1 acpi= [HW,ACPI,X86,ARM64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force" 14 are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 { vendor | video | native | none } 26 If set to vendor, prefer vendor-specific driver 27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 28 of the ACPI video.ko driver. 29 If set to video, use the ACPI video.ko driver. 30 If set to native, use the device's native backlight mode. 31 If set to none, disable the ACPI backlight interface. 32 33 acpi_force_32bit_fadt_addr 34 force FADT to use 32 bit addresses rather than the 35 64 bit X_* addresses. Some firmware have broken 64 36 bit addresses for force ACPI ignore these and use 37 the older legacy 32 bit addresses. 38 39 acpica_no_return_repair [HW, ACPI] 40 Disable AML predefined validation mechanism 41 This mechanism can repair the evaluation result to make 42 the return objects more ACPI specification compliant. 43 This option is useful for developers to identify the 44 root cause of an AML interpreter issue when the issue 45 has something to do with the repair mechanism. 46 47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 49 Format: <int> 50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 51 debug output. Bits in debug_layer correspond to a 52 _COMPONENT in an ACPI source file, e.g., 53 #define _COMPONENT ACPI_PCI_COMPONENT 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 PCI/PCI interrupt routing info messages: 64 acpi.debug_layer=0x400000 65 Enable AML "Debug" output, i.e., stores to the Debug 66 object while interpreting AML: 67 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 68 Enable all messages related to ACPI hardware: 69 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 70 71 Some values produce so much output that the system is 72 unusable. The "log_buf_len" parameter may be useful 73 if you need to capture more output. 74 75 acpi_enforce_resources= [ACPI] 76 { strict | lax | no } 77 Check for resource conflicts between native drivers 78 and ACPI OperationRegions (SystemIO and SystemMemory 79 only). IO ports and memory declared in ACPI might be 80 used by the ACPI subsystem in arbitrary AML code and 81 can interfere with legacy drivers. 82 strict (default): access to resources claimed by ACPI 83 is denied; legacy drivers trying to access reserved 84 resources will fail to bind to device using them. 85 lax: access to resources claimed by ACPI is allowed; 86 legacy drivers trying to access reserved resources 87 will bind successfully but a warning message is logged. 88 no: ACPI OperationRegions are not marked as reserved, 89 no further checks are performed. 90 91 acpi_force_table_verification [HW,ACPI] 92 Enable table checksum verification during early stage. 93 By default, this is disabled due to x86 early mapping 94 size limitation. 95 96 acpi_irq_balance [HW,ACPI] 97 ACPI will balance active IRQs 98 default in APIC mode 99 100 acpi_irq_nobalance [HW,ACPI] 101 ACPI will not move active IRQs (default) 102 default in PIC mode 103 104 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 105 Format: <irq>,<irq>... 106 107 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 108 use by PCI 109 Format: <irq>,<irq>... 110 111 acpi_mask_gpe= [HW,ACPI] 112 Due to the existence of _Lxx/_Exx, some GPEs triggered 113 by unsupported hardware/firmware features can result in 114 GPE floodings that cannot be automatically disabled by 115 the GPE dispatcher. 116 This facility can be used to prevent such uncontrolled 117 GPE floodings. 118 Format: <byte> 119 120 acpi_no_auto_serialize [HW,ACPI] 121 Disable auto-serialization of AML methods 122 AML control methods that contain the opcodes to create 123 named objects will be marked as "Serialized" by the 124 auto-serialization feature. 125 This feature is enabled by default. 126 This option allows to turn off the feature. 127 128 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 129 kernels. 130 131 acpi_no_static_ssdt [HW,ACPI] 132 Disable installation of static SSDTs at early boot time 133 By default, SSDTs contained in the RSDT/XSDT will be 134 installed automatically and they will appear under 135 /sys/firmware/acpi/tables. 136 This option turns off this feature. 137 Note that specifying this option does not affect 138 dynamic table installation which will install SSDT 139 tables to /sys/firmware/acpi/tables/dynamic. 140 141 acpi_no_watchdog [HW,ACPI,WDT] 142 Ignore the ACPI-based watchdog interface (WDAT) and let 143 a native driver control the watchdog device instead. 144 145 acpi_rsdp= [ACPI,EFI,KEXEC] 146 Pass the RSDP address to the kernel, mostly used 147 on machines running EFI runtime service to boot the 148 second kernel for kdump. 149 150 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 151 Format: To spoof as Windows 98: ="Microsoft Windows" 152 153 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 154 of 2 which is mandated by ACPI 6) as the supported ACPI 155 specification revision (when using this switch, it may 156 be necessary to carry out a cold reboot _twice_ in a 157 row to make it take effect on the platform firmware). 158 159 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 160 acpi_osi="string1" # add string1 161 acpi_osi="!string2" # remove string2 162 acpi_osi=!* # remove all strings 163 acpi_osi=! # disable all built-in OS vendor 164 strings 165 acpi_osi=!! # enable all built-in OS vendor 166 strings 167 acpi_osi= # disable all strings 168 169 'acpi_osi=!' can be used in combination with single or 170 multiple 'acpi_osi="string1"' to support specific OS 171 vendor string(s). Note that such command can only 172 affect the default state of the OS vendor strings, thus 173 it cannot affect the default state of the feature group 174 strings and the current state of the OS vendor strings, 175 specifying it multiple times through kernel command line 176 is meaningless. This command is useful when one do not 177 care about the state of the feature group strings which 178 should be controlled by the OSPM. 179 Examples: 180 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 181 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 182 can make '_OSI("Windows 2000")' TRUE. 183 184 'acpi_osi=' cannot be used in combination with other 185 'acpi_osi=' command lines, the _OSI method will not 186 exist in the ACPI namespace. NOTE that such command can 187 only affect the _OSI support state, thus specifying it 188 multiple times through kernel command line is also 189 meaningless. 190 Examples: 191 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 192 FALSE. 193 194 'acpi_osi=!*' can be used in combination with single or 195 multiple 'acpi_osi="string1"' to support specific 196 string(s). Note that such command can affect the 197 current state of both the OS vendor strings and the 198 feature group strings, thus specifying it multiple times 199 through kernel command line is meaningful. But it may 200 still not able to affect the final state of a string if 201 there are quirks related to this string. This command 202 is useful when one want to control the state of the 203 feature group strings to debug BIOS issues related to 204 the OSPM features. 205 Examples: 206 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 207 '_OSI("Module Device")' FALSE. 208 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 209 '_OSI("Module Device")' TRUE. 210 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 211 equivalent to 212 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 213 and 214 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 215 they all will make '_OSI("Windows 2000")' TRUE. 216 217 acpi_pm_good [X86] 218 Override the pmtimer bug detection: force the kernel 219 to assume that this machine's pmtimer latches its value 220 and always returns good values. 221 222 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 223 Format: { level | edge | high | low } 224 225 acpi_skip_timer_override [HW,ACPI] 226 Recognize and ignore IRQ0/pin2 Interrupt Override. 227 For broken nForce2 BIOS resulting in XT-PIC timer. 228 229 acpi_sleep= [HW,ACPI] Sleep options 230 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig, 231 old_ordering, nonvs, sci_force_enable, nobl } 232 See Documentation/power/video.rst for information on 233 s3_bios and s3_mode. 234 s3_beep is for debugging; it makes the PC's speaker beep 235 as soon as the kernel's real-mode entry point is called. 236 s4_nohwsig prevents ACPI hardware signature from being 237 used during resume from hibernation. 238 old_ordering causes the ACPI 1.0 ordering of the _PTS 239 control method, with respect to putting devices into 240 low power states, to be enforced (the ACPI 2.0 ordering 241 of _PTS is used by default). 242 nonvs prevents the kernel from saving/restoring the 243 ACPI NVS memory during suspend/hibernation and resume. 244 sci_force_enable causes the kernel to set SCI_EN directly 245 on resume from S1/S3 (which is against the ACPI spec, 246 but some broken systems don't work without it). 247 nobl causes the internal blacklist of systems known to 248 behave incorrectly in some ways with respect to system 249 suspend and resume to be ignored (use wisely). 250 251 acpi_use_timer_override [HW,ACPI] 252 Use timer override. For some broken Nvidia NF5 boards 253 that require a timer override, but don't have HPET 254 255 add_efi_memmap [EFI; X86] Include EFI memory map in 256 kernel's map of available physical RAM. 257 258 agp= [AGP] 259 { off | try_unsupported } 260 off: disable AGP support 261 try_unsupported: try to drive unsupported chipsets 262 (may crash computer or cause data corruption) 263 264 ALSA [HW,ALSA] 265 See Documentation/sound/alsa-configuration.rst 266 267 alignment= [KNL,ARM] 268 Allow the default userspace alignment fault handler 269 behaviour to be specified. Bit 0 enables warnings, 270 bit 1 enables fixups, and bit 2 sends a segfault. 271 272 align_va_addr= [X86-64] 273 Align virtual addresses by clearing slice [14:12] when 274 allocating a VMA at process creation time. This option 275 gives you up to 3% performance improvement on AMD F15h 276 machines (where it is enabled by default) for a 277 CPU-intensive style benchmark, and it can vary highly in 278 a microbenchmark depending on workload and compiler. 279 280 32: only for 32-bit processes 281 64: only for 64-bit processes 282 on: enable for both 32- and 64-bit processes 283 off: disable for both 32- and 64-bit processes 284 285 alloc_snapshot [FTRACE] 286 Allocate the ftrace snapshot buffer on boot up when the 287 main buffer is allocated. This is handy if debugging 288 and you need to use tracing_snapshot() on boot up, and 289 do not want to use tracing_snapshot_alloc() as it needs 290 to be done where GFP_KERNEL allocations are allowed. 291 292 amd_iommu= [HW,X86-64] 293 Pass parameters to the AMD IOMMU driver in the system. 294 Possible values are: 295 fullflush - enable flushing of IO/TLB entries when 296 they are unmapped. Otherwise they are 297 flushed before they will be reused, which 298 is a lot of faster 299 off - do not initialize any AMD IOMMU found in 300 the system 301 force_isolation - Force device isolation for all 302 devices. The IOMMU driver is not 303 allowed anymore to lift isolation 304 requirements as needed. This option 305 does not override iommu=pt 306 307 amd_iommu_dump= [HW,X86-64] 308 Enable AMD IOMMU driver option to dump the ACPI table 309 for AMD IOMMU. With this option enabled, AMD IOMMU 310 driver will print ACPI tables for AMD IOMMU during 311 IOMMU initialization. 312 313 amd_iommu_intr= [HW,X86-64] 314 Specifies one of the following AMD IOMMU interrupt 315 remapping modes: 316 legacy - Use legacy interrupt remapping mode. 317 vapic - Use virtual APIC mode, which allows IOMMU 318 to inject interrupts directly into guest. 319 This mode requires kvm-amd.avic=1. 320 (Default when IOMMU HW support is present.) 321 322 amijoy.map= [HW,JOY] Amiga joystick support 323 Map of devices attached to JOY0DAT and JOY1DAT 324 Format: <a>,<b> 325 See also Documentation/input/joydev/joystick.rst 326 327 analog.map= [HW,JOY] Analog joystick and gamepad support 328 Specifies type or capabilities of an analog joystick 329 connected to one of 16 gameports 330 Format: <type1>,<type2>,..<type16> 331 332 apc= [HW,SPARC] 333 Power management functions (SPARCstation-4/5 + deriv.) 334 Format: noidle 335 Disable APC CPU standby support. SPARCstation-Fox does 336 not play well with APC CPU idle - disable it if you have 337 APC and your system crashes randomly. 338 339 apic= [APIC,X86] Advanced Programmable Interrupt Controller 340 Change the output verbosity while booting 341 Format: { quiet (default) | verbose | debug } 342 Change the amount of debugging information output 343 when initialising the APIC and IO-APIC components. 344 For X86-32, this can also be used to specify an APIC 345 driver name. 346 Format: apic=driver_name 347 Examples: apic=bigsmp 348 349 apic_extnmi= [APIC,X86] External NMI delivery setting 350 Format: { bsp (default) | all | none } 351 bsp: External NMI is delivered only to CPU 0 352 all: External NMIs are broadcast to all CPUs as a 353 backup of CPU 0 354 none: External NMI is masked for all CPUs. This is 355 useful so that a dump capture kernel won't be 356 shot down by NMI 357 358 autoconf= [IPV6] 359 See Documentation/networking/ipv6.rst. 360 361 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 362 Limit apic dumping. The parameter defines the maximal 363 number of local apics being dumped. Also it is possible 364 to set it to "all" by meaning -- no limit here. 365 Format: { 1 (default) | 2 | ... | all }. 366 The parameter valid if only apic=debug or 367 apic=verbose is specified. 368 Example: apic=debug show_lapic=all 369 370 apm= [APM] Advanced Power Management 371 See header of arch/x86/kernel/apm_32.c. 372 373 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 374 Format: <io>,<irq>,<nodeID> 375 376 ataflop= [HW,M68k] 377 378 atarimouse= [HW,MOUSE] Atari Mouse 379 380 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 381 EzKey and similar keyboards 382 383 atkbd.reset= [HW] Reset keyboard during initialization 384 385 atkbd.set= [HW] Select keyboard code set 386 Format: <int> (2 = AT (default), 3 = PS/2) 387 388 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 389 keyboards 390 391 atkbd.softraw= [HW] Choose between synthetic and real raw mode 392 Format: <bool> (0 = real, 1 = synthetic (default)) 393 394 atkbd.softrepeat= [HW] 395 Use software keyboard repeat 396 397 audit= [KNL] Enable the audit sub-system 398 Format: { "0" | "1" | "off" | "on" } 399 0 | off - kernel audit is disabled and can not be 400 enabled until the next reboot 401 unset - kernel audit is initialized but disabled and 402 will be fully enabled by the userspace auditd. 403 1 | on - kernel audit is initialized and partially 404 enabled, storing at most audit_backlog_limit 405 messages in RAM until it is fully enabled by the 406 userspace auditd. 407 Default: unset 408 409 audit_backlog_limit= [KNL] Set the audit queue size limit. 410 Format: <int> (must be >=0) 411 Default: 64 412 413 bau= [X86_UV] Enable the BAU on SGI UV. The default 414 behavior is to disable the BAU (i.e. bau=0). 415 Format: { "0" | "1" } 416 0 - Disable the BAU. 417 1 - Enable the BAU. 418 unset - Disable the BAU. 419 420 baycom_epp= [HW,AX25] 421 Format: <io>,<mode> 422 423 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 424 Format: <io>,<mode> 425 See header of drivers/net/hamradio/baycom_par.c. 426 427 baycom_ser_fdx= [HW,AX25] 428 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 429 Format: <io>,<irq>,<mode>[,<baud>] 430 See header of drivers/net/hamradio/baycom_ser_fdx.c. 431 432 baycom_ser_hdx= [HW,AX25] 433 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 434 Format: <io>,<irq>,<mode> 435 See header of drivers/net/hamradio/baycom_ser_hdx.c. 436 437 blkdevparts= Manual partition parsing of block device(s) for 438 embedded devices based on command line input. 439 See Documentation/block/cmdline-partition.rst 440 441 boot_delay= Milliseconds to delay each printk during boot. 442 Values larger than 10 seconds (10000) are changed to 443 no delay (0). 444 Format: integer 445 446 bootconfig [KNL] 447 Extended command line options can be added to an initrd 448 and this will cause the kernel to look for it. 449 450 See Documentation/admin-guide/bootconfig.rst 451 452 bert_disable [ACPI] 453 Disable BERT OS support on buggy BIOSes. 454 455 bgrt_disable [ACPI][X86] 456 Disable BGRT to avoid flickering OEM logo. 457 458 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 459 bttv.radio= Most important insmod options are available as 460 kernel args too. 461 bttv.pll= See Documentation/admin-guide/media/bttv.rst 462 bttv.tuner= 463 464 bulk_remove=off [PPC] This parameter disables the use of the pSeries 465 firmware feature for flushing multiple hpte entries 466 at a time. 467 468 c101= [NET] Moxa C101 synchronous serial card 469 470 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 471 Sometimes CPU hardware bugs make them report the cache 472 size incorrectly. The kernel will attempt work arounds 473 to fix known problems, but for some CPUs it is not 474 possible to determine what the correct size should be. 475 This option provides an override for these situations. 476 477 carrier_timeout= 478 [NET] Specifies amount of time (in seconds) that 479 the kernel should wait for a network carrier. By default 480 it waits 120 seconds. 481 482 ca_keys= [KEYS] This parameter identifies a specific key(s) on 483 the system trusted keyring to be used for certificate 484 trust validation. 485 format: { id:<keyid> | builtin } 486 487 cca= [MIPS] Override the kernel pages' cache coherency 488 algorithm. Accepted values range from 0 to 7 489 inclusive. See arch/mips/include/asm/pgtable-bits.h 490 for platform specific values (SB1, Loongson3 and 491 others). 492 493 ccw_timeout_log [S390] 494 See Documentation/s390/common_io.rst for details. 495 496 cgroup_disable= [KNL] Disable a particular controller 497 Format: {name of the controller(s) to disable} 498 The effects of cgroup_disable=foo are: 499 - foo isn't auto-mounted if you mount all cgroups in 500 a single hierarchy 501 - foo isn't visible as an individually mountable 502 subsystem 503 {Currently only "memory" controller deal with this and 504 cut the overhead, others just disable the usage. So 505 only cgroup_disable=memory is actually worthy} 506 507 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 508 Format: { { controller | "all" | "named" } 509 [,{ controller | "all" | "named" }...] } 510 Like cgroup_disable, but only applies to cgroup v1; 511 the blacklisted controllers remain available in cgroup2. 512 "all" blacklists all controllers and "named" disables 513 named mounts. Specifying both "all" and "named" disables 514 all v1 hierarchies. 515 516 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 517 Format: <string> 518 nosocket -- Disable socket memory accounting. 519 nokmem -- Disable kernel memory accounting. 520 521 checkreqprot [SELINUX] Set initial checkreqprot flag value. 522 Format: { "0" | "1" } 523 See security/selinux/Kconfig help text. 524 0 -- check protection applied by kernel (includes 525 any implied execute protection). 526 1 -- check protection requested by application. 527 Default value is set via a kernel config option. 528 Value can be changed at runtime via 529 /sys/fs/selinux/checkreqprot. 530 Setting checkreqprot to 1 is deprecated. 531 532 cio_ignore= [S390] 533 See Documentation/s390/common_io.rst for details. 534 clk_ignore_unused 535 [CLK] 536 Prevents the clock framework from automatically gating 537 clocks that have not been explicitly enabled by a Linux 538 device driver but are enabled in hardware at reset or 539 by the bootloader/firmware. Note that this does not 540 force such clocks to be always-on nor does it reserve 541 those clocks in any way. This parameter is useful for 542 debug and development, but should not be needed on a 543 platform with proper driver support. For more 544 information, see Documentation/driver-api/clk.rst. 545 546 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 547 [Deprecated] 548 Forces specified clocksource (if available) to be used 549 when calculating gettimeofday(). If specified 550 clocksource is not available, it defaults to PIT. 551 Format: { pit | tsc | cyclone | pmtmr } 552 553 clocksource= Override the default clocksource 554 Format: <string> 555 Override the default clocksource and use the clocksource 556 with the name specified. 557 Some clocksource names to choose from, depending on 558 the platform: 559 [all] jiffies (this is the base, fallback clocksource) 560 [ACPI] acpi_pm 561 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 562 pxa_timer,timer3,32k_counter,timer0_1 563 [X86-32] pit,hpet,tsc; 564 scx200_hrt on Geode; cyclone on IBM x440 565 [MIPS] MIPS 566 [PARISC] cr16 567 [S390] tod 568 [SH] SuperH 569 [SPARC64] tick 570 [X86-64] hpet,tsc 571 572 clocksource.arm_arch_timer.evtstrm= 573 [ARM,ARM64] 574 Format: <bool> 575 Enable/disable the eventstream feature of the ARM 576 architected timer so that code using WFE-based polling 577 loops can be debugged more effectively on production 578 systems. 579 580 clearcpuid=BITNUM [X86] 581 Disable CPUID feature X for the kernel. See 582 arch/x86/include/asm/cpufeatures.h for the valid bit 583 numbers. Note the Linux specific bits are not necessarily 584 stable over kernel options, but the vendor specific 585 ones should be. 586 Also note that user programs calling CPUID directly 587 or using the feature without checking anything 588 will still see it. This just prevents it from 589 being used by the kernel or shown in /proc/cpuinfo. 590 Also note the kernel might malfunction if you disable 591 some critical bits. 592 593 cma=nn[MG]@[start[MG][-end[MG]]] 594 [ARM,X86,KNL] 595 Sets the size of kernel global memory area for 596 contiguous memory allocations and optionally the 597 placement constraint by the physical address range of 598 memory allocations. A value of 0 disables CMA 599 altogether. For more information, see 600 include/linux/dma-contiguous.h 601 602 cmo_free_hint= [PPC] Format: { yes | no } 603 Specify whether pages are marked as being inactive 604 when they are freed. This is used in CMO environments 605 to determine OS memory pressure for page stealing by 606 a hypervisor. 