwise: 6958 6959 vdso=1: enable VDSO (the default) 6960 vdso=0: disable VDSO mapping 6961 6962 vdso32= [X86] Control the 32-bit vDSO 6963 vdso32=1: enable 32-bit VDSO 6964 vdso32=0 or vdso32=2: disable 32-bit VDSO 6965 6966 See the help text for CONFIG_COMPAT_VDSO for more 6967 details. If CONFIG_COMPAT_VDSO is set, the default is 6968 vdso32=0; otherwise, the default is vdso32=1. 6969 6970 For compatibility with older kernels, vdso32=2 is an 6971 alias for vdso32=0. 6972 6973 Try vdso32=0 if you encounter an error that says: 6974 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 6975 6976 vector= [IA-64,SMP] 6977 vector=percpu: enable percpu vector domain 6978 6979 video= [FB] Frame buffer configuration 6980 See Documentation/fb/modedb.rst. 6981 6982 video.brightness_switch_enabled= [ACPI] 6983 Format: [0|1] 6984 If set to 1, on receiving an ACPI notify event 6985 generated by hotkey, video driver will adjust brightness 6986 level and then send out the event to user space through 6987 the allocated input device. If set to 0, video driver 6988 will only send out the event without touching backlight 6989 brightness level. 6990 default: 1 6991 6992 virtio_mmio.device= 6993 [VMMIO] Memory mapped virtio (platform) device. 6994 6995 <size>@<baseaddr>:<irq>[:<id>] 6996 where: 6997 <size> := size (can use standard suffixes 6998 like K, M and G) 6999 <baseaddr> := physical base address 7000 <irq> := interrupt number (as passed to 7001 request_irq()) 7002 <id> := (optional) platform device id 7003 example: 7004 virtio_mmio.device=1K@0x100b0000:48:7 7005 7006 Can be used multiple times for multiple devices. 7007 7008 vga= [BOOT,X86-32] Select a particular video mode 7009 See Documentation/arch/x86/boot.rst and 7010 Documentation/admin-guide/svga.rst. 7011 Use vga=ask for menu. 7012 This is actually a boot loader parameter; the value is 7013 passed to the kernel using a special protocol. 7014 7015 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 7016 May slow down system boot speed, especially when 7017 enabled on systems with a large amount of memory. 7018 All options are enabled by default, and this 7019 interface is meant to allow for selectively 7020 enabling or disabling specific virtual memory 7021 debugging features. 7022 7023 Available options are: 7024 P Enable page structure init time poisoning 7025 - Disable all of the above options 7026 7027 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 7028 size of <nn>. This can be used to increase the 7029 minimum size (128MB on x86). It can also be used to 7030 decrease the size and leave more room for directly 7031 mapped kernel RAM. 7032 7033 vmcp_cma=nn[MG] [KNL,S390] 7034 Sets the memory size reserved for contiguous memory 7035 allocations for the vmcp device driver. 7036 7037 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 7038 Format: <command> 7039 7040 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 7041 Format: <command> 7042 7043 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 7044 Format: <command> 7045 7046 vsyscall= [X86-64] 7047 Controls the behavior of vsyscalls (i.e. calls to 7048 fixed addresses of 0xffffffffff600x00 from legacy 7049 code). Most statically-linked binaries and older 7050 versions of glibc use these calls. Because these 7051 functions are at fixed addresses, they make nice 7052 targets for exploits that can control RIP. 7053 7054 emulate Vsyscalls turn into traps and are emulated 7055 reasonably safely. The vsyscall page is 7056 readable. 7057 7058 xonly [default] Vsyscalls turn into traps and are 7059 emulated reasonably safely. The vsyscall 7060 page is not readable. 7061 7062 none Vsyscalls don't work at all. This makes 7063 them quite hard to use for exploits but 7064 might break your system. 7065 7066 vt.color= [VT] Default text color. 7067 Format: 0xYX, X = foreground, Y = background. 7068 Default: 0x07 = light gray on black. 7069 7070 vt.cur_default= [VT] Default cursor shape. 7071 Format: 0xCCBBAA, where AA, BB, and CC are the same as 7072 the parameters of the <Esc>[?A;B;Cc escape sequence; 7073 see VGA-softcursor.txt. Default: 2 = underline. 7074 7075 vt.default_blu= [VT] 7076 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 7077 Change the default blue palette of the console. 7078 This is a 16-member array composed of values 7079 ranging from 0-255. 7080 7081 vt.default_grn= [VT] 7082 Format: <green0>,<green1>,<green2>,...,<green15> 7083 Change the default green palette of the console. 7084 This is a 16-member array composed of values 7085 ranging from 0-255. 7086 7087 vt.default_red= [VT] 7088 Format: <red0>,<red1>,<red2>,...,<red15> 7089 Change the default red palette of the console. 