607 Default: yes 608 609 coherent_pool=nn[KMG] [ARM,KNL] 610 Sets the size of memory pool for coherent, atomic dma 611 allocations, by default set to 256K. 612 613 com20020= [HW,NET] ARCnet - COM20020 chipset 614 Format: 615 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 616 617 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 618 Format: <io>[,<irq>] 619 620 com90xx= [HW,NET] 621 ARCnet - COM90xx chipset (memory-mapped buffers) 622 Format: <io>[,<irq>[,<memstart>]] 623 624 condev= [HW,S390] console device 625 conmode= 626 627 console= [KNL] Output console device and options. 628 629 tty<n> Use the virtual console device <n>. 630 631 ttyS<n>[,options] 632 ttyUSB0[,options] 633 Use the specified serial port. The options are of 634 the form "bbbbpnf", where "bbbb" is the baud rate, 635 "p" is parity ("n", "o", or "e"), "n" is number of 636 bits, and "f" is flow control ("r" for RTS or 637 omit it). Default is "9600n8". 638 639 See Documentation/admin-guide/serial-console.rst for more 640 information. See 641 Documentation/networking/netconsole.rst for an 642 alternative. 643 644 uart[8250],io,<addr>[,options] 645 uart[8250],mmio,<addr>[,options] 646 uart[8250],mmio16,<addr>[,options] 647 uart[8250],mmio32,<addr>[,options] 648 uart[8250],0x<addr>[,options] 649 Start an early, polled-mode console on the 8250/16550 650 UART at the specified I/O port or MMIO address, 651 switching to the matching ttyS device later. 652 MMIO inter-register address stride is either 8-bit 653 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 654 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 655 to be equivalent to 'mmio'. 'options' are specified in 656 the same format described for ttyS above; if unspecified, 657 the h/w is not re-initialized. 658 659 hvc<n> Use the hypervisor console device <n>. This is for 660 both Xen and PowerPC hypervisors. 661 662 If the device connected to the port is not a TTY but a braille 663 device, prepend "brl," before the device type, for instance 664 console=brl,ttyS0 665 For now, only VisioBraille is supported. 666 667 console_msg_format= 668 [KNL] Change console messages format 669 default 670 By default we print messages on consoles in 671 "[time stamp] text\n" format (time stamp may not be 672 printed, depending on CONFIG_PRINTK_TIME or 673 `printk_time' param). 674 syslog 675 Switch to syslog format: "<%u>[time stamp] text\n" 676 IOW, each message will have a facility and loglevel 677 prefix. The format is similar to one used by syslog() 678 syscall, or to executing "dmesg -S --raw" or to reading 679 from /proc/kmsg. 680 681 consoleblank= [KNL] The console blank (screen saver) timeout in 682 seconds. A value of 0 disables the blank timer. 683 Defaults to 0. 684 685 coredump_filter= 686 [KNL] Change the default value for 687 /proc/<pid>/coredump_filter. 688 See also Documentation/filesystems/proc.rst. 689 690 coresight_cpu_debug.enable 691 [ARM,ARM64] 692 Format: <bool> 693 Enable/disable the CPU sampling based debugging. 694 0: default value, disable debugging 695 1: enable debugging at boot time 696 697 cpuidle.off=1 [CPU_IDLE] 698 disable the cpuidle sub-system 699 700 cpuidle.governor= 701 [CPU_IDLE] Name of the cpuidle governor to use. 702 703 cpufreq.off=1 [CPU_FREQ] 704 disable the cpufreq sub-system 705 706 cpu_init_udelay=N 707 [X86] Delay for N microsec between assert and de-assert 708 of APIC INIT to start processors. This delay occurs 709 on every CPU online, such as boot, and resume from suspend. 710 Default: 10000 711 712 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 713 Format: 714 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 715 716 crashkernel=size[KMG][@offset[KMG]] 717 [KNL] Using kexec, Linux can switch to a 'crash kernel' 718 upon panic. This parameter reserves the physical 719 memory region [offset, offset + size] for that kernel 720 image. If '@offset' is omitted, then a suitable offset 721 is selected automatically. 722 [KNL, x86_64] select a region under 4G first, and 723 fall back to reserve region above 4G when '@offset' 724 hasn't been specified. 725 See Documentation/admin-guide/kdump/kdump.rst for further details. 726 727 crashkernel=range1:size1[,range2:size2,...][@offset] 728 [KNL] Same as above, but depends on the memory 729 in the running system. The syntax of range is 730 start-[end] where start and end are both 731 a memory unit (amount[KMG]). See also 732 Documentation/admin-guide/kdump/kdump.rst for an example. 733 734 crashkernel=size[KMG],high 735 [KNL, x86_64] range could be above 4G. Allow kernel 736 to allocate physical memory region from top, so could 737 be above 4G if system have more than 4G ram installed. 738 Otherwise memory region will be allocated below 4G, if 739 available. 740 It will be ignored if crashkernel=X is specified. 741 crashkernel=size[KMG],low 742 [KNL, x86_64] range under 4G. When crashkernel=X,high 743 is passed, kernel could allocate physical memory region 744 above 4G, that cause second kernel crash on system 745 that require some amount of low memory, e.g. swiotlb 746 requires at least 64M+32K low memory, also enough extra 747 low memory is needed to make sure DMA buffers for 32-bit 748 devices won't run out. Kernel would try to allocate at 749 at least 256M below 4G automatically. 750 This one let user to specify own low range under 4G 751 for second kernel instead. 752 0: to disable low allocation. 753 It will be ignored when crashkernel=X,high is not used 754 or memory reserved is below 4G. 755 756 cryptomgr.notests 757 [KNL] Disable crypto self-tests 758 759 cs89x0_dma= [HW,NET] 760 Format: <dma> 761 762 cs89x0_media= [HW,NET] 763 Format: { rj45 | aui | bnc } 764 765 dasd= [HW,NET] 766 See header of drivers/s390/block/dasd_devmap.c. 767 768 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 769 (one device per port) 770 Format: <port#>,<type> 771 See also Documentation/input/devices/joystick-parport.rst 772 773 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 774 time. See 775 Documentation/admin-guide/dynamic-debug-howto.rst for 776 details. Deprecated, see dyndbg. 777 778 debug [KNL] Enable kernel debugging (events log level). 779 780 debug_boot_weak_hash 781 [KNL] Enable printing [hashed] pointers early in the 782 boot sequence. If enabled, we use a weak hash instead 783 of siphash to hash pointers. Use this option if you are 784 seeing instances of '(___ptrval___)') and need to see a 785 value (hashed pointer) instead. Cryptographically 786 insecure, please do not use on production kernels. 787 788 debug_locks_verbose= 789 [KNL] verbose self-tests 790 Format=<0|1> 791 Print debugging info while doing the locking API 792 self-tests. 793 We default to 0 (no extra messages), setting it to 794 1 will print _a lot_ more information - normally 795 only useful to kernel developers. 796 797 debug_objects [KNL] Enable object debugging 798 799 no_debug_objects 800 [KNL] Disable object debugging 801 802 debug_guardpage_minorder= 803 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 804 parameter allows control of the order of pages that will 805 be intentionally kept free (and hence protected) by the 806 buddy allocator. Bigger value increase the probability 807 of catching random memory corruption, but reduce the 808 amount of memory for normal system use. The maximum 809 possible value is MAX_ORDER/2. Setting this parameter 810 to 1 or 2 should be enough to identify most random 811 memory corruption problems caused by bugs in kernel or 812 driver code when a CPU writes to (or reads from) a 813 random memory location. Note that there exists a class 814 of memory corruptions problems caused by buggy H/W or 815 F/W or by drivers badly programing DMA (basically when 816 memory is written at bus level and the CPU MMU is 817 bypassed) which are not detectable by 818 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 819 tracking down these problems. 820 821 debug_pagealloc= 822 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 823 enables the feature at boot time. By default, it is 824 disabled and the system will work mostly the same as a 825 kernel built without CONFIG_DEBUG_PAGEALLOC. 826 Note: to get most of debug_pagealloc error reports, it's 827 useful to also enable the page_owner functionality. 828 on: enable the feature 829 830 debugpat [X86] Enable PAT debugging 831 832 decnet.addr= [HW,NET] 833 Format: <area>[,<node>] 834 See also Documentation/networking/decnet.rst. 835 836 default_hugepagesz= 837 [HW] The size of the default HugeTLB page. This is 838 the size represented by the legacy /proc/ hugepages 839 APIs. In addition, this is the default hugetlb size 840 used for shmget(), mmap() and mounting hugetlbfs 841 filesystems. If not specified, defaults to the 842 architecture's default huge page size. Huge page 843 sizes are architecture dependent. See also 844 Documentation/admin-guide/mm/hugetlbpage.rst. 845 Format: size[KMG] 846 847 deferred_probe_timeout= 848 [KNL] Debugging option to set a timeout in seconds for 849 deferred probe to give up waiting on dependencies to 850 probe. Only specific dependencies (subsystems or 851 drivers) that have opted in will be ignored. A timeout of 0 852 will timeout at the end of initcalls. This option will also 853 dump out devices still on the deferred probe list after 854 retrying. 855 856 dfltcc= [HW,S390] 857 Format: { on | off | def_only | inf_only | always } 858 on: s390 zlib hardware support for compression on 859 level 1 and decompression (default) 860 off: No s390 zlib hardware support 861 def_only: s390 zlib hardware support for deflate 862 only (compression on level 1) 863 inf_only: s390 zlib hardware support for inflate 864 only (decompression) 865 always: Same as 'on' but ignores the selected compression 866 level always using hardware support (used for debugging) 867 868 dhash_entries= [KNL] 869 Set number of hash buckets for dentry cache. 870 871 disable_1tb_segments [PPC] 872 Disables the use of 1TB hash page table segments. This 873 causes the kernel to fall back to 256MB segments which 874 can be useful when debugging issues that require an SLB 875 miss to occur. 876 877 stress_slb [PPC] 878 Limits the number of kernel SLB entries, and flushes 879 them frequently to increase the rate of SLB faults 880 on kernel addresses. 881 882 disable= [IPV6] 883 See Documentation/networking/ipv6.rst. 884 885 hardened_usercopy= 886 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 887 hardening is enabled for this boot. Hardened 888 usercopy checking is used to protect the kernel 889 from reading or writing beyond known memory 890 allocation boundaries as a proactive defense 891 against bounds-checking flaws in the kernel's 892 copy_to_user()/copy_from_user() interface. 893 on Perform hardened usercopy checks (default). 894 off Disable hardened usercopy checks. 895 896 disable_radix [PPC] 897 Disable RADIX MMU mode on POWER9 898 899 radix_hcall_invalidate=on [PPC/PSERIES] 900 Disable RADIX GTSE feature and use hcall for TLB 901 invalidate. 902 903 disable_tlbie [PPC] 904 Disable TLBIE instruction. Currently does not work 905 with KVM, with HASH MMU, or with coherent accelerators. 906 907 disable_cpu_apicid= [X86,APIC,SMP] 908 Format: <int> 909 The number of initial APIC ID for the 910 corresponding CPU to be disabled at boot, 911 mostly used for the kdump 2nd kernel to 912 disable BSP to wake up multiple CPUs without 913 causing system reset or hang due to sending 914 INIT from AP to BSP. 915 916 perf_v4_pmi= [X86,INTEL] 917 Format: <bool> 918 Disable Intel PMU counter freezing feature. 919 The feature only exists starting from 920 Arch Perfmon v4 (Skylake and newer). 921 922 disable_ddw [PPC/PSERIES] 923 Disable Dynamic DMA Window support. Use this if 924 to workaround buggy firmware. 925 926 disable_ipv6= [IPV6] 927 See Documentation/networking/ipv6.rst. 928 929 disable_mtrr_cleanup [X86] 930 The kernel tries to adjust MTRR layout from continuous 931 to discrete, to make X server driver able to add WB 932 entry later. This parameter disables that. 933 934 disable_mtrr_trim [X86, Intel and AMD only] 935 By default the kernel will trim any uncacheable 936 memory out of your available memory pool based on 937 MTRR settings. This parameter disables that behavior, 938 possibly causing your machine to run very slowly. 939 940 disable_timer_pin_1 [X86] 941 Disable PIN 1 of APIC timer 942 Can be useful to work around chipset bugs. 943 944 dis_ucode_ldr [X86] Disable the microcode loader. 945 946 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 947 this option disables the debugging code at boot. 948 949 dma_debug_entries=<number> 950 This option allows to tune the number of preallocated 951 entries for DMA-API debugging code. One entry is 952 required per DMA-API allocation. Use this if the 953 DMA-API debugging code disables itself because the 954 architectural default is too low. 955 956 dma_debug_driver=<driver_name> 957 With this option the DMA-API debugging driver 958 filter feature can be enabled at boot time. Just 959 pass the driver to filter for as the parameter. 960 The filter can be disabled or changed to another 961 driver later using sysfs. 962 963 driver_async_probe= [KNL] 964 List of driver names to be probed asynchronously. 965 Format: <driver_name1>,<driver_name2>... 966 967 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 968 Broken monitors, graphic adapters, KVMs and EDIDless 969 panels may send no or incorrect EDID data sets. 970 This parameter allows to specify an EDID data sets 971 in the /lib/firmware directory that are used instead. 972 Generic built-in EDID data sets are used, if one of 973 edid/1024x768.bin, edid/1280x1024.bin, 974 edid/1680x1050.bin, or edid/1920x1080.bin is given 975 and no file with the same name exists. Details and 976 instructions how to build your own EDID data are 977 available in Documentation/admin-guide/edid.rst. An EDID 978 data set will only be used for a particular connector, 979 if its name and a colon are prepended to the EDID 980 name. Each connector may use a unique EDID data 981 set by separating the files with a comma. An EDID 982 data set with no connector name will be used for 983 any connectors not explicitly specified. 984 985 dscc4.setup= [NET] 986 987 dt_cpu_ftrs= [PPC] 988 Format: {"off" | "known"} 989 Control how the dt_cpu_ftrs device-tree binding is 990 used for CPU feature discovery and setup (if it 991 exists). 992 off: Do not use it, fall back to legacy cpu table. 993 known: Do not pass through unknown features to guests 994 or userspace, only those that the kernel is aware of. 995 996 dump_apple_properties [X86] 997 Dump name and content of EFI device properties on 998 x86 Macs. Useful for driver authors to determine 999 what data is available or for reverse-engineering. 1000 1001 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1002 module.dyndbg[="val"] 1003 Enable debug messages at boot time. See 1004 Documentation/admin-guide/dynamic-debug-howto.rst 1005 for details. 1006 1007 nopku [X86] Disable Memory Protection Keys CPU feature found 1008 in some Intel CPUs. 1009 1010 module.async_probe [KNL] 1011 Enable asynchronous probe on this module. 1012 1013 early_ioremap_debug [KNL] 1014 Enable debug messages in early_ioremap support. This 1015 is useful for tracking down temporary early mappings 1016 which are not unmapped. 1017 1018 earlycon= [KNL] Output early console device and options. 1019 1020 When used with no options, the early console is 1021 determined by stdout-path property in device tree's 1022 chosen node or the ACPI SPCR table if supported by 1023 the platform. 1024 1025 cdns,<addr>[,options] 1026 Start an early, polled-mode console on a Cadence 1027 (xuartps) serial port at the specified address. Only 1028 supported option is baud rate. If baud rate is not 1029 specified, the serial port must already be setup and 1030 configured. 1031 1032 uart[8250],io,<addr>[,options] 1033 uart[8250],mmio,<addr>[,options] 1034 uart[8250],mmio32,<addr>[,options] 1035 uart[8250],mmio32be,<addr>[,options] 1036 uart[8250],0x<addr>[,options] 1037 Start an early, polled-mode console on the 8250/16550 1038 UART at the specified I/O port or MMIO address. 1039 MMIO inter-register address stride is either 8-bit 1040 (mmio) or 32-bit (mmio32 or mmio32be). 1041 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1042 to be equivalent to 'mmio'. 'options' are specified 1043 in the same format described for "console=ttyS<n>"; if 1044 unspecified, the h/w is not initialized. 1045 1046 pl011,<addr> 1047 pl011,mmio32,<addr> 1048 Start an early, polled-mode console on a pl011 serial 1049 port at the specified address. The pl011 serial port 1050 must already be setup and configured. Options are not 1051 yet supported. If 'mmio32' is specified, then only 1052 the driver will use only 32-bit accessors to read/write 1053 the device registers. 1054 1055 meson,<addr> 1056 Start an early, polled-mode console on a meson serial 1057 port at the specified address. The serial port must 1058 already be setup and configured. Options are not yet 1059 supported. 1060 1061 msm_serial,<addr> 1062 Start an early, polled-mode console on an msm serial 1063 port at the specified address. The serial port 1064 must already be setup and configured. Options are not 1065 yet supported. 1066 1067 msm_serial_dm,<addr> 1068 Start an early, polled-mode console on an msm serial 1069 dm port at the specified address. The serial port 1070 must already be setup and configured. Options are not 1071 yet supported. 1072 1073 owl,<addr> 1074 Start an early, polled-mode console on a serial port 1075 of an Actions Semi SoC, such as S500 or S900, at the 1076 specified address. The serial port must already be 1077 setup and configured. Options are not yet supported. 1078 1079 rda,<addr> 1080 Start an early, polled-mode console on a serial port 1081 of an RDA Micro SoC, such as RDA8810PL, at the 1082 specified address. The serial port must already be 1083 setup and configured. Options are not yet supported. 1084 1085 sbi 1086 Use RISC-V SBI (Supervisor Binary Interface) for early 1087 console. 1088 1089 smh Use ARM semihosting calls for early console. 1090 1091 s3c2410,<addr> 1092 s3c2412,<addr> 1093 s3c2440,<addr> 1094 s3c6400,<addr> 1095 s5pv210,<addr> 1096 exynos4210,<addr> 1097 Use early console provided by serial driver available 1098 on Samsung SoCs, requires selecting proper type and 1099 a correct base address of the selected UART port. The 1100 serial port must already be setup and configured. 1101 Options are not yet supported. 1102 1103 lantiq,<addr> 1104 Start an early, polled-mode console on a lantiq serial 1105 (lqasc) port at the specified address. The serial port 1106 must already be setup and configured. Options are not 1107 yet supported. 1108 1109 lpuart,<addr> 1110 lpuart32,<addr> 1111 Use early console provided by Freescale LP UART driver 1112 found on Freescale Vybrid and QorIQ LS1021A processors. 1113 A valid base address must be provided, and the serial 1114 port must already be setup and configured. 1115 1116 ec_imx21,<addr> 1117 ec_imx6q,<addr> 1118 Start an early, polled-mode, output-only console on the 1119 Freescale i.MX UART at the specified address. The UART 1120 must already be setup and configured. 1121 1122 ar3700_uart,<addr> 1123 Start an early, polled-mode console on the 1124 Armada 3700 serial port at the specified 1125 address. The serial port must already be setup 1126 and configured. Options are not yet supported. 1127 1128 qcom_geni,<addr> 1129 Start an early, polled-mode console on a Qualcomm 1130 Generic Interface (GENI) based serial port at the 1131 specified address. The serial port must already be 1132 setup and configured. Options are not yet supported. 1133 1134 efifb,[options] 1135 Start an early, unaccelerated console on the EFI 1136 memory mapped framebuffer (if available). On cache 1137 coherent non-x86 systems that use system memory for 1138 the framebuffer, pass the 'ram' option so that it is 1139 mapped with the correct attributes. 1140 1141 linflex,<addr> 1142 Use early console provided by Freescale LINFlexD UART 1143 serial driver for NXP S32V234 SoCs. A valid base 1144 address must be provided, and the serial port must 1145 already be setup and configured. 1146 1147 earlyprintk= [X86,SH,ARM,M68k,S390] 1148 earlyprintk=vga 1149 earlyprintk=sclp 1150 earlyprintk=xen 1151 earlyprintk=serial[,ttySn[,baudrate]] 1152 earlyprintk=serial[,0x...[,baudrate]] 1153 earlyprintk=ttySn[,baudrate] 1154 earlyprintk=dbgp[debugController#] 1155 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1156 earlyprintk=xdbc[xhciController#] 1157 1158 earlyprintk is useful when the kernel crashes before 1159 the normal console is initialized. It is not enabled by 1160 default because it has some cosmetic problems. 1161 1162 Append ",keep" to not disable it when the real console 1163 takes over. 1164 1165 Only one of vga, efi, serial, or usb debug port can 1166 be used at a time. 1167 1168 Currently only ttyS0 and ttyS1 may be specified by 1169 name. Other I/O ports may be explicitly specified 1170 on some architectures (x86 and arm at least) by 1171 replacing ttySn with an I/O port address, like this: 1172 earlyprintk=serial,0x1008,115200 1173 You can find the port for a given device in 1174 /proc/tty/driver/serial: 1175 2: uart:ST16650V2 port:00001008 irq:18 ... 1176 1177 Interaction with the standard serial driver is not 1178 very good. 1179 1180 The VGA and EFI output is eventually overwritten by 1181 the real console. 1182 1183 The xen output can only be used by Xen PV guests. 1184 1185 The sclp output can only be used on s390. 1186 1187 The optional "force" to "pciserial" enables use of a 1188 PCI device even when its classcode is not of the 1189 UART class. 1190 1191 edac_report= [HW,EDAC] Control how to report EDAC event 1192 Format: {"on" | "off" | "force"} 1193 on: enable EDAC to report H/W event. May be overridden 1194 by other higher priority error reporting module. 1195 off: disable H/W event reporting through EDAC. 1196 force: enforce the use of EDAC to report H/W event. 1197 default: on. 1198 1199 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1200 ekgdboc=kbd 1201 1202 This is designed to be used in conjunction with 1203 the boot argument: earlyprintk=vga 1204 1205 This parameter works in place of the kgdboc parameter 1206 but can only be used if the backing tty is available 1207 very early in the boot process. For early debugging 1208 via a serial port see kgdboc_earlycon instead. 