7090 This is a 16-member array composed of values 7091 ranging from 0-255. 7092 7093 vt.default_utf8= 7094 [VT] 7095 Format=<0|1> 7096 Set system-wide default UTF-8 mode for all tty's. 7097 Default is 1, i.e. UTF-8 mode is enabled for all 7098 newly opened terminals. 7099 7100 vt.global_cursor_default= 7101 [VT] 7102 Format=<-1|0|1> 7103 Set system-wide default for whether a cursor 7104 is shown on new VTs. Default is -1, 7105 i.e. cursors will be created by default unless 7106 overridden by individual drivers. 0 will hide 7107 cursors, 1 will display them. 7108 7109 vt.italic= [VT] Default color for italic text; 0-15. 7110 Default: 2 = green. 7111 7112 vt.underline= [VT] Default color for underlined text; 0-15. 7113 Default: 3 = cyan. 7114 7115 watchdog timers [HW,WDT] For information on watchdog timers, 7116 see Documentation/watchdog/watchdog-parameters.rst 7117 or other driver-specific files in the 7118 Documentation/watchdog/ directory. 7119 7120 watchdog_thresh= 7121 [KNL] 7122 Set the hard lockup detector stall duration 7123 threshold in seconds. The soft lockup detector 7124 threshold is set to twice the value. A value of 0 7125 disables both lockup detectors. Default is 10 7126 seconds. 7127 7128 workqueue.unbound_cpus= 7129 [KNL,SMP] Specify to constrain one or some CPUs 7130 to use in unbound workqueues. 7131 Format: <cpu-list> 7132 By default, all online CPUs are available for 7133 unbound workqueues. 7134 7135 workqueue.watchdog_thresh= 7136 If CONFIG_WQ_WATCHDOG is configured, workqueue can 7137 warn stall conditions and dump internal state to 7138 help debugging. 0 disables workqueue stall 7139 detection; otherwise, it's the stall threshold 7140 duration in seconds. The default value is 30 and 7141 it can be updated at runtime by writing to the 7142 corresponding sysfs file. 7143 7144 workqueue.cpu_intensive_thresh_us= 7145 Per-cpu work items which run for longer than this 7146 threshold are automatically considered CPU intensive 7147 and excluded from concurrency management to prevent 7148 them from noticeably delaying other per-cpu work 7149 items. Default is 10000 (10ms). 7150 7151 If CONFIG_WQ_CPU_INTENSIVE_REPORT is set, the kernel 7152 will report the work functions which violate this 7153 threshold repeatedly. They are likely good 7154 candidates for using WQ_UNBOUND workqueues instead. 7155 7156 workqueue.power_efficient 7157 Per-cpu workqueues are generally preferred because 7158 they show better performance thanks to cache 7159 locality; unfortunately, per-cpu workqueues tend to 7160 be more power hungry than unbound workqueues. 7161 7162 Enabling this makes the per-cpu workqueues which 7163 were observed to contribute significantly to power 7164 consumption unbound, leading to measurably lower 7165 power usage at the cost of small performance 7166 overhead. 7167 7168 The default value of this parameter is determined by 7169 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 7170 7171 workqueue.default_affinity_scope= 7172 Select the default affinity scope to use for unbound 7173 workqueues. Can be one of "cpu", "smt", "cache", 7174 "numa" and "system". Default is "cache". For more 7175 information, see the Affinity Scopes section in 7176 Documentation/core-api/workqueue.rst. 7177 7178 This can be changed after boot by writing to the 7179 matching /sys/module/workqueue/parameters file. All 7180 workqueues with the "default" affinity scope will be 7181 updated accordignly. 7182 7183 workqueue.debug_force_rr_cpu 7184 Workqueue used to implicitly guarantee that work 7185 items queued without explicit CPU specified are put 7186 on the local CPU. This guarantee is no longer true 7187 and while local CPU is still preferred work items 7188 may be put on foreign CPUs. This debug option 7189 forces round-robin CPU selection to flush out 7190 usages which depend on the now broken guarantee. 7191 When enabled, memory and cache locality will be 7192 impacted. 7193 7194 writecombine= [LOONGARCH] Control the MAT (Memory Access Type) of 7195 ioremap_wc(). 7196 7197 on - Enable writecombine, use WUC for ioremap_wc() 7198 off - Disable writecombine, use SUC for ioremap_wc() 7199 7200 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 7201 default x2apic cluster mode on platforms 7202 supporting x2apic. 7203 7204 xen_512gb_limit [KNL,X86-64,XEN] 7205 Restricts the kernel running paravirtualized under Xen 7206 to use only up to 512 GB of RAM. The reason to do so is 7207 crash analysis tools and Xen tools for doing domain 7208 save/restore/migration must be enabled to handle larger 7209 domains. 