1209 1210 edd= [EDD] 1211 Format: {"off" | "on" | "skip[mbr]"} 1212 1213 efi= [EFI] 1214 Format: { "old_map", "nochunk", "noruntime", "debug", 1215 "nosoftreserve", "disable_early_pci_dma", 1216 "no_disable_early_pci_dma" } 1217 old_map [X86-64]: switch to the old ioremap-based EFI 1218 runtime services mapping. [Needs CONFIG_X86_UV=y] 1219 nochunk: disable reading files in "chunks" in the EFI 1220 boot stub, as chunking can cause problems with some 1221 firmware implementations. 1222 noruntime : disable EFI runtime services support 1223 debug: enable misc debug output 1224 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1225 attribute may cause the kernel to reserve the 1226 memory range for a memory mapping driver to 1227 claim. Specify efi=nosoftreserve to disable this 1228 reservation and treat the memory by its base type 1229 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1230 disable_early_pci_dma: Disable the busmaster bit on all 1231 PCI bridges while in the EFI boot stub 1232 no_disable_early_pci_dma: Leave the busmaster bit set 1233 on all PCI bridges while in the EFI boot stub 1234 1235 efi_no_storage_paranoia [EFI; X86] 1236 Using this parameter you can use more than 50% of 1237 your efi variable storage. Use this parameter only if 1238 you are really sure that your UEFI does sane gc and 1239 fulfills the spec otherwise your board may brick. 1240 1241 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1242 Add arbitrary attribute to specific memory range by 1243 updating original EFI memory map. 1244 Region of memory which aa attribute is added to is 1245 from ss to ss+nn. 1246 1247 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1248 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1249 attribute is added to range 0x100000000-0x180000000 and 1250 0x10a0000000-0x1120000000. 1251 1252 If efi_fake_mem=8G@9G:0x40000 is specified, the 1253 EFI_MEMORY_SP(0x40000) attribute is added to 1254 range 0x240000000-0x43fffffff. 1255 1256 Using this parameter you can do debugging of EFI memmap 1257 related features. For example, you can do debugging of 1258 Address Range Mirroring feature even if your box 1259 doesn't support it, or mark specific memory as 1260 "soft reserved". 1261 1262 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1263 that is to be dynamically loaded by Linux. If there are 1264 multiple variables with the same name but with different 1265 vendor GUIDs, all of them will be loaded. See 1266 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1267 1268 1269 eisa_irq_edge= [PARISC,HW] 1270 See header of drivers/parisc/eisa.c. 1271 1272 elanfreq= [X86-32] 1273 See comment before function elanfreq_setup() in 1274 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1275 1276 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1277 Specifies physical address of start of kernel core 1278 image elf header and optionally the size. Generally 1279 kexec loader will pass this option to capture kernel. 1280 See Documentation/admin-guide/kdump/kdump.rst for details. 1281 1282 enable_mtrr_cleanup [X86] 1283 The kernel tries to adjust MTRR layout from continuous 1284 to discrete, to make X server driver able to add WB 1285 entry later. This parameter enables that. 1286 1287 enable_timer_pin_1 [X86] 1288 Enable PIN 1 of APIC timer 1289 Can be useful to work around chipset bugs 1290 (in particular on some ATI chipsets). 1291 The kernel tries to set a reasonable default. 1292 1293 enforcing [SELINUX] Set initial enforcing status. 1294 Format: {"0" | "1"} 1295 See security/selinux/Kconfig help text. 1296 0 -- permissive (log only, no denials). 1297 1 -- enforcing (deny and log). 1298 Default value is 0. 1299 Value can be changed at runtime via 1300 /sys/fs/selinux/enforce. 1301 1302 erst_disable [ACPI] 1303 Disable Error Record Serialization Table (ERST) 1304 support. 1305 1306 ether= [HW,NET] Ethernet cards parameters 1307 This option is obsoleted by the "netdev=" option, which 1308 has equivalent usage. See its documentation for details. 1309 1310 evm= [EVM] 1311 Format: { "fix" } 1312 Permit 'security.evm' to be updated regardless of 1313 current integrity status. 1314 1315 failslab= 1316 fail_page_alloc= 1317 fail_make_request=[KNL] 1318 General fault injection mechanism. 1319 Format: <interval>,<probability>,<space>,<times> 1320 See also Documentation/fault-injection/. 1321 1322 floppy= [HW] 1323 See Documentation/admin-guide/blockdev/floppy.rst. 1324 1325 force_pal_cache_flush 1326 [IA-64] Avoid check_sal_cache_flush which may hang on 1327 buggy SAL_CACHE_FLUSH implementations. Using this 1328 parameter will force ia64_sal_cache_flush to call 1329 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1330 1331 forcepae [X86-32] 1332 Forcefully enable Physical Address Extension (PAE). 1333 Many Pentium M systems disable PAE but may have a 1334 functionally usable PAE implementation. 1335 Warning: use of this parameter will taint the kernel 1336 and may cause unknown problems. 1337 1338 ftrace=[tracer] 1339 [FTRACE] will set and start the specified tracer 1340 as early as possible in order to facilitate early 1341 boot debugging. 1342 1343 ftrace_dump_on_oops[=orig_cpu] 1344 [FTRACE] will dump the trace buffers on oops. 1345 If no parameter is passed, ftrace will dump 1346 buffers of all CPUs, but if you pass orig_cpu, it will 1347 dump only the buffer of the CPU that triggered the 1348 oops. 1349 1350 ftrace_filter=[function-list] 1351 [FTRACE] Limit the functions traced by the function 1352 tracer at boot up. function-list is a comma separated 1353 list of functions. This list can be changed at run 1354 time by the set_ftrace_filter file in the debugfs 1355 tracing directory. 1356 1357 ftrace_notrace=[function-list] 1358 [FTRACE] Do not trace the functions specified in 1359 function-list. This list can be changed at run time 1360 by the set_ftrace_notrace file in the debugfs 1361 tracing directory. 1362 1363 ftrace_graph_filter=[function-list] 1364 [FTRACE] Limit the top level callers functions traced 1365 by the function graph tracer at boot up. 1366 function-list is a comma separated list of functions 1367 that can be changed at run time by the 1368 set_graph_function file in the debugfs tracing directory. 1369 1370 ftrace_graph_notrace=[function-list] 1371 [FTRACE] Do not trace from the functions specified in 1372 function-list. This list is a comma separated list of 1373 functions that can be changed at run time by the 1374 set_graph_notrace file in the debugfs tracing directory. 1375 1376 ftrace_graph_max_depth=<uint> 1377 [FTRACE] Used with the function graph tracer. This is 1378 the max depth it will trace into a function. This value 1379 can be changed at run time by the max_graph_depth file 1380 in the tracefs tracing directory. default: 0 (no limit) 1381 1382 fw_devlink= [KNL] Create device links between consumer and supplier 1383 devices by scanning the firmware to infer the 1384 consumer/supplier relationships. This feature is 1385 especially useful when drivers are loaded as modules as 1386 it ensures proper ordering of tasks like device probing 1387 (suppliers first, then consumers), supplier boot state 1388 clean up (only after all consumers have probed), 1389 suspend/resume & runtime PM (consumers first, then 1390 suppliers). 1391 Format: { off | permissive | on | rpm } 1392 off -- Don't create device links from firmware info. 1393 permissive -- Create device links from firmware info 1394 but use it only for ordering boot state clean 1395 up (sync_state() calls). 1396 on -- Create device links from firmware info and use it 1397 to enforce probe and suspend/resume ordering. 1398 rpm -- Like "on", but also use to order runtime PM. 1399 1400 gamecon.map[2|3]= 1401 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1402 support via parallel port (up to 5 devices per port) 1403 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1404 See also Documentation/input/devices/joystick-parport.rst 1405 1406 gamma= [HW,DRM] 1407 1408 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART 1409 Format: off | on 1410 default: on 1411 1412 gcov_persist= [GCOV] When non-zero (default), profiling data for 1413 kernel modules is saved and remains accessible via 1414 debugfs, even when the module is unloaded/reloaded. 1415 When zero, profiling data is discarded and associated 1416 debugfs files are removed at module unload time. 1417 1418 goldfish [X86] Enable the goldfish android emulator platform. 1419 Don't use this when you are not running on the 1420 android emulator 1421 1422 gpt [EFI] Forces disk with valid GPT signature but 1423 invalid Protective MBR to be treated as GPT. If the 1424 primary GPT is corrupted, it enables the backup/alternate 1425 GPT to be used instead. 1426 1427 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1428 the "Enable 0" bit of the configuration register. 1429 Format: 0 | 1 1430 Default: 0 1431 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1432 the "Enable 0" bit of the configuration register. 1433 Format: 0 | 1 1434 Default: 0 1435 grcan.select= [HW] Select which physical interface to use. 1436 Format: 0 | 1 1437 Default: 0 1438 grcan.txsize= [HW] Sets the size of the tx buffer. 1439 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1440 Default: 1024 1441 grcan.rxsize= [HW] Sets the size of the rx buffer. 1442 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1443 Default: 1024 1444 1445 gpio-mockup.gpio_mockup_ranges 1446 [HW] Sets the ranges of gpiochip of for this device. 1447 Format: <start1>,<end1>,<start2>,<end2>... 1448 1449 hardlockup_all_cpu_backtrace= 1450 [KNL] Should the hard-lockup detector generate 1451 backtraces on all cpus. 1452 Format: 0 | 1 1453 1454 hashdist= [KNL,NUMA] Large hashes allocated during boot 1455 are distributed across NUMA nodes. Defaults on 1456 for 64-bit NUMA, off otherwise. 1457 Format: 0 | 1 (for off | on) 1458 1459 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1460 1461 hd= [EIDE] (E)IDE hard drive subsystem geometry 1462 Format: <cyl>,<head>,<sect> 1463 1464 hest_disable [ACPI] 1465 Disable Hardware Error Source Table (HEST) support; 1466 corresponding firmware-first mode error processing 1467 logic will be disabled. 1468 1469 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1470 size of <nn>. This works even on boxes that have no 1471 highmem otherwise. This also works to reduce highmem 1472 size on bigger boxes. 1473 1474 highres= [KNL] Enable/disable high resolution timer mode. 1475 Valid parameters: "on", "off" 1476 Default: "on" 1477 1478 hlt [BUGS=ARM,SH] 1479 1480 hpet= [X86-32,HPET] option to control HPET usage 1481 Format: { enable (default) | disable | force | 1482 verbose } 1483 disable: disable HPET and use PIT instead 1484 force: allow force enabled of undocumented chips (ICH4, 1485 VIA, nVidia) 1486 verbose: show contents of HPET registers during setup 1487 1488 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1489 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1490 1491 hugetlb_cma= [HW] The size of a cma area used for allocation 1492 of gigantic hugepages. 1493 Format: nn[KMGTPE] 1494 1495 Reserve a cma area of given size and allocate gigantic 1496 hugepages using the cma allocator. If enabled, the 1497 boot-time allocation of gigantic hugepages is skipped. 1498 1499 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1500 If this follows hugepagesz (below), it specifies 1501 the number of pages of hugepagesz to be allocated. 1502 If this is the first HugeTLB parameter on the command 1503 line, it specifies the number of pages to allocate for 1504 the default huge page size. See also 1505 Documentation/admin-guide/mm/hugetlbpage.rst. 1506 Format: <integer> 1507 1508 hugepagesz= 1509 [HW] The size of the HugeTLB pages. This is used in 1510 conjunction with hugepages (above) to allocate huge 1511 pages of a specific size at boot. The pair 1512 hugepagesz=X hugepages=Y can be specified once for 1513 each supported huge page size. Huge page sizes are 1514 architecture dependent. See also 1515 Documentation/admin-guide/mm/hugetlbpage.rst. 1516 Format: size[KMG] 1517 1518 hung_task_panic= 1519 [KNL] Should the hung task detector generate panics. 1520 Format: 0 | 1 1521 1522 A value of 1 instructs the kernel to panic when a 1523 hung task is detected. The default value is controlled 1524 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1525 option. The value selected by this boot parameter can 1526 be changed later by the kernel.hung_task_panic sysctl. 1527 1528 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1529 terminal devices. Valid values: 0..8 1530 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1531 If specified, z/VM IUCV HVC accepts connections 1532 from listed z/VM user IDs only. 1533 1534 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1535 which allow the hypervisor to 'idle' the 1536 guest on lock contention. 1537 1538 keep_bootcon [KNL] 1539 Do not unregister boot console at start. This is only 1540 useful for debugging when something happens in the window 1541 between unregistering the boot console and initializing 1542 the real console. 1543 1544 i2c_bus= [HW] Override the default board specific I2C bus speed 1545 or register an additional I2C bus that is not 1546 registered from board initialization code. 1547 Format: 1548 <bus_id>,<clkrate> 1549 1550 i8042.debug [HW] Toggle i8042 debug mode 1551 i8042.unmask_kbd_data 1552 [HW] Enable printing of interrupt data from the KBD port 1553 (disabled by default, and as a pre-condition 1554 requires that i8042.debug=1 be enabled) 1555 i8042.direct [HW] Put keyboard port into non-translated mode 1556 i8042.dumbkbd [HW] Pretend that controller can only read data from 1557 keyboard and cannot control its state 1558 (Don't attempt to blink the leds) 1559 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1560 i8042.nokbd [HW] Don't check/create keyboard port 1561 i8042.noloop [HW] Disable the AUX Loopback command while probing 1562 for the AUX port 1563 i8042.nomux [HW] Don't check presence of an active multiplexing 1564 controller 1565 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1566 controllers 1567 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1568 i8042.reset [HW] Reset the controller during init, cleanup and 1569 suspend-to-ram transitions, only during s2r 1570 transitions, or never reset 1571 Format: { 1 | Y | y | 0 | N | n } 1572 1, Y, y: always reset controller 1573 0, N, n: don't ever reset controller 1574 Default: only on s2r transitions on x86; most other 1575 architectures force reset to be always executed 1576 i8042.unlock [HW] Unlock (ignore) the keylock 1577 i8042.kbdreset [HW] Reset device connected to KBD port 1578 1579 i810= [HW,DRM] 1580 1581 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1582 indicates that the driver is running on unsupported 1583 hardware. 1584 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1585 does not match list of supported models. 1586 i8k.power_status 1587 [HW] Report power status in /proc/i8k 1588 (disabled by default) 1589 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1590 capability is set. 1591 1592 i915.invert_brightness= 1593 [DRM] Invert the sense of the variable that is used to 1594 set the brightness of the panel backlight. Normally a 1595 brightness value of 0 indicates backlight switched off, 1596 and the maximum of the brightness value sets the backlight 1597 to maximum brightness. If this parameter is set to 0 1598 (default) and the machine requires it, or this parameter 1599 is set to 1, a brightness value of 0 sets the backlight 1600 to maximum brightness, and the maximum of the brightness 1601 value switches the backlight off. 1602 -1 -- never invert brightness 1603 0 -- machine default 1604 1 -- force brightness inversion 1605 1606 icn= [HW,ISDN] 1607 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1608 1609 ide-core.nodma= [HW] (E)IDE subsystem 1610 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1611 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1612 .cdrom .chs .ignore_cable are additional options 1613 See Documentation/ide/ide.rst. 1614 1615 ide-generic.probe-mask= [HW] (E)IDE subsystem 1616 Format: <int> 1617 Probe mask for legacy ISA IDE ports. Depending on 1618 platform up to 6 ports are supported, enabled by 1619 setting corresponding bits in the mask to 1. The 1620 default value is 0x0, which has a special meaning. 1621 On systems that have PCI, it triggers scanning the 1622 PCI bus for the first and the second port, which 1623 are then probed. On systems without PCI the value 1624 of 0x0 enables probing the two first ports as if it 1625 was 0x3. 1626 1627 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1628 Claim all unknown PCI IDE storage controllers. 1629 1630 idle= [X86] 1631 Format: idle=poll, idle=halt, idle=nomwait 1632 Poll forces a polling idle loop that can slightly 1633 improve the performance of waking up a idle CPU, but 1634 will use a lot of power and make the system run hot. 1635 Not recommended. 1636 idle=halt: Halt is forced to be used for CPU idle. 1637 In such case C2/C3 won't be used again. 1638 idle=nomwait: Disable mwait for CPU C-states 1639 1640 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1641 Format: { strict | legacy | 2008 | relaxed } 1642 Default: strict 1643 1644 Choose which programs will be accepted for execution 1645 based on the IEEE 754 NaN encoding(s) supported by 1646 the FPU and the NaN encoding requested with the value 1647 of an ELF file header flag individually set by each 1648 binary. Hardware implementations are permitted to 1649 support either or both of the legacy and the 2008 NaN 1650 encoding mode. 1651 1652 Available settings are as follows: 1653 strict accept binaries that request a NaN encoding 1654 supported by the FPU 1655 legacy only accept legacy-NaN binaries, if supported 1656 by the FPU 1657 2008 only accept 2008-NaN binaries, if supported 1658 by the FPU 1659 relaxed accept any binaries regardless of whether 1660 supported by the FPU 1661 1662 The FPU emulator is always able to support both NaN 1663 encodings, so if no FPU hardware is present or it has 1664 been disabled with 'nofpu', then the settings of 1665 'legacy' and '2008' strap the emulator accordingly, 1666 'relaxed' straps the emulator for both legacy-NaN and 1667 2008-NaN, whereas 'strict' enables legacy-NaN only on 1668 legacy processors and both NaN encodings on MIPS32 or 1669 MIPS64 CPUs. 1670 1671 The setting for ABS.fmt/NEG.fmt instruction execution 1672 mode generally follows that for the NaN encoding, 1673 except where unsupported by hardware. 1674 1675 ignore_loglevel [KNL] 1676 Ignore loglevel setting - this will print /all/ 1677 kernel messages to the console. Useful for debugging. 1678 We also add it as printk module parameter, so users 1679 could change it dynamically, usually by 1680 /sys/module/printk/parameters/ignore_loglevel. 1681 1682 ignore_rlimit_data 1683 Ignore RLIMIT_DATA setting for data mappings, 1684 print warning at first misuse. Can be changed via 1685 /sys/module/kernel/parameters/ignore_rlimit_data. 1686 1687 ihash_entries= [KNL] 1688 Set number of hash buckets for inode cache. 1689 1690 ima_appraise= [IMA] appraise integrity measurements 1691 Format: { "off" | "enforce" | "fix" | "log" } 1692 default: "enforce" 1693 1694 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1695 The builtin appraise policy appraises all files 1696 owned by uid=0. 1697 1698 ima_canonical_fmt [IMA] 1699 Use the canonical format for the binary runtime 1700 measurements, instead of host native format. 1701 1702 ima_hash= [IMA] 1703 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1704 | sha512 | ... } 1705 default: "sha1" 1706 1707 The list of supported hash algorithms is defined 1708 in crypto/hash_info.h. 1709 1710 ima_policy= [IMA] 1711 The builtin policies to load during IMA setup. 1712 Format: "tcb | appraise_tcb | secure_boot | 1713 fail_securely" 1714 1715 The "tcb" policy measures all programs exec'd, files 1716 mmap'd for exec, and all files opened with the read 1717 mode bit set by either the effective uid (euid=0) or 1718 uid=0. 1719 1720 The "appraise_tcb" policy appraises the integrity of 1721 all files owned by root. 1722 1723 The "secure_boot" policy appraises the integrity 1724 of files (eg. kexec kernel image, kernel modules, 1725 firmware, policy, etc) based on file signatures. 1726 1727 The "fail_securely" policy forces file signature 1728 verification failure also on privileged mounted 1729 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1730 flag. 1731 1732 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1733 Load a policy which meets the needs of the Trusted 1734 Computing Base. This means IMA will measure all 1735 programs exec'd, files mmap'd for exec, and all files 1736 opened for read by uid=0. 1737 1738 ima_template= [IMA] 1739 Select one of defined IMA measurements template formats. 1740 Formats: { "ima" | "ima-ng" | "ima-sig" } 1741 Default: "ima-ng" 1742 1743 ima_template_fmt= 1744 [IMA] Define a custom template format. 1745 Format: { "field1|...|fieldN" } 1746 1747 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1748 Format: <min_file_size> 1749 Set the minimal file size for using asynchronous hash. 1750 If left unspecified, ahash usage is disabled. 1751 1752 ahash performance varies for different data sizes on 1753 different crypto accelerators. This option can be used 1754 to achieve the best performance for a particular HW. 1755 1756 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1757 Format: <bufsize> 1758 Set hashing buffer size. Default: 4k. 1759 1760 ahash performance varies for different chunk sizes on 1761 different crypto accelerators. This option can be used 1762 to achieve best performance for particular HW. 1763 1764 init= [KNL] 1765 Format: <full_path> 1766 Run specified binary instead of /sbin/init as init 1767 process. 1768 1769 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1770 for working out where the kernel is dying during 1771 startup. 1772 1773 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1774 initcall functions. Useful for debugging built-in 1775 modules and initcalls. 1776 1777 initrd= [BOOT] Specify the location of the initial ramdisk 1778 1779 initrdmem= [KNL] Specify a physical address and size from which to 1780 load the initrd. If an initrd is compiled in or 1781 specified in the bootparams, it takes priority over this 1782 setting. 1783 Format: ss[KMG],nn[KMG] 1784 Default is 0, 0 1785 1786 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1787 zeroes. 1788 Format: 0 | 1 1789 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1790 1791 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1792 Format: 0 | 1 1793 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1794 1795 init_pkru= [x86] Specify the default memory protection keys rights 1796 register contents for all processes. 0x55555554 by 1797 default (disallow access to all but pkey 0). Can 1798 override in debugfs after boot. 1799 1800 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1801 Format: <irq> 1802 1803 int_pln_enable [x86] Enable power limit notification interrupt 1804 1805 integrity_audit=[IMA] 1806 Format: { "0" | "1" } 1807 0 -- basic integrity auditing messages. (Default) 1808 1 -- additional integrity auditing messages. 1809 1810 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1811 on 1812 Enable intel iommu driver. 1813 off 1814 Disable intel iommu driver. 1815 igfx_off [Default Off] 1816 By default, gfx is mapped as normal device. If a gfx 1817 device has a dedicated DMAR unit, the DMAR unit is 1818 bypassed by not enabling DMAR with this option. In 1819 this case, gfx device will use physical address for 1820 DMA. 1821 forcedac [x86_64] 1822 With this option iommu will not optimize to look 1823 for io virtual address below 32-bit forcing dual 1824 address cycle on pci bus for cards supporting greater 1825 than 32-bit addressing. The default is to look 1826 for translation below 32-bit and if not available 1827 then look in the higher range. 1828 strict [Default Off] 1829 With this option on every unmap_single operation will 1830 result in a hardware IOTLB flush operation as opposed 1831 to batching them for performance. 1832 sp_off [Default Off] 1833 By default, super page will be supported if Intel IOMMU 1834 has the capability. With this option, super page will 1835 not be supported. 1836 sm_on [Default Off] 1837 By default, scalable mode will be disabled even if the 1838 hardware advertises that it has support for the scalable 1839 mode translation. With this option set, scalable mode 1840 will be used on hardware which claims to support it. 1841 tboot_noforce [Default Off] 1842 Do not force the Intel IOMMU enabled under tboot. 1843 By default, tboot will force Intel IOMMU on, which 1844 could harm performance of some high-throughput 1845 devices like 40GBit network cards, even if identity 1846 mapping is enabled. 1847 Note that using this option lowers the security 1848 provided by tboot because it makes the system 1849 vulnerable to DMA attacks. 1850 nobounce [Default off] 1851 Disable bounce buffer for untrusted devices such as 1852 the Thunderbolt devices. This will treat the untrusted 1853 devices as the trusted ones, hence might expose security 1854 risks of DMA attacks. 1855 1856 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1857 0 disables intel_idle and fall back on acpi_idle. 1858 1 to 9 specify maximum depth of C-state. 1859 1860 intel_pstate= [X86] 1861 disable 1862 Do not enable intel_pstate as the default 1863 scaling driver for the supported processors 1864 passive 1865 Use intel_pstate as a scaling driver, but configure it 1866 to work with generic cpufreq governors (instead of 1867 enabling its internal governor). This mode cannot be 1868 used along with the hardware-managed P-states (HWP) 1869 feature. 1870 force 1871 Enable intel_pstate on systems that prohibit it by default 1872 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1873 instead of acpi-cpufreq may disable platform features, such 1874 as thermal controls and power capping, that rely on ACPI 1875 P-States information being indicated to OSPM and therefore 1876 should be used with caution. This option does not work with 1877 processors that aren't supported by the intel_pstate driver 1878 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1879 no_hwp 1880 Do not enable hardware P state control (HWP) 1881 if available. 1882 hwp_only 1883 Only load intel_pstate on systems which support 1884 hardware P state control (HWP) if available. 1885 support_acpi_ppc 1886 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1887 Description Table, specifies preferred power management 1888 profile as "Enterprise Server" or "Performance Server", 1889 then this feature is turned on by default. 1890 per_cpu_perf_limits 1891 Allow per-logical-CPU P-State performance control limits using 1892 cpufreq sysfs interface 1893 1894 intremap= [X86-64, Intel-IOMMU] 1895 on enable Interrupt Remapping (default) 1896 off disable Interrupt Remapping 1897 nosid disable Source ID checking 1898 no_x2apic_optout 1899 BIOS x2APIC opt-out request will be ignored 1900 nopost disable Interrupt Posting 1901 1902 iomem= Disable strict checking of access to MMIO memory 1903 strict regions from userspace. 1904 relaxed 1905 1906 iommu= [x86] 1907 off 1908 force 1909 noforce 1910 biomerge 1911 panic 1912 nopanic 1913 merge 1914 nomerge 1915 soft 1916 pt [x86] 1917 nopt [x86] 1918 nobypass [PPC/POWERNV] 1919 Disable IOMMU bypass, using IOMMU for PCI devices. 1920 1921 iommu.strict= [ARM64] Configure TLB invalidation behaviour 1922 Format: { "0" | "1" } 1923 0 - Lazy mode. 1924 Request that DMA unmap operations use deferred 1925 invalidation of hardware TLBs, for increased 1926 throughput at the cost of reduced device isolation. 1927 Will fall back to strict mode if not supported by 1928 the relevant IOMMU driver. 1929 1 - Strict mode (default). 1930 DMA unmap operations invalidate IOMMU hardware TLBs 1931 synchronously. 1932 1933 iommu.passthrough= 1934 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 1935 Format: { "0" | "1" } 1936 0 - Use IOMMU translation for DMA. 1937 1 - Bypass the IOMMU for DMA. 1938 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 1939 1940 io7= [HW] IO7 for Marvel based alpha systems 1941 See comment before marvel_specify_io7 in 1942 arch/alpha/kernel/core_marvel.c. 1943 1944 io_delay= [X86] I/O delay method 1945 0x80 1946 Standard port 0x80 based delay 1947 0xed 1948 Alternate port 0xed based delay (needed on some systems) 1949 udelay 1950 Simple two microseconds delay 1951 none 1952 No delay 1953 1954 ip= [IP_PNP] 1955 See Documentation/admin-guide/nfs/nfsroot.rst. 1956 1957 ipcmni_extend [KNL] Extend the maximum number of unique System V 1958 IPC identifiers from 32,768 to 16,777,216. 1959 1960 irqaffinity= [SMP] Set the default irq affinity mask 1961 The argument is a cpu list, as described above. 1962 1963 irqchip.gicv2_force_probe= 1964 [ARM, ARM64] 1965 Format: <bool> 1966 Force the kernel to look for the second 4kB page 1967 of a GICv2 controller even if the memory range 1968 exposed by the device tree is too small. 1969 1970 irqchip.gicv3_nolpi= 1971 [ARM, ARM64] 1972 Force the kernel to ignore the availability of 1973 LPIs (and by consequence ITSs). Intended for system 1974 that use the kernel as a bootloader, and thus want 1975 to let secondary kernels in charge of setting up 1976 LPIs. 1977 1978 irqchip.gicv3_pseudo_nmi= [ARM64] 1979 Enables support for pseudo-NMIs in the kernel. This 1980 requires the kernel to be built with 1981 CONFIG_ARM64_PSEUDO_NMI. 1982 1983 irqfixup [HW] 1984 When an interrupt is not handled search all handlers 1985 for it. Intended to get systems with badly broken 1986 firmware running. 1987 1988 irqpoll [HW] 1989 When an interrupt is not handled search all handlers 1990 for it. Also check all handlers each timer 1991 interrupt. Intended to get systems with badly broken 1992 firmware running. 1993 1994 isapnp= [ISAPNP] 1995 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 1996 1997 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 1998 [Deprecated - use cpusets instead] 1999 Format: [flag-list,]<cpu-list> 2000 2001 Specify one or more CPUs to isolate from disturbances 2002 specified in the flag list (default: domain): 2003 2004 nohz 2005 Disable the tick when a single task runs. 2006 2007 A residual 1Hz tick is offloaded to workqueues, which you 2008 need to affine to housekeeping through the global 2009 workqueue's affinity configured via the 2010 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2011 by using the 'domain' flag described below. 2012 2013 NOTE: by default the global workqueue runs on all CPUs, 2014 so to protect individual CPUs the 'cpumask' file has to 2015 be configured manually after bootup. 2016 2017 domain 2018 Isolate from the general SMP balancing and scheduling 2019 algorithms. Note that performing domain isolation this way 2020 is irreversible: it's not possible to bring back a CPU to 2021 the domains once isolated through isolcpus. It's strongly 2022 advised to use cpusets instead to disable scheduler load 2023 balancing through the "cpuset.sched_load_balance" file. 2024 It offers a much more flexible interface where CPUs can 2025 move in and out of an isolated set anytime. 2026 2027 You can move a process onto or off an "isolated" CPU via 2028 the CPU affinity syscalls or cpuset. 2029 <cpu number> begins at 0 and the maximum value is 2030 "number of CPUs in system - 1". 2031 2032 managed_irq 2033 2034 Isolate from being targeted by managed interrupts 2035 which have an interrupt mask containing isolated 2036 CPUs. The affinity of managed interrupts is 2037 handled by the kernel and cannot be changed via 2038 the /proc/irq/* interfaces. 2039 2040 This isolation is best effort and only effective 2041 if the automatically assigned interrupt mask of a 2042 device queue contains isolated and housekeeping 2043 CPUs. If housekeeping CPUs are online then such 2044 interrupts are directed to the housekeeping CPU 2045 so that IO submitted on the housekeeping CPU 2046 cannot disturb the isolated CPU. 2047 2048 If a queue's affinity mask contains only isolated 2049 CPUs then this parameter has no effect on the 2050 interrupt routing decision, though interrupts are 2051 only delivered when tasks running on those 2052 isolated CPUs submit IO. IO submitted on 2053 housekeeping CPUs has no influence on those 2054 queues. 2055 2056 The format of <cpu-list> is described above. 2057 2058 iucv= [HW,NET] 2059 2060 ivrs_ioapic [HW,X86_64] 2061 Provide an override to the IOAPIC-ID<->DEVICE-ID 2062 mapping provided in the IVRS ACPI table. For 2063 example, to map IOAPIC-ID decimal 10 to 2064 PCI device 00:14.0 write the parameter as: 2065 ivrs_ioapic[10]=00:14.0 2066 2067 ivrs_hpet [HW,X86_64] 2068 Provide an override to the HPET-ID<->DEVICE-ID 2069 mapping provided in the IVRS ACPI table. For 2070 example, to map HPET-ID decimal 0 to 2071 PCI device 00:14.0 write the parameter as: 2072 ivrs_hpet[0]=00:14.0 2073 2074 ivrs_acpihid [HW,X86_64] 2075 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2076 mapping provided in the IVRS ACPI table. For 2077 example, to map UART-HID:UID AMD0020:0 to 2078 PCI device 00:14.5 write the parameter as: 2079 ivrs_acpihid[00:14.5]=AMD0020:0 2080 2081 js= [HW,JOY] Analog joystick 2082 See Documentation/input/joydev/joystick.rst. 2083 2084 nokaslr [KNL] 2085 When CONFIG_RANDOMIZE_BASE is set, this disables 2086 kernel and module base offset ASLR (Address Space 2087 Layout Randomization). 2088 2089 kasan_multi_shot 2090 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2091 report on every invalid memory access. Without this 2092 parameter KASAN will print report only for the first 2093 invalid access. 2094 2095 keepinitrd [HW,ARM] 2096 2097 kernelcore= [KNL,X86,IA-64,PPC] 2098 Format: nn[KMGTPE] | nn% | "mirror" 2099 This parameter specifies the amount of memory usable by 2100 the kernel for non-movable allocations. The requested 2101 amount is spread evenly throughout all nodes in the 2102 system as ZONE_NORMAL. The remaining memory is used for 2103 movable memory in its own zone, ZONE_MOVABLE. In the 2104 event, a node is too small to have both ZONE_NORMAL and 2105 ZONE_MOVABLE, kernelcore memory will take priority and 2106 other nodes will have a larger ZONE_MOVABLE. 2107 2108 ZONE_MOVABLE is used for the allocation of pages that 2109 may be reclaimed or moved by the page migration 2110 subsystem. Note that allocations like PTEs-from-HighMem 2111 still use the HighMem zone if it exists, and the Normal 2112 zone if it does not. 2113 2114 It is possible to specify the exact amount of memory in 2115 the form of "nn[KMGTPE]", a percentage of total system 2116 memory in the form of "nn%", or "mirror". If "mirror" 2117 option is specified, mirrored (reliable) memory is used 2118 for non-movable allocations and remaining memory is used 2119 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2120 are exclusive, so you cannot specify multiple forms. 2121 2122 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2123 Format: <Controller#>[,poll interval] 2124 The controller # is the number of the ehci usb debug 2125 port as it is probed via PCI. The poll interval is 2126 optional and is the number seconds in between 2127 each poll cycle to the debug port in case you need 2128 the functionality for interrupting the kernel with 2129 gdb or control-c on the dbgp connection. When 2130 not using this parameter you use sysrq-g to break into 2131 the kernel debugger. 2132 2133 kgdboc= [KGDB,HW] kgdb over consoles. 2134 Requires a tty driver that supports console polling, 2135 or a supported polling keyboard driver (non-usb). 2136 Serial only format: <serial_device>[,baud] 2137 keyboard only format: kbd 2138 keyboard and serial format: kbd,<serial_device>[,baud] 2139 Optional Kernel mode setting: 2140 kms, kbd format: kms,kbd 2141 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2142 2143 kgdboc_earlycon= [KGDB,HW] 2144 If the boot console provides the ability to read 2145 characters and can work in polling mode, you can use 2146 this parameter to tell kgdb to use it as a backend 2147 until the normal console is registered. Intended to 2148 be used together with the kgdboc parameter which 2149 specifies the normal console to transition to. 2150 2151 The name of the early console should be specified 2152 as the value of this parameter. Note that the name of 2153 the early console might be different than the tty 2154 name passed to kgdboc. It's OK to leave the value 2155 blank and the first boot console that implements 2156 read() will be picked. 2157 2158 kgdbwait [KGDB] Stop kernel execution and enter the 2159 kernel debugger at the earliest opportunity. 2160 2161 kmac= [MIPS] korina ethernet MAC address. 2162 Configure the RouterBoard 532 series on-chip 2163 Ethernet adapter MAC address. 2164 2165 kmemleak= [KNL] Boot-time kmemleak enable/disable 2166 Valid arguments: on, off 2167 Default: on 2168 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2169 the default is off. 2170 2171 kprobe_event=[probe-list] 2172 [FTRACE] Add kprobe events and enable at boot time. 2173 The probe-list is a semicolon delimited list of probe 2174 definitions. Each definition is same as kprobe_events 2175 interface, but the parameters are comma delimited. 2176 For example, to add a kprobe event on vfs_read with 2177 arg1 and arg2, add to the command line; 2178 2179 kprobe_event=p,vfs_read,$arg1,$arg2 2180 2181 See also Documentation/trace/kprobetrace.rst "Kernel 2182 Boot Parameter" section. 2183 2184 kpti= [ARM64] Control page table isolation of user 2185 and kernel address spaces. 2186 Default: enabled on cores which need mitigation. 2187 0: force disabled 2188 1: force enabled 2189 2190 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2191 Default is 0 (don't ignore, but inject #GP) 2192 2193 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2194 Default is false (don't support). 2195 2196 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2197 KVM MMU at runtime. 2198 Default is 0 (off) 2199 2200 kvm.nx_huge_pages= 2201 [KVM] Controls the software workaround for the 2202 X86_BUG_ITLB_MULTIHIT bug. 2203 force : Always deploy workaround. 2204 off : Never deploy workaround. 2205 auto : Deploy workaround based on the presence of 2206 X86_BUG_ITLB_MULTIHIT. 2207 2208 Default is 'auto'. 2209 2210 If the software workaround is enabled for the host, 2211 guests do need not to enable it for nested guests. 2212 2213 kvm.nx_huge_pages_recovery_ratio= 2214 [KVM] Controls how many 4KiB pages are periodically zapped 2215 back to huge pages. 0 disables the recovery, otherwise if 2216 the value is N KVM will zap 1/Nth of the 4KiB pages every 2217 minute. The default is 60. 2218 2219 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2220 Default is 1 (enabled) 2221 2222 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2223 for all guests. 2224 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2225 2226 kvm-arm.vgic_v3_group0_trap= 2227 [KVM,ARM] Trap guest accesses to GICv3 group-0 2228 system registers 2229 2230 kvm-arm.vgic_v3_group1_trap= 2231 [KVM,ARM] Trap guest accesses to GICv3 group-1 2232 system registers 2233 2234 kvm-arm.vgic_v3_common_trap= 2235 [KVM,ARM] Trap guest accesses to GICv3 common 2236 system registers 2237 2238 kvm-arm.vgic_v4_enable= 2239 [KVM,ARM] Allow use of GICv4 for direct injection of 2240 LPIs. 2241 2242 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2243 (virtualized MMU) support on capable Intel chips. 2244 Default is 1 (enabled) 2245 2246 kvm-intel.emulate_invalid_guest_state= 2247 [KVM,Intel] Enable emulation of invalid guest states 2248 Default is 0 (disabled) 2249 2250 kvm-intel.flexpriority= 2251 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2252 Default is 1 (enabled) 2253 2254 kvm-intel.nested= 2255 [KVM,Intel] Enable VMX nesting (nVMX). 2256 Default is 0 (disabled) 2257 2258 kvm-intel.unrestricted_guest= 2259 [KVM,Intel] Disable unrestricted guest feature 2260 (virtualized real and unpaged mode) on capable 2261 Intel chips. Default is 1 (enabled) 2262 2263 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2264 CVE-2018-3620. 2265 2266 Valid arguments: never, cond, always 2267 2268 always: L1D cache flush on every VMENTER. 2269 cond: Flush L1D on VMENTER only when the code between 2270 VMEXIT and VMENTER can leak host memory. 2271 never: Disables the mitigation 2272 2273 Default is cond (do L1 cache flush in specific instances) 2274 2275 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2276 feature (tagged TLBs) on capable Intel chips. 2277 Default is 1 (enabled) 2278 2279 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2280 affected CPUs 2281 2282 The kernel PTE inversion protection is unconditionally 2283 enabled and cannot be disabled. 2284 2285 full 2286 Provides all available mitigations for the 2287 L1TF vulnerability. Disables SMT and 2288 enables all mitigations in the 2289 hypervisors, i.e. unconditional L1D flush. 2290 2291 SMT control and L1D flush control via the 2292 sysfs interface is still possible after 2293 boot. Hypervisors will issue a warning 2294 when the first VM is started in a 2295 potentially insecure configuration, 2296 i.e. SMT enabled or L1D flush disabled. 2297 2298 full,force 2299 Same as 'full', but disables SMT and L1D 2300 flush runtime control. Implies the 2301 'nosmt=force' command line option. 2302 (i.e. sysfs control of SMT is disabled.) 2303 2304 flush 2305 Leaves SMT enabled and enables the default 2306 hypervisor mitigation, i.e. conditional 2307 L1D flush. 2308 2309 SMT control and L1D flush control via the 2310 sysfs interface is still possible after 2311 boot. Hypervisors will issue a warning 2312 when the first VM is started in a 2313 potentially insecure configuration, 2314 i.e. SMT enabled or L1D flush disabled. 2315 2316 flush,nosmt 2317 2318 Disables SMT and enables the default 2319 hypervisor mitigation. 2320 2321 SMT control and L1D flush control via the 2322 sysfs interface is still possible after 2323 boot. Hypervisors will issue a warning 2324 when the first VM is started in a 2325 potentially insecure configuration, 2326 i.e. SMT enabled or L1D flush disabled. 2327 2328 flush,nowarn 2329 Same as 'flush', but hypervisors will not 2330 warn when a VM is started in a potentially 2331 insecure configuration. 2332 2333 off 2334 Disables hypervisor mitigations and doesn't 2335 emit any warnings. 2336 It also drops the swap size and available 2337 RAM limit restriction on both hypervisor and 2338 bare metal. 2339 2340 Default is 'flush'. 2341 2342 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2343 2344 l2cr= [PPC] 2345 2346 l3cr= [PPC] 2347 2348 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2349 disabled it. 2350 2351 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline 2352 value for LAPIC timer one-shot implementation. Default 2353 back to the programmable timer unit in the LAPIC. 2354 2355 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2356 in C2 power state. 2357 2358 libata.dma= [LIBATA] DMA control 2359 libata.dma=0 Disable all PATA and SATA DMA 2360 libata.dma=1 PATA and SATA Disk DMA only 2361 libata.dma=2 ATAPI (CDROM) DMA only 2362 libata.dma=4 Compact Flash DMA only 2363 Combinations also work, so libata.dma=3 enables DMA 2364 for disks and CDROMs, but not CFs. 2365 2366 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2367 libata.ignore_hpa=0 keep BIOS limits (default) 2368 libata.ignore_hpa=1 ignore limits, using full disk 2369 2370 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2371 when set. 2372 Format: <int> 2373 2374 libata.force= [LIBATA] Force configurations. The format is comma 2375 separated list of "[ID:]VAL" where ID is 2376 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2377 matching port, link or device. Basically, it matches 2378 the ATA ID string printed on console by libata. If 2379 the whole ID part is omitted, the last PORT and DEVICE 2380 values are used. If ID hasn't been specified yet, the 2381 configuration applies to all ports, links and devices. 2382 2383 If only DEVICE is omitted, the parameter applies to 2384 the port and all links and devices behind it. DEVICE 2385 number of 0 either selects the first device or the 2386 first fan-out link behind PMP device. It does not 2387 select the host link. DEVICE number of 15 selects the 2388 host link and device attached to it. 2389 2390 The VAL specifies the configuration to force. As long 2391 as there's no ambiguity shortcut notation is allowed. 2392 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2393 The following configurations can be forced. 2394 2395 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2396 Any ID with matching PORT is used. 2397 2398 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2399 2400 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2401 udma[/][16,25,33,44,66,100,133] notation is also 2402 allowed. 2403 2404 * [no]ncq: Turn on or off NCQ. 2405 2406 * [no]ncqtrim: Turn off queued DSM TRIM. 2407 2408 * nohrst, nosrst, norst: suppress hard, soft 2409 and both resets. 