7210 7211 xen_emul_unplug= [HW,X86,XEN] 7212 Unplug Xen emulated devices 7213 Format: [unplug0,][unplug1] 7214 ide-disks -- unplug primary master IDE devices 7215 aux-ide-disks -- unplug non-primary-master IDE devices 7216 nics -- unplug network devices 7217 all -- unplug all emulated devices (NICs and IDE disks) 7218 unnecessary -- unplugging emulated devices is 7219 unnecessary even if the host did not respond to 7220 the unplug protocol 7221 never -- do not unplug even if version check succeeds 7222 7223 xen_legacy_crash [X86,XEN] 7224 Crash from Xen panic notifier, without executing late 7225 panic() code such as dumping handler. 7226 7227 xen_msr_safe= [X86,XEN] 7228 Format: <bool> 7229 Select whether to always use non-faulting (safe) MSR 7230 access functions when running as Xen PV guest. The 7231 default value is controlled by CONFIG_XEN_PV_MSR_SAFE. 7232 7233 xen_nopvspin [X86,XEN] 7234 Disables the qspinlock slowpath using Xen PV optimizations. 7235 This parameter is obsoleted by "nopvspin" parameter, which 7236 has equivalent effect for XEN platform. 7237 7238 xen_nopv [X86] 7239 Disables the PV optimizations forcing the HVM guest to 7240 run as generic HVM guest with no PV drivers. 7241 This option is obsoleted by the "nopv" option, which 7242 has equivalent effect for XEN platform. 7243 7244 xen_no_vector_callback 7245 [KNL,X86,XEN] Disable the vector callback for Xen 7246 event channel interrupts. 7247 7248 xen_scrub_pages= [XEN] 7249 Boolean option to control scrubbing pages before giving them back 7250 to Xen, for use by other domains. Can be also changed at runtime 7251 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 7252 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 7253 7254 xen_timer_slop= [X86-64,XEN] 7255 Set the timer slop (in nanoseconds) for the virtual Xen 7256 timers (default is 100000). This adjusts the minimum 7257 delta of virtualized Xen timers, where lower values 7258 improve timer resolution at the expense of processing 7259 more timer interrupts. 7260 7261 xen.balloon_boot_timeout= [XEN] 7262 The time (in seconds) to wait before giving up to boot 7263 in case initial ballooning fails to free enough memory. 7264 Applies only when running as HVM or PVH guest and 7265 started with less memory configured than allowed at 7266 max. Default is 180. 7267 7268 xen.event_eoi_delay= [XEN] 7269 How long to delay EOI handling in case of event 7270 storms (jiffies). Default is 10. 7271 7272 xen.event_loop_timeout= [XEN] 7273 After which time (jiffies) the event handling loop 7274 should start to delay EOI handling. Default is 2. 7275 7276 xen.fifo_events= [XEN] 7277 Boolean parameter to disable using fifo event handling 7278 even if available. Normally fifo event handling is 7279 preferred over the 2-level event handling, as it is 7280 fairer and the number of possible event channels is 7281 much higher. Default is on (use fifo events). 7282 7283 xirc2ps_cs= [NET,PCMCIA] 7284 Format: 7285 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 7286 7287 xive= [PPC] 7288 By default on POWER9 and above, the kernel will 7289 natively use the XIVE interrupt controller. This option 7290 allows the fallback firmware mode to be used: 7291 7292 off Fallback to firmware control of XIVE interrupt 7293 controller on both pseries and powernv 7294 platforms. Only useful on POWER9 and above. 7295 7296 xive.store-eoi=off [PPC] 7297 By default on POWER10 and above, the kernel will use 7298 stores for EOI handling when the XIVE interrupt mode 7299 is active. This option allows the XIVE driver to use 7300 loads instead, as on POWER9. 7301 7302 xhci-hcd.quirks [USB,KNL] 7303 A hex value specifying bitmask with supplemental xhci 7304 host controller quirks. Meaning of each bit can be 7305 consulted in header drivers/usb/host/xhci.h. 7306 7307 xmon [PPC] 7308 Format: { early | on | rw | ro | off } 7309 Controls if xmon debugger is enabled. Default is off. 7310 Passing only "xmon" is equivalent to "xmon=early". 7311 early Call xmon as early as possible on boot; xmon 7312 debugger is called from setup_arch(). 7313 on xmon debugger hooks will be installed so xmon 7314 is only called on a kernel crash. Default mode, 7315 i.e. either "ro" or "rw" mode, is controlled 7316 with CONFIG_XMON_DEFAULT_RO_MODE. 7317 rw xmon debugger hooks will be installed so xmon 7318 is called only on a kernel crash, mode is write, 7319 meaning SPR registers, memory and, other data 7320 can be written using xmon commands. 7321 ro same as "rw" option above but SPR registers, 7322 memory, and other data can't be written using 7323 xmon commands. 7324 off xmon is disabled. 7325 7326