2410 2411 * rstonce: only attempt one reset during 2412 hot-unplug link recovery 2413 2414 * dump_id: dump IDENTIFY data. 2415 2416 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2417 2418 * disable: Disable this device. 2419 2420 If there are multiple matching configurations changing 2421 the same attribute, the last one is used. 2422 2423 memblock=debug [KNL] Enable memblock debug messages. 2424 2425 load_ramdisk= [RAM] List of ramdisks to load from floppy 2426 See Documentation/admin-guide/blockdev/ramdisk.rst. 2427 2428 lockd.nlm_grace_period=P [NFS] Assign grace period. 2429 Format: <integer> 2430 2431 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2432 Format: <integer> 2433 2434 lockd.nlm_timeout=T [NFS] Assign timeout value. 2435 Format: <integer> 2436 2437 lockd.nlm_udpport=M [NFS] Assign UDP port. 2438 Format: <integer> 2439 2440 lockdown= [SECURITY] 2441 { integrity | confidentiality } 2442 Enable the kernel lockdown feature. If set to 2443 integrity, kernel features that allow userland to 2444 modify the running kernel are disabled. If set to 2445 confidentiality, kernel features that allow userland 2446 to extract confidential information from the kernel 2447 are also disabled. 2448 2449 locktorture.nreaders_stress= [KNL] 2450 Set the number of locking read-acquisition kthreads. 2451 Defaults to being automatically set based on the 2452 number of online CPUs. 2453 2454 locktorture.nwriters_stress= [KNL] 2455 Set the number of locking write-acquisition kthreads. 2456 2457 locktorture.onoff_holdoff= [KNL] 2458 Set time (s) after boot for CPU-hotplug testing. 2459 2460 locktorture.onoff_interval= [KNL] 2461 Set time (s) between CPU-hotplug operations, or 2462 zero to disable CPU-hotplug testing. 2463 2464 locktorture.shuffle_interval= [KNL] 2465 Set task-shuffle interval (jiffies). Shuffling 2466 tasks allows some CPUs to go into dyntick-idle 2467 mode during the locktorture test. 2468 2469 locktorture.shutdown_secs= [KNL] 2470 Set time (s) after boot system shutdown. This 2471 is useful for hands-off automated testing. 2472 2473 locktorture.stat_interval= [KNL] 2474 Time (s) between statistics printk()s. 2475 2476 locktorture.stutter= [KNL] 2477 Time (s) to stutter testing, for example, 2478 specifying five seconds causes the test to run for 2479 five seconds, wait for five seconds, and so on. 2480 This tests the locking primitive's ability to 2481 transition abruptly to and from idle. 2482 2483 locktorture.torture_type= [KNL] 2484 Specify the locking implementation to test. 2485 2486 locktorture.verbose= [KNL] 2487 Enable additional printk() statements. 2488 2489 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2490 Format: <irq> 2491 2492 loglevel= All Kernel Messages with a loglevel smaller than the 2493 console loglevel will be printed to the console. It can 2494 also be changed with klogd or other programs. The 2495 loglevels are defined as follows: 2496 2497 0 (KERN_EMERG) system is unusable 2498 1 (KERN_ALERT) action must be taken immediately 2499 2 (KERN_CRIT) critical conditions 2500 3 (KERN_ERR) error conditions 2501 4 (KERN_WARNING) warning conditions 2502 5 (KERN_NOTICE) normal but significant condition 2503 6 (KERN_INFO) informational 2504 7 (KERN_DEBUG) debug-level messages 2505 2506 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2507 in bytes. n must be a power of two and greater 2508 than the minimal size. The minimal size is defined 2509 by LOG_BUF_SHIFT kernel config parameter. There is 2510 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2511 that allows to increase the default size depending on 2512 the number of CPUs. See init/Kconfig for more details. 2513 2514 logo.nologo [FB] Disables display of the built-in Linux logo. 2515 This may be used to provide more screen space for 2516 kernel log messages and is useful when debugging 2517 kernel boot problems. 2518 2519 lp=0 [LP] Specify parallel ports to use, e.g, 2520 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2521 lp=reset first parallel port). 'lp=0' disables the 2522 lp=auto printer driver. 'lp=reset' (which can be 2523 specified in addition to the ports) causes 2524 attached printers to be reset. Using 2525 lp=port1,port2,... specifies the parallel ports 2526 to associate lp devices with, starting with 2527 lp0. A port specification may be 'none' to skip 2528 that lp device, or a parport name such as 2529 'parport0'. Specifying 'lp=auto' instead of a 2530 port specification list means that device IDs 2531 from each port should be examined, to see if 2532 an IEEE 1284-compliant printer is attached; if 2533 so, the driver will manage that printer. 2534 See also header of drivers/char/lp.c. 2535 2536 lpj=n [KNL] 2537 Sets loops_per_jiffy to given constant, thus avoiding 2538 time-consuming boot-time autodetection (up to 250 ms per 2539 CPU). 0 enables autodetection (default). To determine 2540 the correct value for your kernel, boot with normal 2541 autodetection and see what value is printed. Note that 2542 on SMP systems the preset will be applied to all CPUs, 2543 which is likely to cause problems if your CPUs need 2544 significantly divergent settings. An incorrect value 2545 will cause delays in the kernel to be wrong, leading to 2546 unpredictable I/O errors and other breakage. Although 2547 unlikely, in the extreme case this might damage your 2548 hardware. 2549 2550 ltpc= [NET] 2551 Format: <io>,<irq>,<dma> 2552 2553 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2554 2555 lsm=lsm1,...,lsmN 2556 [SECURITY] Choose order of LSM initialization. This 2557 overrides CONFIG_LSM, and the "security=" parameter. 2558 2559 machvec= [IA-64] Force the use of a particular machine-vector 2560 (machvec) in a generic kernel. 2561 Example: machvec=hpzx1 2562 2563 machtype= [Loongson] Share the same kernel image file between different 2564 yeeloong laptop. 2565 Example: machtype=lemote-yeeloong-2f-7inch 2566 2567 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2568 than or equal to this physical address is ignored. 2569 2570 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2571 will bring up during bootup. maxcpus=n : n >= 0 limits 2572 the kernel to bring up 'n' processors. Surely after 2573 bootup you can bring up the other plugged cpu by executing 2574 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2575 only takes effect during system bootup. 2576 While n=0 is a special case, it is equivalent to "nosmp", 2577 which also disables the IO APIC. 2578 2579 max_loop= [LOOP] The number of loop block devices that get 2580 (loop.max_loop) unconditionally pre-created at init time. The default 2581 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2582 of statically allocating a predefined number, loop 2583 devices can be requested on-demand with the 2584 /dev/loop-control interface. 2585 2586 mce [X86-32] Machine Check Exception 2587 2588 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2589 2590 md= [HW] RAID subsystems devices and level 2591 See Documentation/admin-guide/md.rst. 2592 2593 mdacon= [MDA] 2594 Format: <first>,<last> 2595 Specifies range of consoles to be captured by the MDA. 2596 2597 mds= [X86,INTEL] 2598 Control mitigation for the Micro-architectural Data 2599 Sampling (MDS) vulnerability. 2600 2601 Certain CPUs are vulnerable to an exploit against CPU 2602 internal buffers which can forward information to a 2603 disclosure gadget under certain conditions. 2604 2605 In vulnerable processors, the speculatively 2606 forwarded data can be used in a cache side channel 2607 attack, to access data to which the attacker does 2608 not have direct access. 2609 2610 This parameter controls the MDS mitigation. The 2611 options are: 2612 2613 full - Enable MDS mitigation on vulnerable CPUs 2614 full,nosmt - Enable MDS mitigation and disable 2615 SMT on vulnerable CPUs 2616 off - Unconditionally disable MDS mitigation 2617 2618 On TAA-affected machines, mds=off can be prevented by 2619 an active TAA mitigation as both vulnerabilities are 2620 mitigated with the same mechanism so in order to disable 2621 this mitigation, you need to specify tsx_async_abort=off 2622 too. 2623 2624 Not specifying this option is equivalent to 2625 mds=full. 2626 2627 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2628 2629 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2630 Amount of memory to be used in cases as follows: 2631 2632 1 for test; 2633 2 when the kernel is not able to see the whole system memory; 2634 3 memory that lies after 'mem=' boundary is excluded from 2635 the hypervisor, then assigned to KVM guests. 2636 2637 [X86] Work as limiting max address. Use together 2638 with memmap= to avoid physical address space collisions. 2639 Without memmap= PCI devices could be placed at addresses 2640 belonging to unused RAM. 2641 2642 Note that this only takes effects during boot time since 2643 in above case 3, memory may need be hot added after boot 2644 if system memory of hypervisor is not sufficient. 2645 2646 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2647 memory. 2648 2649 memchunk=nn[KMG] 2650 [KNL,SH] Allow user to override the default size for 2651 per-device physically contiguous DMA buffers. 2652 2653 memhp_default_state=online/offline 2654 [KNL] Set the initial state for the memory hotplug 2655 onlining policy. If not specified, the default value is 2656 set according to the 2657 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2658 option. 2659 See Documentation/admin-guide/mm/memory-hotplug.rst. 2660 2661 memmap=exactmap [KNL,X86] Enable setting of an exact 2662 E820 memory map, as specified by the user. 2663 Such memmap=exactmap lines can be constructed based on 2664 BIOS output or other requirements. See the memmap=nn@ss 2665 option description. 2666 2667 memmap=nn[KMG]@ss[KMG] 2668 [KNL] Force usage of a specific region of memory. 2669 Region of memory to be used is from ss to ss+nn. 2670 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2671 which limits max address to nn[KMG]. 2672 Multiple different regions can be specified, 2673 comma delimited. 2674 Example: 2675 memmap=100M@2G,100M#3G,1G!1024G 2676 2677 memmap=nn[KMG]#ss[KMG] 2678 [KNL,ACPI] Mark specific memory as ACPI data. 2679 Region of memory to be marked is from ss to ss+nn. 2680 2681 memmap=nn[KMG]$ss[KMG] 2682 [KNL,ACPI] Mark specific memory as reserved. 2683 Region of memory to be reserved is from ss to ss+nn. 2684 Example: Exclude memory from 0x18690000-0x1869ffff 2685 memmap=64K$0x18690000 2686 or 2687 memmap=0x10000$0x18690000 2688 Some bootloaders may need an escape character before '$', 2689 like Grub2, otherwise '$' and the following number 2690 will be eaten. 2691 2692 memmap=nn[KMG]!ss[KMG] 2693 [KNL,X86] Mark specific memory as protected. 2694 Region of memory to be used, from ss to ss+nn. 2695 The memory region may be marked as e820 type 12 (0xc) 2696 and is NVDIMM or ADR memory. 2697 2698 memmap=<size>%<offset>-<oldtype>+<newtype> 2699 [KNL,ACPI] Convert memory within the specified region 2700 from <oldtype> to <newtype>. If "-<oldtype>" is left 2701 out, the whole region will be marked as <newtype>, 2702 even if previously unavailable. If "+<newtype>" is left 2703 out, matching memory will be removed. Types are 2704 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2705 3 = ACPI, 12 = PRAM. 2706 2707 memory_corruption_check=0/1 [X86] 2708 Some BIOSes seem to corrupt the first 64k of 2709 memory when doing things like suspend/resume. 2710 Setting this option will scan the memory 2711 looking for corruption. Enabling this will 2712 both detect corruption and prevent the kernel 2713 from using the memory being corrupted. 2714 However, its intended as a diagnostic tool; if 2715 repeatable BIOS-originated corruption always 2716 affects the same memory, you can use memmap= 2717 to prevent the kernel from using that memory. 2718 2719 memory_corruption_check_size=size [X86] 2720 By default it checks for corruption in the low 2721 64k, making this memory unavailable for normal 2722 use. Use this parameter to scan for 2723 corruption in more or less memory. 2724 2725 memory_corruption_check_period=seconds [X86] 2726 By default it checks for corruption every 60 2727 seconds. Use this parameter to check at some 2728 other rate. 0 disables periodic checking. 2729 2730 memtest= [KNL,X86,ARM,PPC] Enable memtest 2731 Format: <integer> 2732 default : 0 <disable> 2733 Specifies the number of memtest passes to be 2734 performed. Each pass selects another test 2735 pattern from a given set of patterns. Memtest 2736 fills the memory with this pattern, validates 2737 memory contents and reserves bad memory 2738 regions that are detected. 2739 2740 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2741 Valid arguments: on, off 2742 Default (depends on kernel configuration option): 2743 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2744 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2745 mem_encrypt=on: Activate SME 2746 mem_encrypt=off: Do not activate SME 2747 2748 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2749 for details on when memory encryption can be activated. 2750 2751 mem_sleep_default= [SUSPEND] Default system suspend mode: 2752 s2idle - Suspend-To-Idle 2753 shallow - Power-On Suspend or equivalent (if supported) 2754 deep - Suspend-To-RAM or equivalent (if supported) 2755 See Documentation/admin-guide/pm/sleep-states.rst. 2756 2757 meye.*= [HW] Set MotionEye Camera parameters 2758 See Documentation/admin-guide/media/meye.rst. 2759 2760 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2761 Multi-Function General Purpose Timers on AMD Geode 2762 platforms. 2763 2764 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2765 the BIOS has incorrectly applied a workaround. TinyBIOS 2766 version 0.98 is known to be affected, 0.99 fixes the 2767 problem by letting the user disable the workaround. 2768 2769 mga= [HW,DRM] 2770 2771 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2772 physical address is ignored. 2773 2774 mini2440= [ARM,HW,KNL] 2775 Format:[0..2][b][c][t] 2776 Default: "0tb" 2777 MINI2440 configuration specification: 2778 0 - The attached screen is the 3.5" TFT 2779 1 - The attached screen is the 7" TFT 2780 2 - The VGA Shield is attached (1024x768) 2781 Leaving out the screen size parameter will not load 2782 the TFT driver, and the framebuffer will be left 2783 unconfigured. 2784 b - Enable backlight. The TFT backlight pin will be 2785 linked to the kernel VESA blanking code and a GPIO 2786 LED. This parameter is not necessary when using the 2787 VGA shield. 2788 c - Enable the s3c camera interface. 2789 t - Reserved for enabling touchscreen support. The 2790 touchscreen support is not enabled in the mainstream 2791 kernel as of 2.6.30, a preliminary port can be found 2792 in the "bleeding edge" mini2440 support kernel at 2793 http://repo.or.cz/w/linux-2.6/mini2440.git 2794 2795 mitigations= 2796 [X86,PPC,S390,ARM64] Control optional mitigations for 2797 CPU vulnerabilities. This is a set of curated, 2798 arch-independent options, each of which is an 2799 aggregation of existing arch-specific options. 2800 2801 off 2802 Disable all optional CPU mitigations. This 2803 improves system performance, but it may also 2804 expose users to several CPU vulnerabilities. 2805 Equivalent to: nopti [X86,PPC] 2806 kpti=0 [ARM64] 2807 nospectre_v1 [X86,PPC] 2808 nobp=0 [S390] 2809 nospectre_v2 [X86,PPC,S390,ARM64] 2810 spectre_v2_user=off [X86] 2811 spec_store_bypass_disable=off [X86,PPC] 2812 ssbd=force-off [ARM64] 2813 l1tf=off [X86] 2814 mds=off [X86] 2815 tsx_async_abort=off [X86] 2816 kvm.nx_huge_pages=off [X86] 2817 2818 Exceptions: 2819 This does not have any effect on 2820 kvm.nx_huge_pages when 2821 kvm.nx_huge_pages=force. 2822 2823 auto (default) 2824 Mitigate all CPU vulnerabilities, but leave SMT 2825 enabled, even if it's vulnerable. This is for 2826 users who don't want to be surprised by SMT 2827 getting disabled across kernel upgrades, or who 2828 have other ways of avoiding SMT-based attacks. 2829 Equivalent to: (default behavior) 2830 2831 auto,nosmt 2832 Mitigate all CPU vulnerabilities, disabling SMT 2833 if needed. This is for users who always want to 2834 be fully mitigated, even if it means losing SMT. 2835 Equivalent to: l1tf=flush,nosmt [X86] 2836 mds=full,nosmt [X86] 2837 tsx_async_abort=full,nosmt [X86] 2838 2839 mminit_loglevel= 2840 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2841 parameter allows control of the logging verbosity for 2842 the additional memory initialisation checks. A value 2843 of 0 disables mminit logging and a level of 4 will 2844 log everything. Information is printed at KERN_DEBUG 2845 so loglevel=8 may also need to be specified. 2846 2847 module.sig_enforce 2848 [KNL] When CONFIG_MODULE_SIG is set, this means that 2849 modules without (valid) signatures will fail to load. 2850 Note that if CONFIG_MODULE_SIG_FORCE is set, that 2851 is always true, so this option does nothing. 2852 2853 module_blacklist= [KNL] Do not load a comma-separated list of 2854 modules. Useful for debugging problem modules. 2855 2856 mousedev.tap_time= 2857 [MOUSE] Maximum time between finger touching and 2858 leaving touchpad surface for touch to be considered 2859 a tap and be reported as a left button click (for 2860 touchpads working in absolute mode only). 2861 Format: <msecs> 2862 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 2863 reporting absolute coordinates, such as tablets 2864 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 2865 reporting absolute coordinates, such as tablets 2866 2867 movablecore= [KNL,X86,IA-64,PPC] 2868 Format: nn[KMGTPE] | nn% 2869 This parameter is the complement to kernelcore=, it 2870 specifies the amount of memory used for migratable 2871 allocations. If both kernelcore and movablecore is 2872 specified, then kernelcore will be at *least* the 2873 specified value but may be more. If movablecore on its 2874 own is specified, the administrator must be careful 2875 that the amount of memory usable for all allocations 2876 is not too small. 2877 2878 movable_node [KNL] Boot-time switch to make hotplugable memory 2879 NUMA nodes to be movable. This means that the memory 2880 of such nodes will be usable only for movable 2881 allocations which rules out almost all kernel 2882 allocations. Use with caution! 2883 2884 MTD_Partition= [MTD] 2885 Format: <name>,<region-number>,<size>,<offset> 2886 2887 MTD_Region= [MTD] Format: 2888 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 2889 2890 mtdparts= [MTD] 2891 See drivers/mtd/parsers/cmdlinepart.c 2892 2893 multitce=off [PPC] This parameter disables the use of the pSeries 2894 firmware feature for updating multiple TCE entries 2895 at a time. 2896 2897 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 2898 2899 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 2900 2901 boundary - index of last SLC block on Flex-OneNAND. 2902 The remaining blocks are configured as MLC blocks. 2903 lock - Configure if Flex-OneNAND boundary should be locked. 2904 Once locked, the boundary cannot be changed. 2905 1 indicates lock status, 0 indicates unlock status. 2906 2907 mtdset= [ARM] 2908 ARM/S3C2412 JIVE boot control 2909 2910 See arch/arm/mach-s3c2412/mach-jive.c 2911 2912 mtouchusb.raw_coordinates= 2913 [HW] Make the MicroTouch USB driver use raw coordinates 2914 ('y', default) or cooked coordinates ('n') 2915 2916 mtrr_chunk_size=nn[KMG] [X86] 2917 used for mtrr cleanup. It is largest continuous chunk 2918 that could hold holes aka. UC entries. 2919 2920 mtrr_gran_size=nn[KMG] [X86] 2921 Used for mtrr cleanup. It is granularity of mtrr block. 2922 Default is 1. 2923 Large value could prevent small alignment from 2924 using up MTRRs. 2925 2926 mtrr_spare_reg_nr=n [X86] 2927 Format: <integer> 2928 Range: 0,7 : spare reg number 2929 Default : 1 2930 Used for mtrr cleanup. It is spare mtrr entries number. 2931 Set to 2 or more if your graphical card needs more. 2932 2933 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 2934 2935 netdev= [NET] Network devices parameters 2936 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 2937 Note that mem_start is often overloaded to mean 2938 something different and driver-specific. 2939 This usage is only documented in each driver source 2940 file if at all. 2941 2942 nf_conntrack.acct= 2943 [NETFILTER] Enable connection tracking flow accounting 2944 0 to disable accounting 2945 1 to enable accounting 2946 Default value is 0. 2947 2948 nfsaddrs= [NFS] Deprecated. Use ip= instead. 2949 See Documentation/admin-guide/nfs/nfsroot.rst. 2950 2951 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 2952 See Documentation/admin-guide/nfs/nfsroot.rst. 2953 2954 nfsrootdebug [NFS] enable nfsroot debugging messages. 2955 See Documentation/admin-guide/nfs/nfsroot.rst. 2956 2957 nfs.callback_nr_threads= 2958 [NFSv4] set the total number of threads that the 2959 NFS client will assign to service NFSv4 callback 2960 requests. 2961 2962 nfs.callback_tcpport= 2963 [NFS] set the TCP port on which the NFSv4 callback 2964 channel should listen. 2965 2966 nfs.cache_getent= 2967 [NFS] sets the pathname to the program which is used 2968 to update the NFS client cache entries. 2969 2970 nfs.cache_getent_timeout= 2971 [NFS] sets the timeout after which an attempt to 2972 update a cache entry is deemed to have failed. 2973 2974 nfs.idmap_cache_timeout= 2975 [NFS] set the maximum lifetime for idmapper cache 2976 entries. 2977 2978 nfs.enable_ino64= 2979 [NFS] enable 64-bit inode numbers. 2980 If zero, the NFS client will fake up a 32-bit inode 2981 number for the readdir() and stat() syscalls instead 2982 of returning the full 64-bit number. 2983 The default is to return 64-bit inode numbers. 2984 2985 nfs.max_session_cb_slots= 2986 [NFSv4.1] Sets the maximum number of session 2987 slots the client will assign to the callback 2988 channel. This determines the maximum number of 2989 callbacks the client will process in parallel for 2990 a particular server. 2991 2992 nfs.max_session_slots= 2993 [NFSv4.1] Sets the maximum number of session slots 2994 the client will attempt to negotiate with the server. 2995 This limits the number of simultaneous RPC requests 2996 that the client can send to the NFSv4.1 server. 2997 Note that there is little point in setting this 2998 value higher than the max_tcp_slot_table_limit. 2999 3000 nfs.nfs4_disable_idmapping= 3001 [NFSv4] When set to the default of '1', this option 3002 ensures that both the RPC level authentication 3003 scheme and the NFS level operations agree to use 3004 numeric uids/gids if the mount is using the 3005 'sec=sys' security flavour. In effect it is 3006 disabling idmapping, which can make migration from 3007 legacy NFSv2/v3 systems to NFSv4 easier. 3008 Servers that do not support this mode of operation 3009 will be autodetected by the client, and it will fall 3010 back to using the idmapper. 3011 To turn off this behaviour, set the value to '0'. 3012 nfs.nfs4_unique_id= 3013 [NFS4] Specify an additional fixed unique ident- 3014 ification string that NFSv4 clients can insert into 3015 their nfs_client_id4 string. This is typically a 3016 UUID that is generated at system install time. 3017 3018 nfs.send_implementation_id = 3019 [NFSv4.1] Send client implementation identification 3020 information in exchange_id requests. 3021 If zero, no implementation identification information 3022 will be sent. 3023 The default is to send the implementation identification 3024 information. 3025 3026 nfs.recover_lost_locks = 3027 [NFSv4] Attempt to recover locks that were lost due 3028 to a lease timeout on the server. Please note that 3029 doing this risks data corruption, since there are 3030 no guarantees that the file will remain unchanged 3031 after the locks are lost. 3032 If you want to enable the kernel legacy behaviour of 3033 attempting to recover these locks, then set this 3034 parameter to '1'. 3035 The default parameter value of '0' causes the kernel 3036 not to attempt recovery of lost locks. 3037 3038 nfs4.layoutstats_timer = 3039 [NFSv4.2] Change the rate at which the kernel sends 3040 layoutstats to the pNFS metadata server. 3041 3042 Setting this to value to 0 causes the kernel to use 3043 whatever value is the default set by the layout 3044 driver. A non-zero value sets the minimum interval 3045 in seconds between layoutstats transmissions. 3046 3047 nfsd.nfs4_disable_idmapping= 3048 [NFSv4] When set to the default of '1', the NFSv4 3049 server will return only numeric uids and gids to 3050 clients using auth_sys, and will accept numeric uids 3051 and gids from such clients. This is intended to ease 3052 migration from NFSv2/v3. 3053 3054 nmi_debug= [KNL,SH] Specify one or more actions to take 3055 when a NMI is triggered. 3056 Format: [state][,regs][,debounce][,die] 3057 3058 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3059 Format: [panic,][nopanic,][num] 3060 Valid num: 0 or 1 3061 0 - turn hardlockup detector in nmi_watchdog off 3062 1 - turn hardlockup detector in nmi_watchdog on 3063 When panic is specified, panic when an NMI watchdog 3064 timeout occurs (or 'nopanic' to not panic on an NMI 3065 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3066 To disable both hard and soft lockup detectors, 3067 please see 'nowatchdog'. 3068 This is useful when you use a panic=... timeout and 3069 need the box quickly up again. 3070 3071 These settings can be accessed at runtime via 3072 the nmi_watchdog and hardlockup_panic sysctls. 3073 3074 netpoll.carrier_timeout= 3075 [NET] Specifies amount of time (in seconds) that 3076 netpoll should wait for a carrier. By default netpoll 3077 waits 4 seconds. 3078 3079 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3080 emulation library even if a 387 maths coprocessor 3081 is present. 3082 3083 no5lvl [X86-64] Disable 5-level paging mode. Forces 3084 kernel to use 4-level paging instead. 3085 3086 no_console_suspend 3087 [HW] Never suspend the console 3088 Disable suspending of consoles during suspend and 3089 hibernate operations. Once disabled, debugging 3090 messages can reach various consoles while the rest 3091 of the system is being put to sleep (ie, while 3092 debugging driver suspend/resume hooks). This may 3093 not work reliably with all consoles, but is known 3094 to work with serial and VGA consoles. 3095 To facilitate more flexible debugging, we also add 3096 console_suspend, a printk module parameter to control 3097 it. Users could use console_suspend (usually 3098 /sys/module/printk/parameters/console_suspend) to 3099 turn on/off it dynamically. 3100 3101 novmcoredd [KNL,KDUMP] 3102 Disable device dump. Device dump allows drivers to 3103 append dump data to vmcore so you can collect driver 3104 specified debug info. Drivers can append the data 3105 without any limit and this data is stored in memory, 3106 so this may cause significant memory stress. Disabling 3107 device dump can help save memory but the driver debug 3108 data will be no longer available. This parameter 3109 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3110 is set. 3111 3112 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3113 caches in the slab allocator. Saves per-node memory, 3114 but will impact performance. 3115 3116 noalign [KNL,ARM] 3117 3118 noaltinstr [S390] Disables alternative instructions patching 3119 (CPU alternatives feature). 3120 3121 noapic [SMP,APIC] Tells the kernel to not make use of any 3122 IOAPICs that may be present in the system. 3123 3124 noautogroup Disable scheduler automatic task group creation. 3125 3126 nobats [PPC] Do not use BATs for mapping kernel lowmem 3127 on "Classic" PPC cores. 3128 3129 nocache [ARM] 3130 3131 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 3132 3133 nodelayacct [KNL] Disable per-task delay accounting 3134 3135 nodsp [SH] Disable hardware DSP at boot time. 3136 3137 noefi Disable EFI runtime services support. 3138 3139 noexec [IA-64] 3140 3141 noexec [X86] 3142 On X86-32 available only on PAE configured kernels. 3143 noexec=on: enable non-executable mappings (default) 3144 noexec=off: disable non-executable mappings 3145 3146 nosmap [X86,PPC] 3147 Disable SMAP (Supervisor Mode Access Prevention) 3148 even if it is supported by processor. 3149 3150 nosmep [X86,PPC] 3151 Disable SMEP (Supervisor Mode Execution Prevention) 3152 even if it is supported by processor. 3153 3154 noexec32 [X86-64] 3155 This affects only 32-bit executables. 3156 noexec32=on: enable non-executable mappings (default) 3157 read doesn't imply executable mappings 3158 noexec32=off: disable non-executable mappings 3159 read implies executable mappings 3160 3161 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3162 3163 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3164 register save and restore. The kernel will only save 3165 legacy floating-point registers on task switch. 3166 3167 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings. 3168 3169 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3170 Equivalent to smt=1. 3171 3172 [KNL,x86] Disable symmetric multithreading (SMT). 3173 nosmt=force: Force disable SMT, cannot be undone 3174 via the sysfs control file. 3175 3176 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3177 (bounds check bypass). With this option data leaks are 3178 possible in the system. 3179 3180 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3181 the Spectre variant 2 (indirect branch prediction) 3182 vulnerability. System may allow data leaks with this 3183 option. 3184 3185 nospec_store_bypass_disable 3186 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3187 3188 noxsave [BUGS=X86] Disables x86 extended register state save 3189 and restore using xsave. The kernel will fallback to 3190 enabling legacy floating-point and sse state. 3191 3192 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3193 register states. The kernel will fall back to use 3194 xsave to save the states. By using this parameter, 3195 performance of saving the states is degraded because 3196 xsave doesn't support modified optimization while 3197 xsaveopt supports it on xsaveopt enabled systems. 3198 3199 noxsaves [X86] Disables xsaves and xrstors used in saving and 3200 restoring x86 extended register state in compacted 3201 form of xsave area. The kernel will fall back to use 3202 xsaveopt and xrstor to save and restore the states 3203 in standard form of xsave area. By using this 3204 parameter, xsave area per process might occupy more 3205 memory on xsaves enabled systems. 3206 3207 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or 3208 wfi(ARM) instruction doesn't work correctly and not to 3209 use it. This is also useful when using JTAG debugger. 3210 3211 no_file_caps Tells the kernel not to honor file capabilities. The 3212 only way then for a file to be executed with privilege 3213 is to be setuid root or executed by root. 3214 3215 nohalt [IA-64] Tells the kernel not to use the power saving 3216 function PAL_HALT_LIGHT when idle. This increases 3217 power-consumption. On the positive side, it reduces 3218 interrupt wake-up latency, which may improve performance 3219 in certain environments such as networked servers or 3220 real-time systems. 3221 3222 nohibernate [HIBERNATION] Disable hibernation and resume. 3223 3224 nohz= [KNL] Boottime enable/disable dynamic ticks 3225 Valid arguments: on, off 3226 Default: on 3227 3228 nohz_full= [KNL,BOOT,SMP,ISOL] 3229 The argument is a cpu list, as described above. 3230 In kernels built with CONFIG_NO_HZ_FULL=y, set 3231 the specified list of CPUs whose tick will be stopped 3232 whenever possible. The boot CPU will be forced outside 3233 the range to maintain the timekeeping. Any CPUs 3234 in this list will have their RCU callbacks offloaded, 3235 just as if they had also been called out in the 3236 rcu_nocbs= boot parameter. 3237 3238 noiotrap [SH] Disables trapped I/O port accesses. 3239 3240 noirqdebug [X86-32] Disables the code which attempts to detect and 3241 disable unhandled interrupt sources. 3242 3243 no_timer_check [X86,APIC] Disables the code which tests for 3244 broken timer IRQ sources. 3245 3246 noisapnp [ISAPNP] Disables ISA PnP code. 3247 3248 noinitrd [RAM] Tells the kernel not to load any configured 3249 initial RAM disk. 3250 3251 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3252 remapping. 3253 [Deprecated - use intremap=off] 3254 3255 nointroute [IA-64] 3256 3257 noinvpcid [X86] Disable the INVPCID cpu feature. 3258 3259 nojitter [IA-64] Disables jitter checking for ITC timers. 3260 3261 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3262 3263 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3264 fault handling. 3265 3266 no-vmw-sched-clock 3267 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3268 clock and use the default one. 3269 3270 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time 3271 accounting. steal time is computed, but won't 3272 influence scheduler behaviour 3273 3274 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3275 3276 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3277 3278 noltlbs [PPC] Do not use large page/tlb entries for kernel 3279 lowmem mapping on PPC40x and PPC8xx 3280 3281 nomca [IA-64] Disable machine check abort handling 3282 3283 nomce [X86-32] Disable Machine Check Exception 3284 3285 nomfgpt [X86-32] Disable Multi-Function General Purpose 3286 Timer usage (for AMD Geode machines). 3287 3288 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3289 shutdown the other cpus. Instead use the REBOOT_VECTOR 3290 irq. 3291 3292 nomodule Disable module load 3293 3294 nopat [X86] Disable PAT (page attribute table extension of 3295 pagetables) support. 3296 3297 nopcid [X86-64] Disable the PCID cpu feature. 3298 3299 norandmaps Don't use address space randomization. Equivalent to 3300 echo 0 > /proc/sys/kernel/randomize_va_space 3301 3302 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3303 with UP alternatives 3304 3305 nordrand [X86] Disable kernel use of the RDRAND and 3306 RDSEED instructions even if they are supported 3307 by the processor. RDRAND and RDSEED are still 3308 available to user space applications. 3309 3310 noresume [SWSUSP] Disables resume and restores original swap 3311 space. 3312 3313 no-scroll [VGA] Disables scrollback. 3314 This is required for the Braillex ib80-piezo Braille 3315 reader made by F.H. Papenmeier (Germany). 3316 3317 nosbagart [IA-64] 3318 3319 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3320 3321 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3322 and disable the IO APIC. legacy for "maxcpus=0". 3323 3324 nosoftlockup [KNL] Disable the soft-lockup detector. 3325 3326 nosync [HW,M68K] Disables sync negotiation for all devices. 3327 3328 nowatchdog [KNL] Disable both lockup detectors, i.e. 3329 soft-lockup and NMI watchdog (hard-lockup). 3330 3331 nowb [ARM] 3332 3333 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3334 3335 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3336 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3337 Some features depend on CPU0. Known dependencies are: 3338 1. Resume from suspend/hibernate depends on CPU0. 3339 Suspend/hibernate will fail if CPU0 is offline and you 3340 need to online CPU0 before suspend/hibernate. 3341 2. PIC interrupts also depend on CPU0. CPU0 can't be 3342 removed if a PIC interrupt is detected. 3343 It's said poweroff/reboot may depend on CPU0 on some 3344 machines although I haven't seen such issues so far 3345 after CPU0 is offline on a few tested machines. 3346 If the dependencies are under your control, you can 3347 turn on cpu0_hotplug. 3348 3349 nps_mtm_hs_ctr= [KNL,ARC] 3350 This parameter sets the maximum duration, in 3351 cycles, each HW thread of the CTOP can run 3352 without interruptions, before HW switches it. 3353 The actual maximum duration is 16 times this 3354 parameter's value. 3355 Format: integer between 1 and 255 3356 Default: 255 3357 3358 nptcg= [IA-64] Override max number of concurrent global TLB 3359 purges which is reported from either PAL_VM_SUMMARY or 3360 SAL PALO. 3361 3362 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3363 could support. nr_cpus=n : n >= 1 limits the kernel to 3364 support 'n' processors. It could be larger than the 3365 number of already plugged CPU during bootup, later in 3366 runtime you can physically add extra cpu until it reaches 3367 n. So during boot up some boot time memory for per-cpu 3368 variables need be pre-allocated for later physical cpu 3369 hot plugging. 3370 3371 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3372 3373 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing. 3374 Allowed values are enable and disable 3375 3376 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3377 'node', 'default' can be specified 3378 This can be set from sysctl after boot. 3379 See Documentation/admin-guide/sysctl/vm.rst for details. 3380 3381 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3382 See Documentation/core-api/debugging-via-ohci1394.rst for more 3383 info. 3384 3385 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3386 Rather than timing out after 20 ms if an EC 3387 command is not properly ACKed, override the length 3388 of the timeout. We have interrupts disabled while 3389 waiting for the ACK, so if this is set too high 3390 interrupts *may* be lost! 3391 3392 omap_mux= [OMAP] Override bootloader pin multiplexing. 3393 Format: <mux_mode0.mode_name=value>... 3394 For example, to override I2C bus2: 3395 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3396 3397 oprofile.timer= [HW] 3398 Use timer interrupt instead of performance counters 3399 3400 oprofile.cpu_type= Force an oprofile cpu type 3401 This might be useful if you have an older oprofile 3402 userland or if you want common events. 3403 Format: { arch_perfmon } 3404 arch_perfmon: [X86] Force use of architectural 3405 perfmon on Intel CPUs instead of the 3406 CPU specific event set. 3407 timer: [X86] Force use of architectural NMI 3408 timer mode (see also oprofile.timer 3409 for generic hr timer mode) 3410 3411 oops=panic Always panic on oopses. Default is to just kill the 3412 process, but there is a small probability of 3413 deadlocking the machine. 3414 This will also cause panics on machine check exceptions. 3415 Useful together with panic=30 to trigger a reboot. 3416 3417 page_alloc.shuffle= 3418 [KNL] Boolean flag to control whether the page allocator 3419 should randomize its free lists. The randomization may 3420 be automatically enabled if the kernel detects it is 3421 running on a platform with a direct-mapped memory-side 3422 cache, and this parameter can be used to 3423 override/disable that behavior. The state of the flag 3424 can be read from sysfs at: 3425 /sys/module/page_alloc/parameters/shuffle. 3426 3427 page_owner= [KNL] Boot-time page_owner enabling option. 3428 Storage of the information about who allocated 3429 each page is disabled in default. With this switch, 3430 we can turn it on. 3431 on: enable the feature 3432 3433 page_poison= [KNL] Boot-time parameter changing the state of 3434 poisoning on the buddy allocator, available with 3435 CONFIG_PAGE_POISONING=y. 3436 off: turn off poisoning (default) 3437 on: turn on poisoning 3438 3439 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3440 timeout > 0: seconds before rebooting 3441 timeout = 0: wait forever 3442 timeout < 0: reboot immediately 3443 Format: <timeout> 3444 3445 panic_print= Bitmask for printing system info when panic happens. 3446 User can chose combination of the following bits: 3447 bit 0: print all tasks info 3448 bit 1: print system memory info 3449 bit 2: print timer info 3450 bit 3: print locks info if CONFIG_LOCKDEP is on 3451 bit 4: print ftrace buffer 3452 bit 5: print all printk messages in buffer 3453 3454 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 3455 Format: <hex>[,nousertaint] 3456 Hexadecimal bitmask representing the set of TAINT flags 3457 that will cause the kernel to panic when add_taint() is 3458 called with any of the flags in this set. 3459 The optional switch "nousertaint" can be utilized to 3460 prevent userspace forced crashes by writing to sysctl 3461 /proc/sys/kernel/tainted any flagset matching with the 3462 bitmask set on panic_on_taint. 3463 See Documentation/admin-guide/tainted-kernels.rst for 3464 extra details on the taint flags that users can pick 3465 to compose the bitmask to assign to panic_on_taint. 3466 3467 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3468 on a WARN(). 3469 3470 crash_kexec_post_notifiers 3471 Run kdump after running panic-notifiers and dumping 3472 kmsg. This only for the users who doubt kdump always 3473 succeeds in any situation. 3474 Note that this also increases risks of kdump failure, 3475 because some panic notifiers can make the crashed 3476 kernel more unstable. 3477 3478 parkbd.port= [HW] Parallel port number the keyboard adapter is 3479 connected to, default is 0. 3480 Format: <parport#> 3481 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3482 0 for XT, 1 for AT (default is AT). 3483 Format: <mode> 3484 3485 parport= [HW,PPT] Specify parallel ports. 0 disables. 3486 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3487 Use 'auto' to force the driver to use any 3488 IRQ/DMA settings detected (the default is to 3489 ignore detected IRQ/DMA settings because of 3490 possible conflicts). You can specify the base 3491 address, IRQ, and DMA settings; IRQ and DMA 3492 should be numbers, or 'auto' (for using detected 3493 settings on that particular port), or 'nofifo' 3494 (to avoid using a FIFO even if it is detected). 3495 Parallel ports are assigned in the order they 3496 are specified on the command line, starting 3497 with parport0. 3498 3499 parport_init_mode= [HW,PPT] 3500 Configure VIA parallel port to operate in 3501 a specific mode. This is necessary on Pegasos 3502 computer where firmware has no options for setting 3503 up parallel port mode and sets it to spp. 3504 Currently this function knows 686a and 8231 chips. 3505 Format: [spp|ps2|epp|ecp|ecpepp] 3506 3507 pause_on_oops= 3508 Halt all CPUs after the first oops has been printed for 3509 the specified number of seconds. This is to be used if 3510 your oopses keep scrolling off the screen. 3511 3512 pcbit= [HW,ISDN] 3513 3514 pcd. [PARIDE] 3515 See header of drivers/block/paride/pcd.c. 3516 See also Documentation/admin-guide/blockdev/paride.rst. 3517 3518 pci=option[,option...] [PCI] various PCI subsystem options. 3519 3520 Some options herein operate on a specific device 3521 or a set of devices (<pci_dev>). These are 3522 specified in one of the following formats: 3523 3524 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3525 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3526 3527 Note: the first format specifies a PCI 3528 bus/device/function address which may change 3529 if new hardware is inserted, if motherboard 3530 firmware changes, or due to changes caused 3531 by other kernel parameters. If the 3532 domain is left unspecified, it is 3533 taken to be zero. Optionally, a path 3534 to a device through multiple device/function 3535 addresses can be specified after the base 3536 address (this is more robust against 3537 renumbering issues). The second format 3538 selects devices using IDs from the 3539 configuration space which may match multiple 3540 devices in the system. 3541 3542 earlydump dump PCI config space before the kernel 3543 changes anything 3544 off [X86] don't probe for the PCI bus 3545 bios [X86-32] force use of PCI BIOS, don't access 3546 the hardware directly. Use this if your machine 3547 has a non-standard PCI host bridge. 3548 nobios [X86-32] disallow use of PCI BIOS, only direct 3549 hardware access methods are allowed. Use this 3550 if you experience crashes upon bootup and you 3551 suspect they are caused by the BIOS. 3552 conf1 [X86] Force use of PCI Configuration Access 3553 Mechanism 1 (config address in IO port 0xCF8, 3554 data in IO port 0xCFC, both 32-bit). 3555 conf2 [X86] Force use of PCI Configuration Access 3556 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3557 the function, IO port 0xCFA, also 8-bit, sets 3558 bus number. The config space is then accessed 3559 through ports 0xC000-0xCFFF). 3560 See http://wiki.osdev.org/PCI for more info 3561 on the configuration access mechanisms. 3562 noaer [PCIE] If the PCIEAER kernel config parameter is 3563 enabled, this kernel boot option can be used to 3564 disable the use of PCIE advanced error reporting. 3565 nodomains [PCI] Disable support for multiple PCI 3566 root domains (aka PCI segments, in ACPI-speak). 3567 nommconf [X86] Disable use of MMCONFIG for PCI 3568 Configuration 3569 check_enable_amd_mmconf [X86] check for and enable 3570 properly configured MMIO access to PCI 3571 config space on AMD family 10h CPU 3572 nomsi [MSI] If the PCI_MSI kernel config parameter is 3573 enabled, this kernel boot option can be used to 3574 disable the use of MSI interrupts system-wide. 3575 noioapicquirk [APIC] Disable all boot interrupt quirks. 3576 Safety option to keep boot IRQs enabled. This 3577 should never be necessary. 3578 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3579 primary IO-APIC for bridges that cannot disable 3580 boot IRQs. This fixes a source of spurious IRQs 3581 when the system masks IRQs. 3582 noioapicreroute [APIC] Disable workaround that uses the 3583 boot IRQ equivalent of an IRQ that connects to 3584 a chipset where boot IRQs cannot be disabled. 3585 The opposite of ioapicreroute. 3586 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3587 routing table. These calls are known to be buggy 3588 on several machines and they hang the machine 3589 when used, but on other computers it's the only 3590 way to get the interrupt routing table. Try 3591 this option if the kernel is unable to allocate 3592 IRQs or discover secondary PCI buses on your 3593 motherboard. 3594 rom [X86] Assign address space to expansion ROMs. 3595 Use with caution as certain devices share 3596 address decoders between ROMs and other 3597 resources. 3598 norom [X86] Do not assign address space to 3599 expansion ROMs that do not already have 3600 BIOS assigned address ranges. 3601 nobar [X86] Do not assign address space to the 3602 BARs that weren't assigned by the BIOS. 3603 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3604 assigned automatically to PCI devices. You can 3605 make the kernel exclude IRQs of your ISA cards 3606 this way. 3607 pirqaddr=0xAAAAA [X86] Specify the physical address 3608 of the PIRQ table (normally generated 3609 by the BIOS) if it is outside the 3610 F0000h-100000h range. 3611 lastbus=N [X86] Scan all buses thru bus #N. Can be 3612 useful if the kernel is unable to find your 3613 secondary buses and you want to tell it 3614 explicitly which ones they are. 3615 assign-busses [X86] Always assign all PCI bus 3616 numbers ourselves, overriding 3617 whatever the firmware may have done. 3618 usepirqmask [X86] Honor the possible IRQ mask stored 3619 in the BIOS $PIR table. This is needed on 3620 some systems with broken BIOSes, notably 3621 some HP Pavilion N5400 and Omnibook XE3 3622 notebooks. This will have no effect if ACPI 3623 IRQ routing is enabled. 3624 noacpi [X86] Do not use ACPI for IRQ routing 3625 or for PCI scanning. 3626 use_crs [X86] Use PCI host bridge window information 3627 from ACPI. On BIOSes from 2008 or later, this 3628 is enabled by default. If you need to use this, 3629 please report a bug. 3630 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3631 If you need to use this, please report a bug. 3632 routeirq Do IRQ routing for all PCI devices. 3633 This is normally done in pci_enable_device(), 3634 so this option is a temporary workaround 3635 for broken drivers that don't call it. 3636 skip_isa_align [X86] do not align io start addr, so can 3637 handle more pci cards 3638 noearly [X86] Don't do any early type 1 scanning. 3639 This might help on some broken boards which 3640 machine check when some devices' config space 3641 is read. But various workarounds are disabled 3642 and some IOMMU drivers will not work. 3643 bfsort Sort PCI devices into breadth-first order. 3644 This sorting is done to get a device 3645 order compatible with older (<= 2.4) kernels. 3646 nobfsort Don't sort PCI devices into breadth-first order. 3647 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3648 tuning and use the BIOS-configured MPS defaults. 3649 pcie_bus_safe Set every device's MPS to the largest value 3650 supported by all devices below the root complex. 3651 pcie_bus_perf Set device MPS to the largest allowable MPS 3652 based on its parent bus. Also set MRRS (Max 3653 Read Request Size) to the largest supported 3654 value (no larger than the MPS that the device 3655 or bus can support) for best performance. 3656 pcie_bus_peer2peer Set every device's MPS to 128B, which 3657 every device is guaranteed to support. This 3658 configuration allows peer-to-peer DMA between 3659 any pair of devices, possibly at the cost of 3660 reduced performance. This also guarantees 3661 that hot-added devices will work. 3662 cbiosize=nn[KMG] The fixed amount of bus space which is 3663 reserved for the CardBus bridge's IO window. 3664 The default value is 256 bytes. 3665 cbmemsize=nn[KMG] The fixed amount of bus space which is 3666 reserved for the CardBus bridge's memory 3667 window. The default value is 64 megabytes. 3668 resource_alignment= 3669 Format: 3670 [<order of align>@]<pci_dev>[; ...] 3671 Specifies alignment and device to reassign 3672 aligned memory resources. How to 3673 specify the device is described above. 3674 If <order of align> is not specified, 3675 PAGE_SIZE is used as alignment. 3676 A PCI-PCI bridge can be specified if resource 3677 windows need to be expanded. 3678 To specify the alignment for several 3679 instances of a device, the PCI vendor, 3680 device, subvendor, and subdevice may be 3681 specified, e.g., 12@pci:8086:9c22:103c:198f 3682 for 4096-byte alignment. 3683 ecrc= Enable/disable PCIe ECRC (transaction layer 3684 end-to-end CRC checking). 3685 bios: Use BIOS/firmware settings. This is the 3686 the default. 3687 off: Turn ECRC off 3688 on: Turn ECRC on. 3689 hpiosize=nn[KMG] The fixed amount of bus space which is 3690 reserved for hotplug bridge's IO window. 3691 Default size is 256 bytes. 3692 hpmmiosize=nn[KMG] The fixed amount of bus space which is 3693 reserved for hotplug bridge's MMIO window. 3694 Default size is 2 megabytes. 3695 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 3696 reserved for hotplug bridge's MMIO_PREF window. 3697 Default size is 2 megabytes. 3698 hpmemsize=nn[KMG] The fixed amount of bus space which is 3699 reserved for hotplug bridge's MMIO and 3700 MMIO_PREF window. 3701 Default size is 2 megabytes. 3702 hpbussize=nn The minimum amount of additional bus numbers 3703 reserved for buses below a hotplug bridge. 3704 Default is 1. 3705 realloc= Enable/disable reallocating PCI bridge resources 3706 if allocations done by BIOS are too small to 3707 accommodate resources required by all child 3708 devices. 3709 off: Turn realloc off 3710 on: Turn realloc on 3711 realloc same as realloc=on 3712 noari do not use PCIe ARI. 3713 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3714 do not use PCIe ATS (and IOMMU device IOTLB). 3715 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3716 only look for one device below a PCIe downstream 3717 port. 3718 big_root_window Try to add a big 64bit memory window to the PCIe 3719 root complex on AMD CPUs. Some GFX hardware 3720 can resize a BAR to allow access to all VRAM. 3721 Adding the window is slightly risky (it may 3722 conflict with unreported devices), so this 3723 taints the kernel. 3724 disable_acs_redir=<pci_dev>[; ...] 3725 Specify one or more PCI devices (in the format 3726 specified above) separated by semicolons. 3727 Each device specified will have the PCI ACS 3728 redirect capabilities forced off which will 3729 allow P2P traffic between devices through 3730 bridges without forcing it upstream. Note: 3731 this removes isolation between devices and 3732 may put more devices in an IOMMU group. 3733 force_floating [S390] Force usage of floating interrupts. 3734 nomio [S390] Do not use MIO instructions. 3735 norid [S390] ignore the RID field and force use of 3736 one PCI domain per PCI function 3737 3738 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3739 Management. 3740 off Disable ASPM. 3741 force Enable ASPM even on devices that claim not to support it. 3742 WARNING: Forcing ASPM on may cause system lockups. 3743 3744 pcie_ports= [PCIE] PCIe port services handling: 3745 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3746 even if the platform doesn't give the OS permission to 3747 use them. This may cause conflicts if the platform 3748 also tries to use these services. 3749 dpc-native Use native PCIe service for DPC only. May 3750 cause conflicts if firmware uses AER or DPC. 3751 compat Disable native PCIe services (PME, AER, DPC, PCIe 3752 hotplug). 3753 3754 pcie_port_pm= [PCIE] PCIe port power management handling: 3755 off Disable power management of all PCIe ports 3756 force Forcibly enable power management of all PCIe ports 3757 3758 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 3759 nomsi Do not use MSI for native PCIe PME signaling (this makes 3760 all PCIe root ports use INTx for all services). 3761 3762 pcmv= [HW,PCMCIA] BadgePAD 4 3763 3764 pd_ignore_unused 3765 [PM] 3766 Keep all power-domains already enabled by bootloader on, 3767 even if no driver has claimed them. This is useful 3768 for debug and development, but should not be 3769 needed on a platform with proper driver support. 3770 3771 pd. [PARIDE] 3772 See Documentation/admin-guide/blockdev/paride.rst. 3773 3774 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 3775 boot time. 3776 Format: { 0 | 1 } 3777 See arch/parisc/kernel/pdc_chassis.c 3778 3779 percpu_alloc= Select which percpu first chunk allocator to use. 3780 Currently supported values are "embed" and "page". 3781 Archs may support subset or none of the selections. 3782 See comments in mm/percpu.c for details on each 3783 allocator. This parameter is primarily for debugging 3784 and performance comparison. 3785 3786 pf. [PARIDE] 3787 See Documentation/admin-guide/blockdev/paride.rst. 3788 3789 pg. [PARIDE] 3790 See Documentation/admin-guide/blockdev/paride.rst. 3791 3792 pirq= [SMP,APIC] Manual mp-table setup 3793 See Documentation/x86/i386/IO-APIC.rst. 3794 3795 plip= [PPT,NET] Parallel port network link 3796 Format: { parport<nr> | timid | 0 } 3797 See also Documentation/admin-guide/parport.rst. 3798 3799 pmtmr= [X86] Manual setup of pmtmr I/O Port. 3800 Override pmtimer IOPort with a hex value. 3801 e.g. pmtmr=0x508 3802 3803 pm_debug_messages [SUSPEND,KNL] 3804 Enable suspend/resume debug messages during boot up. 3805 3806 pnp.debug=1 [PNP] 3807 Enable PNP debug messages (depends on the 3808 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 3809 via /sys/module/pnp/parameters/debug. We always show 3810 current resource usage; turning this on also shows 3811 possible settings and some assignment information. 3812 3813 pnpacpi= [ACPI] 3814 { off } 3815 3816 pnpbios= [ISAPNP] 3817 { on | off | curr | res | no-curr | no-res } 3818 3819 pnp_reserve_irq= 3820 [ISAPNP] Exclude IRQs for the autoconfiguration 3821 3822 pnp_reserve_dma= 3823 [ISAPNP] Exclude DMAs for the autoconfiguration 3824 3825 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 3826 Ranges are in pairs (I/O port base and size). 3827 3828 pnp_reserve_mem= 3829 [ISAPNP] Exclude memory regions for the 3830 autoconfiguration. 3831 Ranges are in pairs (memory base and size). 3832 3833 ports= [IP_VS_FTP] IPVS ftp helper module 3834 Default is 21. 3835 Up to 8 (IP_VS_APP_MAX_PORTS) ports 3836 may be specified. 3837 Format: <port>,<port>.... 3838 3839 powersave=off [PPC] This option disables power saving features. 3840 It specifically disables cpuidle and sets the 3841 platform machine description specific power_save 3842 function to NULL. On Idle the CPU just reduces 3843 execution priority. 3844 3845 ppc_strict_facility_enable 3846 [PPC] This option catches any kernel floating point, 3847 Altivec, VSX and SPE outside of regions specifically 3848 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 3849 There is some performance impact when enabling this. 3850 3851 ppc_tm= [PPC] 3852 Format: {"off"} 3853 Disable Hardware Transactional Memory 3854 3855 print-fatal-signals= 3856 [KNL] debug: print fatal signals 3857 3858 If enabled, warn about various signal handling 3859 related application anomalies: too many signals, 3860 too many POSIX.1 timers, fatal signals causing a 3861 coredump - etc. 3862 3863 If you hit the warning due to signal overflow, 3864 you might want to try "ulimit -i unlimited". 3865 3866 default: off. 3867 3868 printk.always_kmsg_dump= 3869 Trigger kmsg_dump for cases other than kernel oops or 3870 panics 3871 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3872 default: disabled 3873 3874 printk.devkmsg={on,off,ratelimit} 3875 Control writing to /dev/kmsg. 3876 on - unlimited logging to /dev/kmsg from userspace 3877 off - logging to /dev/kmsg disabled 3878 ratelimit - ratelimit the logging 3879 Default: ratelimit 3880 3881 printk.time= Show timing data prefixed to each printk message line 3882 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3883 3884 processor.max_cstate= [HW,ACPI] 3885 Limit processor to maximum C-state 3886 max_cstate=9 overrides any DMI blacklist limit. 3887 3888 processor.nocst [HW,ACPI] 3889 Ignore the _CST method to determine C-states, 3890 instead using the legacy FADT method 3891 3892 profile= [KNL] Enable kernel profiling via /proc/profile 3893 Format: [<profiletype>,]<number> 3894 Param: <profiletype>: "schedule", "sleep", or "kvm" 3895 [defaults to kernel profiling] 3896 Param: "schedule" - profile schedule points. 3897 Param: "sleep" - profile D-state sleeping (millisecs). 3898 Requires CONFIG_SCHEDSTATS 3899 Param: "kvm" - profile VM exits. 3900 Param: <number> - step/bucket size as a power of 2 for 3901 statistical time based profiling. 3902 3903 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk 3904 before loading. 3905 See Documentation/admin-guide/blockdev/ramdisk.rst. 3906 3907 prot_virt= [S390] enable hosting protected virtual machines 3908 isolated from the hypervisor (if hardware supports 3909 that). 3910 Format: <bool> 3911 3912 psi= [KNL] Enable or disable pressure stall information 3913 tracking. 3914 Format: <bool> 3915 3916 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 3917 probe for; one of (bare|imps|exps|lifebook|any). 3918 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 3919 per second. 3920 psmouse.resetafter= [HW,MOUSE] 3921 Try to reset the device after so many bad packets 3922 (0 = never). 3923 psmouse.resolution= 3924 [HW,MOUSE] Set desired mouse resolution, in dpi. 3925 psmouse.smartscroll= 3926 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 3927 0 = disabled, 1 = enabled (default). 3928 3929 pstore.backend= Specify the name of the pstore backend to use 3930 3931 pt. [PARIDE] 3932 See Documentation/admin-guide/blockdev/paride.rst. 3933 3934 pti= [X86_64] Control Page Table Isolation of user and 3935 kernel address spaces. Disabling this feature 3936 removes hardening, but improves performance of 3937 system calls and interrupts. 3938 3939 on - unconditionally enable 3940 off - unconditionally disable 3941 auto - kernel detects whether your CPU model is 3942 vulnerable to issues that PTI mitigates 3943 3944 Not specifying this option is equivalent to pti=auto. 3945 3946 nopti [X86_64] 3947 Equivalent to pti=off 3948 3949 pty.legacy_count= 3950 [KNL] Number of legacy pty's. Overwrites compiled-in 3951 default number. 3952 3953 quiet [KNL] Disable most log messages 3954 3955 r128= [HW,DRM] 3956 3957 raid= [HW,RAID] 3958 See Documentation/admin-guide/md.rst. 3959 3960 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 3961 See Documentation/admin-guide/blockdev/ramdisk.rst. 3962 3963 random.trust_cpu={on,off} 3964 [KNL] Enable or disable trusting the use of the 3965 CPU's random number generator (if available) to 3966 fully seed the kernel's CRNG. Default is controlled 3967 by CONFIG_RANDOM_TRUST_CPU. 3968 3969 ras=option[,option,...] [KNL] RAS-specific options 3970 3971 cec_disable [X86] 3972 Disable the Correctable Errors Collector, 3973 see CONFIG_RAS_CEC help text. 3974 3975 rcu_nocbs= [KNL] 3976 The argument is a cpu list, as described above, 3977 except that the string "all" can be used to 3978 specify every CPU on the system. 3979 3980 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 3981 the specified list of CPUs to be no-callback CPUs. 3982 Invocation of these CPUs' RCU callbacks will be 3983 offloaded to "rcuox/N" kthreads created for that 3984 purpose, where "x" is "p" for RCU-preempt, and 3985 "s" for RCU-sched, and "N" is the CPU number. 3986 This reduces OS jitter on the offloaded CPUs, 3987 which can be useful for HPC and real-time 3988 workloads. It can also improve energy efficiency 3989 for asymmetric multiprocessors. 3990 3991 rcu_nocb_poll [KNL] 3992 Rather than requiring that offloaded CPUs 3993 (specified by rcu_nocbs= above) explicitly 3994 awaken the corresponding "rcuoN" kthreads, 3995 make these kthreads poll for callbacks. 3996 This improves the real-time response for the 3997 offloaded CPUs by relieving them of the need to 3998 wake up the corresponding kthread, but degrades 3999 energy efficiency by requiring that the kthreads 4000 periodically wake up to do the polling. 4001 4002 rcutree.blimit= [KNL] 4003 Set maximum number of finished RCU callbacks to 4004 process in one batch. 4005 4006 rcutree.dump_tree= [KNL] 4007 Dump the structure of the rcu_node combining tree 4008 out at early boot. This is used for diagnostic 4009 purposes, to verify correct tree setup. 4010 4011 rcutree.gp_cleanup_delay= [KNL] 4012 Set the number of jiffies to delay each step of 4013 RCU grace-period cleanup. 4014 4015 rcutree.gp_init_delay= [KNL] 4016 Set the number of jiffies to delay each step of 4017 RCU grace-period initialization. 4018 4019 rcutree.gp_preinit_delay= [KNL] 4020 Set the number of jiffies to delay each step of 4021 RCU grace-period pre-initialization, that is, 4022 the propagation of recent CPU-hotplug changes up 4023 the rcu_node combining tree. 4024 4025 rcutree.use_softirq= [KNL] 4026 If set to zero, move all RCU_SOFTIRQ processing to 4027 per-CPU rcuc kthreads. Defaults to a non-zero 4028 value, meaning that RCU_SOFTIRQ is used by default. 4029 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4030 4031 rcutree.rcu_fanout_exact= [KNL] 4032 Disable autobalancing of the rcu_node combining 4033 tree. This is used by rcutorture, and might 4034 possibly be useful for architectures having high 4035 cache-to-cache transfer latencies. 4036 4037 rcutree.rcu_fanout_leaf= [KNL] 4038 Change the number of CPUs assigned to each 4039 leaf rcu_node structure. Useful for very 4040 large systems, which will choose the value 64, 4041 and for NUMA systems with large remote-access 4042 latencies, which will choose a value aligned 4043 with the appropriate hardware boundaries. 4044 4045 rcutree.jiffies_till_first_fqs= [KNL] 4046 Set delay from grace-period initialization to 4047 first attempt to force quiescent states. 4048 Units are jiffies, minimum value is zero, 4049 and maximum value is HZ. 4050 4051 rcutree.jiffies_till_next_fqs= [KNL] 4052 Set delay between subsequent attempts to force 4053 quiescent states. Units are jiffies, minimum 4054 value is one, and maximum value is HZ. 4055 4056 rcutree.jiffies_till_sched_qs= [KNL] 4057 Set required age in jiffies for a 4058 given grace period before RCU starts 4059 soliciting quiescent-state help from 4060 rcu_note_context_switch() and cond_resched(). 4061 If not specified, the kernel will calculate 4062 a value based on the most recent settings 4063 of rcutree.jiffies_till_first_fqs 4064 and rcutree.jiffies_till_next_fqs. 4065 This calculated value may be viewed in 4066 rcutree.jiffies_to_sched_qs. Any attempt to set 4067 rcutree.jiffies_to_sched_qs will be cheerfully 4068 overwritten. 4069 4070 rcutree.kthread_prio= [KNL,BOOT] 4071 Set the SCHED_FIFO priority of the RCU per-CPU 4072 kthreads (rcuc/N). This value is also used for 4073 the priority of the RCU boost threads (rcub/N) 4074 and for the RCU grace-period kthreads (rcu_bh, 4075 rcu_preempt, and rcu_sched). If RCU_BOOST is 4076 set, valid values are 1-99 and the default is 1 4077 (the least-favored priority). Otherwise, when 4078 RCU_BOOST is not set, valid values are 0-99 and 4079 the default is zero (non-realtime operation). 4080 4081 rcutree.rcu_nocb_gp_stride= [KNL] 4082 Set the number of NOCB callback kthreads in 4083 each group, which defaults to the square root 4084 of the number of CPUs. Larger numbers reduce 4085 the wakeup overhead on the global grace-period 4086 kthread, but increases that same overhead on 4087 each group's NOCB grace-period kthread. 4088 4089 rcutree.qhimark= [KNL] 4090 Set threshold of queued RCU callbacks beyond which 4091 batch limiting is disabled. 4092 4093 rcutree.qlowmark= [KNL] 4094 Set threshold of queued RCU callbacks below which 4095 batch limiting is re-enabled. 4096 4097 rcutree.qovld= [KNL] 4098 Set threshold of queued RCU callbacks beyond which 4099 RCU's force-quiescent-state scan will aggressively 4100 enlist help from cond_resched() and sched IPIs to 4101 help CPUs more quickly reach quiescent states. 4102 Set to less than zero to make this be set based 4103 on rcutree.qhimark at boot time and to zero to 4104 disable more aggressive help enlistment. 4105 4106 rcutree.rcu_idle_gp_delay= [KNL] 4107 Set wakeup interval for idle CPUs that have 4108 RCU callbacks (RCU_FAST_NO_HZ=y). 4109 4110 rcutree.rcu_idle_lazy_gp_delay= [KNL] 4111 Set wakeup interval for idle CPUs that have 4112 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y). 4113 Lazy RCU callbacks are those which RCU can 4114 prove do nothing more than free memory. 4115 4116 rcutree.rcu_kick_kthreads= [KNL] 4117 Cause the grace-period kthread to get an extra 4118 wake_up() if it sleeps three times longer than 4119 it should at force-quiescent-state time. 4120 This wake_up() will be accompanied by a 4121 WARN_ONCE() splat and an ftrace_dump(). 4122 4123 rcutree.sysrq_rcu= [KNL] 4124 Commandeer a sysrq key to dump out Tree RCU's 4125 rcu_node tree with an eye towards determining 4126 why a new grace period has not yet started. 4127 4128 rcuperf.gp_async= [KNL] 4129 Measure performance of asynchronous 4130 grace-period primitives such as call_rcu(). 4131 4132 rcuperf.gp_async_max= [KNL] 4133 Specify the maximum number of outstanding 4134 callbacks per writer thread. When a writer 4135 thread exceeds this limit, it invokes the 4136 corresponding flavor of rcu_barrier() to allow 4137 previously posted callbacks to drain. 4138 4139 rcuperf.gp_exp= [KNL] 4140 Measure performance of expedited synchronous 4141 grace-period primitives. 4142 4143 rcuperf.holdoff= [KNL] 4144 Set test-start holdoff period. The purpose of 4145 this parameter is to delay the start of the 4146 test until boot completes in order to avoid 4147 interference. 4148 4149 rcuperf.kfree_rcu_test= [KNL] 4150 Set to measure performance of kfree_rcu() flooding. 4151 4152 rcuperf.kfree_nthreads= [KNL] 4153 The number of threads running loops of kfree_rcu(). 4154 4155 rcuperf.kfree_alloc_num= [KNL] 4156 Number of allocations and frees done in an iteration. 4157 4158 rcuperf.kfree_loops= [KNL] 4159 Number of loops doing rcuperf.kfree_alloc_num number 4160 of allocations and frees. 4161 4162 rcuperf.nreaders= [KNL] 4163 Set number of RCU readers. The value -1 selects 4164 N, where N is the number of CPUs. A value 4165 "n" less than -1 selects N-n+1, where N is again 4166 the number of CPUs. For example, -2 selects N 4167 (the number of CPUs), -3 selects N+1, and so on. 4168 A value of "n" less than or equal to -N selects 4169 a single reader. 4170 4171 rcuperf.nwriters= [KNL] 4172 Set number of RCU writers. The values operate 4173 the same as for rcuperf.nreaders. 4174 N, where N is the number of CPUs 4175 4176 rcuperf.perf_type= [KNL] 4177 Specify the RCU implementation to test. 4178 4179 rcuperf.shutdown= [KNL] 4180 Shut the system down after performance tests 4181 complete. This is useful for hands-off automated 4182 testing. 4183 4184 rcuperf.verbose= [KNL] 4185 Enable additional printk() statements. 4186 4187 rcuperf.writer_holdoff= [KNL] 4188 Write-side holdoff between grace periods, 4189 in microseconds. The default of zero says 4190 no holdoff. 4191 4192 rcutorture.fqs_duration= [KNL] 4193 Set duration of force_quiescent_state bursts 4194 in microseconds. 4195 4196 rcutorture.fqs_holdoff= [KNL] 4197 Set holdoff time within force_quiescent_state bursts 4198 in microseconds. 4199 4200 rcutorture.fqs_stutter= [KNL] 4201 Set wait time between force_quiescent_state bursts 4202 in seconds. 4203 4204 rcutorture.fwd_progress= [KNL] 4205 Enable RCU grace-period forward-progress testing 4206 for the types of RCU supporting this notion. 4207 4208 rcutorture.fwd_progress_div= [KNL] 4209 Specify the fraction of a CPU-stall-warning 4210 period to do tight-loop forward-progress testing. 4211 4212 rcutorture.fwd_progress_holdoff= [KNL] 4213 Number of seconds to wait between successive 4214 forward-progress tests. 4215 4216 rcutorture.fwd_progress_need_resched= [KNL] 4217 Enclose cond_resched() calls within checks for 4218 need_resched() during tight-loop forward-progress 4219 testing. 4220 4221 rcutorture.gp_cond= [KNL] 4222 Use conditional/asynchronous update-side 4223 primitives, if available. 4224 4225 rcutorture.gp_exp= [KNL] 4226 Use expedited update-side primitives, if available. 4227 4228 rcutorture.gp_normal= [KNL] 4229 Use normal (non-expedited) asynchronous 4230 update-side primitives, if available. 4231 4232 rcutorture.gp_sync= [KNL] 4233 Use normal (non-expedited) synchronous 4234 update-side primitives, if available. If all 4235 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4236 rcutorture.gp_normal=, and rcutorture.gp_sync= 4237 are zero, rcutorture acts as if is interpreted 4238 they are all non-zero. 4239 4240 rcutorture.n_barrier_cbs= [KNL] 4241 Set callbacks/threads for rcu_barrier() testing. 4242 4243 rcutorture.nfakewriters= [KNL] 4244 Set number of concurrent RCU writers. These just 4245 stress RCU, they don't participate in the actual 4246 test, hence the "fake". 4247 4248 rcutorture.nreaders= [KNL] 4249 Set number of RCU readers. The value -1 selects 4250 N-1, where N is the number of CPUs. A value 4251 "n" less than -1 selects N-n-2, where N is again 4252 the number of CPUs. For example, -2 selects N 4253 (the number of CPUs), -3 selects N+1, and so on. 4254 4255 rcutorture.object_debug= [KNL] 4256 Enable debug-object double-call_rcu() testing. 4257 4258 rcutorture.onoff_holdoff= [KNL] 4259 Set time (s) after boot for CPU-hotplug testing. 4260 4261 rcutorture.onoff_interval= [KNL] 4262 Set time (jiffies) between CPU-hotplug operations, 4263 or zero to disable CPU-hotplug testing. 4264 4265 rcutorture.shuffle_interval= [KNL] 4266 Set task-shuffle interval (s). Shuffling tasks 4267 allows some CPUs to go into dyntick-idle mode 4268 during the rcutorture test. 4269 4270 rcutorture.shutdown_secs= [KNL] 4271 Set time (s) after boot system shutdown. This 4272 is useful for hands-off automated testing. 4273 4274 rcutorture.stall_cpu= [KNL] 4275 Duration of CPU stall (s) to test RCU CPU stall 4276 warnings, zero to disable. 4277 4278 rcutorture.stall_cpu_block= [KNL] 4279 Sleep while stalling if set. This will result 4280 in warnings from preemptible RCU in addition 4281 to any other stall-related activity. 4282 4283 rcutorture.stall_cpu_holdoff= [KNL] 4284 Time to wait (s) after boot before inducing stall. 4285 4286 rcutorture.stall_cpu_irqsoff= [KNL] 4287 Disable interrupts while stalling if set. 4288 4289 rcutorture.stall_gp_kthread= [KNL] 4290 Duration (s) of forced sleep within RCU 4291 grace-period kthread to test RCU CPU stall 4292 warnings, zero to disable. If both stall_cpu 4293 and stall_gp_kthread are specified, the 4294 kthread is starved first, then the CPU. 4295 4296 rcutorture.stat_interval= [KNL] 4297 Time (s) between statistics printk()s. 4298 4299 rcutorture.stutter= [KNL] 4300 Time (s) to stutter testing, for example, specifying 4301 five seconds causes the test to run for five seconds, 4302 wait for five seconds, and so on. This tests RCU's 4303 ability to transition abruptly to and from idle. 4304 4305 rcutorture.test_boost= [KNL] 4306 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4307 "Maybe" means test if the RCU implementation 4308 under test support RCU priority boosting. 4309 4310 rcutorture.test_boost_duration= [KNL] 4311 Duration (s) of each individual boost test. 4312 4313 rcutorture.test_boost_interval= [KNL] 4314 Interval (s) between each boost test. 4315 4316 rcutorture.test_no_idle_hz= [KNL] 4317 Test RCU's dyntick-idle handling. See also the 4318 rcutorture.shuffle_interval parameter. 4319 4320 rcutorture.torture_type= [KNL] 4321 Specify the RCU implementation to test. 4322 4323 rcutorture.verbose= [KNL] 4324 Enable additional printk() statements. 4325 4326 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4327 Dump ftrace buffer after reporting RCU CPU 4328 stall warning. 4329 4330 rcupdate.rcu_cpu_stall_suppress= [KNL] 4331 Suppress RCU CPU stall warning messages. 4332 4333 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 4334 Suppress RCU CPU stall warning messages and 4335 rcutorture writer stall warnings that occur 4336 during early boot, that is, during the time 4337 before the init task is spawned. 4338 4339 rcupdate.rcu_cpu_stall_timeout= [KNL] 4340 Set timeout for RCU CPU stall warning messages. 4341 4342 rcupdate.rcu_expedited= [KNL] 4343 Use expedited grace-period primitives, for 4344 example, synchronize_rcu_expedited() instead 4345 of synchronize_rcu(). This reduces latency, 4346 but can increase CPU utilization, degrade 4347 real-time latency, and degrade energy efficiency. 4348 No effect on CONFIG_TINY_RCU kernels. 4349 4350 rcupdate.rcu_normal= [KNL] 4351 Use only normal grace-period primitives, 4352 for example, synchronize_rcu() instead of 4353 synchronize_rcu_expedited(). This improves 4354 real-time latency, CPU utilization, and 4355 energy efficiency, but can expose users to 4356 increased grace-period latency. This parameter 4357 overrides rcupdate.rcu_expedited. No effect on 4358 CONFIG_TINY_RCU kernels. 4359 4360 rcupdate.rcu_normal_after_boot= [KNL] 4361 Once boot has completed (that is, after 4362 rcu_end_inkernel_boot() has been invoked), use 4363 only normal grace-period primitives. No effect 4364 on CONFIG_TINY_RCU kernels. 4365 4366 rcupdate.rcu_task_ipi_delay= [KNL] 4367 Set time in jiffies during which RCU tasks will 4368 avoid sending IPIs, starting with the beginning 4369 of a given grace period. Setting a large 4370 number avoids disturbing real-time workloads, 4371 but lengthens grace periods. 4372 4373 rcupdate.rcu_task_stall_timeout= [KNL] 4374 Set timeout in jiffies for RCU task stall warning 4375 messages. Disable with a value less than or equal 4376 to zero. 4377 4378 rcupdate.rcu_self_test= [KNL] 4379 Run the RCU early boot self tests 4380 4381 rdinit= [KNL] 4382 Format: <full_path> 4383 Run specified binary instead of /init from the ramdisk, 4384 used for early userspace startup. See initrd. 4385 4386 rdrand= [X86] 4387 force - Override the decision by the kernel to hide the 4388 advertisement of RDRAND support (this affects 4389 certain AMD processors because of buggy BIOS 4390 support, specifically around the suspend/resume 4391 path). 4392 4393 rdt= [HW,X86,RDT] 4394 Turn on/off individual RDT features. List is: 4395 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4396 mba. 4397 E.g. to turn on cmt and turn off mba use: 4398 rdt=cmt,!mba 4399 4400 reboot= [KNL] 4401 Format (x86 or x86_64): 4402 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4403 [[,]s[mp]#### \ 4404 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4405 [[,]f[orce] 4406 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4407 (prefix with 'panic_' to set mode for panic 4408 reboot only), 4409 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4410 reboot_force is either force or not specified, 4411 reboot_cpu is s[mp]#### with #### being the processor 4412 to be used for rebooting. 4413 4414 relax_domain_level= 4415 [KNL, SMP] Set scheduler's default relax_domain_level. 4416 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4417 4418 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4419 Format: <base1>,<size1>[,<base2>,<size2>,...] 4420 Reserve I/O ports or memory so the kernel won't use 4421 them. If <base> is less than 0x10000, the region 4422 is assumed to be I/O ports; otherwise it is memory. 4423 4424 reservetop= [X86-32] 4425 Format: nn[KMG] 4426 Reserves a hole at the top of the kernel virtual 4427 address space. 4428 4429 reservelow= [X86] 4430 Format: nn[K] 4431 Set the amount of memory to reserve for BIOS at 4432 the bottom of the address space. 4433 4434 reset_devices [KNL] Force drivers to reset the underlying device 4435 during initialization. 4436 4437 resume= [SWSUSP] 4438 Specify the partition device for software suspend 4439 Format: 4440 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4441 4442 resume_offset= [SWSUSP] 4443 Specify the offset from the beginning of the partition 4444 given by "resume=" at which the swap header is located, 4445 in <PAGE_SIZE> units (needed only for swap files). 4446 See Documentation/power/swsusp-and-swap-files.rst 4447 4448 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4449 read the resume files 4450 4451 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4452 Useful for devices that are detected asynchronously 4453 (e.g. USB and MMC devices). 4454 4455 hibernate= [HIBERNATION] 4456 noresume Don't check if there's a hibernation image 4457 present during boot. 4458 nocompress Don't compress/decompress hibernation images. 4459 no Disable hibernation and resume. 4460 protect_image Turn on image protection during restoration 4461 (that will set all pages holding image data 4462 during restoration read-only). 4463 4464 retain_initrd [RAM] Keep initrd memory after extraction 4465 4466 rfkill.default_state= 4467 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4468 etc. communication is blocked by default. 4469 1 Unblocked. 4470 4471 rfkill.master_switch_mode= 4472 0 The "airplane mode" button does nothing. 4473 1 The "airplane mode" button toggles between everything 4474 blocked and the previous configuration. 4475 2 The "airplane mode" button toggles between everything 4476 blocked and everything unblocked. 4477 4478 rhash_entries= [KNL,NET] 4479 Set number of hash buckets for route cache 4480 4481 ring3mwait=disable 4482 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4483 CPUs. 4484 4485 ro [KNL] Mount root device read-only on boot 4486 4487 rodata= [KNL] 4488 on Mark read-only kernel memory as read-only (default). 4489 off Leave read-only kernel memory writable for debugging. 4490 4491 rockchip.usb_uart 4492 Enable the uart passthrough on the designated usb port 4493 on Rockchip SoCs. When active, the signals of the 4494 debug-uart get routed to the D+ and D- pins of the usb 4495 port and the regular usb controller gets disabled. 4496 4497 root= [KNL] Root filesystem 4498 See name_to_dev_t comment in init/do_mounts.c. 4499 4500 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4501 mount the root filesystem 4502 4503 rootflags= [KNL] Set root filesystem mount option string 4504 4505 rootfstype= [KNL] Set root filesystem type 4506 4507 rootwait [KNL] Wait (indefinitely) for root device to show up. 4508 Useful for devices that are detected asynchronously 4509 (e.g. USB and MMC devices). 4510 4511 rproc_mem=nn[KMG][@address] 4512 [KNL,ARM,CMA] Remoteproc physical memory block. 4513 Memory area to be used by remote processor image, 4514 managed by CMA. 4515 4516 rw [KNL] Mount root device read-write on boot 4517 4518 S [KNL] Run init in single mode 4519 4520 s390_iommu= [HW,S390] 4521 Set s390 IOTLB flushing mode 4522 strict 4523 With strict flushing every unmap operation will result in 4524 an IOTLB flush. Default is lazy flushing before reuse, 4525 which is faster. 4526 4527 sa1100ir [NET] 4528 See drivers/net/irda/sa1100_ir.c. 4529 4530 sbni= [NET] Granch SBNI12 leased line adapter 4531 4532 sched_debug [KNL] Enables verbose scheduler debug messages. 4533 4534 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4535 Allowed values are enable and disable. This feature 4536 incurs a small amount of overhead in the scheduler 4537 but is useful for debugging and performance tuning. 4538 4539 sched_thermal_decay_shift= 4540 [KNL, SMP] Set a decay shift for scheduler thermal 4541 pressure signal. Thermal pressure signal follows the 4542 default decay period of other scheduler pelt 4543 signals(usually 32 ms but configurable). Setting 4544 sched_thermal_decay_shift will left shift the decay 4545 period for the thermal pressure signal by the shift 4546 value. 4547 i.e. with the default pelt decay period of 32 ms 4548 sched_thermal_decay_shift thermal pressure decay pr 4549 1 64 ms 4550 2 128 ms 4551 and so on. 4552 Format: integer between 0 and 10 4553 Default is 0. 4554 4555 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 4556 xtime_lock contention on larger systems, and/or RCU lock 4557 contention on all systems with CONFIG_MAXSMP set. 4558 Format: { "0" | "1" } 4559 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 4560 1 -- enable. 4561 Note: increases power consumption, thus should only be 4562 enabled if running jitter sensitive (HPC/RT) workloads. 4563 4564 security= [SECURITY] Choose a legacy "major" security module to 4565 enable at boot. This has been deprecated by the 4566 "lsm=" parameter. 4567 4568 selinux= [SELINUX] Disable or enable SELinux at boot time. 4569 Format: { "0" | "1" } 4570 See security/selinux/Kconfig help text. 4571 0 -- disable. 4572 1 -- enable. 4573 Default value is 1. 4574 4575 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 4576 Format: { "0" | "1" } 4577 See security/apparmor/Kconfig help text 4578 0 -- disable. 4579 1 -- enable. 4580 Default value is set via kernel config option. 4581 4582 serialnumber [BUGS=X86-32] 4583 4584 shapers= [NET] 4585 Maximal number of shapers. 4586 4587 simeth= [IA-64] 4588 simscsi= 4589 4590 slram= [HW,MTD] 4591 4592 slab_nomerge [MM] 4593 Disable merging of slabs with similar size. May be 4594 necessary if there is some reason to distinguish 4595 allocs to different slabs, especially in hardened 4596 environments where the risk of heap overflows and 4597 layout control by attackers can usually be 4598 frustrated by disabling merging. This will reduce 4599 most of the exposure of a heap attack to a single 4600 cache (risks via metadata attacks are mostly 4601 unchanged). Debug options disable merging on their 4602 own. 4603 For more information see Documentation/vm/slub.rst. 4604 4605 slab_max_order= [MM, SLAB] 4606 Determines the maximum allowed order for slabs. 4607 A high setting may cause OOMs due to memory 4608 fragmentation. Defaults to 1 for systems with 4609 more than 32MB of RAM, 0 otherwise. 4610 4611 slub_debug[=options[,slabs]] [MM, SLUB] 4612 Enabling slub_debug allows one to determine the 4613 culprit if slab objects become corrupted. Enabling 4614 slub_debug can create guard zones around objects and 4615 may poison objects when not in use. Also tracks the 4616 last alloc / free. For more information see 4617 Documentation/vm/slub.rst. 4618 4619 slub_memcg_sysfs= [MM, SLUB] 4620 Determines whether to enable sysfs directories for 4621 memory cgroup sub-caches. 1 to enable, 0 to disable. 4622 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON. 4623 Enabling this can lead to a very high number of debug 4624 directories and files being created under 4625 /sys/kernel/slub. 4626 4627 slub_max_order= [MM, SLUB] 4628 Determines the maximum allowed order for slabs. 4629 A high setting may cause OOMs due to memory 4630 fragmentation. For more information see 4631 Documentation/vm/slub.rst. 4632 4633 slub_min_objects= [MM, SLUB] 4634 The minimum number of objects per slab. SLUB will 4635 increase the slab order up to slub_max_order to 4636 generate a sufficiently large slab able to contain 4637 the number of objects indicated. The higher the number 4638 of objects the smaller the overhead of tracking slabs 4639 and the less frequently locks need to be acquired. 4640 For more information see Documentation/vm/slub.rst. 4641 4642 slub_min_order= [MM, SLUB] 4643 Determines the minimum page order for slabs. Must be 4644 lower than slub_max_order. 4645 For more information see Documentation/vm/slub.rst. 4646 4647 slub_nomerge [MM, SLUB] 4648 Same with slab_nomerge. This is supported for legacy. 4649 See slab_nomerge for more information. 4650 4651 smart2= [HW] 4652 Format: <io1>[,<io2>[,...,<io8>]] 4653 4654 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 4655 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 4656 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 4657 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 4658 smsc-ircc2.ircc_irq= [HW] IRQ line 4659 smsc-ircc2.ircc_dma= [HW] DMA channel 4660 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 4661 0: Toshiba Satellite 1800 (GP data pin select) 4662 1: Fast pin select (default) 4663 2: ATC IRMode 4664 4665 smt [KNL,S390] Set the maximum number of threads (logical 4666 CPUs) to use per physical CPU on systems capable of 4667 symmetric multithreading (SMT). Will be capped to the 4668 actual hardware limit. 4669 Format: <integer> 4670 Default: -1 (no limit) 4671 4672 softlockup_panic= 4673 [KNL] Should the soft-lockup detector generate panics. 4674 Format: 0 | 1 4675 4676 A value of 1 instructs the soft-lockup detector 4677 to panic the machine when a soft-lockup occurs. It is 4678 also controlled by the kernel.softlockup_panic sysctl 4679 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 4680 respective build-time switch to that functionality. 4681 4682 softlockup_all_cpu_backtrace= 4683 [KNL] Should the soft-lockup detector generate 4684 backtraces on all cpus. 4685 Format: 0 | 1 4686 4687 sonypi.*= [HW] Sony Programmable I/O Control Device driver 4688 See Documentation/admin-guide/laptops/sonypi.rst 4689 4690 spectre_v2= [X86] Control mitigation of Spectre variant 2 4691 (indirect branch speculation) vulnerability. 4692 The default operation protects the kernel from 4693 user space attacks. 4694 4695 on - unconditionally enable, implies 4696 spectre_v2_user=on 4697 off - unconditionally disable, implies 4698 spectre_v2_user=off 4699 auto - kernel detects whether your CPU model is 4700 vulnerable 4701 4702 Selecting 'on' will, and 'auto' may, choose a 4703 mitigation method at run time according to the 4704 CPU, the available microcode, the setting of the 4705 CONFIG_RETPOLINE configuration option, and the 4706 compiler with which the kernel was built. 4707 4708 Selecting 'on' will also enable the mitigation 4709 against user space to user space task attacks. 4710 4711 Selecting 'off' will disable both the kernel and 4712 the user space protections. 4713 4714 Specific mitigations can also be selected manually: 4715 4716 retpoline - replace indirect branches 4717 retpoline,generic - google's original retpoline 4718 retpoline,amd - AMD-specific minimal thunk 4719 4720 Not specifying this option is equivalent to 4721 spectre_v2=auto. 4722 4723 spectre_v2_user= 4724 [X86] Control mitigation of Spectre variant 2 4725 (indirect branch speculation) vulnerability between 4726 user space tasks 4727 4728 on - Unconditionally enable mitigations. Is 4729 enforced by spectre_v2=on 4730 4731 off - Unconditionally disable mitigations. Is 4732 enforced by spectre_v2=off 4733 4734 prctl - Indirect branch speculation is enabled, 4735 but mitigation can be enabled via prctl 4736 per thread. The mitigation control state 4737 is inherited on fork. 4738 4739 prctl,ibpb 4740 - Like "prctl" above, but only STIBP is 4741 controlled per thread. IBPB is issued 4742 always when switching between different user 4743 space processes. 4744 4745 seccomp 4746 - Same as "prctl" above, but all seccomp 4747 threads will enable the mitigation unless 4748 they explicitly opt out. 4749 4750 seccomp,ibpb 4751 - Like "seccomp" above, but only STIBP is 4752 controlled per thread. IBPB is issued 4753 always when switching between different 4754 user space processes. 4755 4756 auto - Kernel selects the mitigation depending on 4757 the available CPU features and vulnerability. 4758 4759 Default mitigation: 4760 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 4761 4762 Not specifying this option is equivalent to 4763 spectre_v2_user=auto. 4764 4765 spec_store_bypass_disable= 4766 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 4767 (Speculative Store Bypass vulnerability) 4768 4769 Certain CPUs are vulnerable to an exploit against a 4770 a common industry wide performance optimization known 4771 as "Speculative Store Bypass" in which recent stores 4772 to the same memory location may not be observed by 4773 later loads during speculative execution. The idea 4774 is that such stores are unlikely and that they can 4775 be detected prior to instruction retirement at the 4776 end of a particular speculation execution window. 4777 4778 In vulnerable processors, the speculatively forwarded 4779 store can be used in a cache side channel attack, for 4780 example to read memory to which the attacker does not 4781 directly have access (e.g. inside sandboxed code). 4782 4783 This parameter controls whether the Speculative Store 4784 Bypass optimization is used. 4785 4786 On x86 the options are: 4787 4788 on - Unconditionally disable Speculative Store Bypass 4789 off - Unconditionally enable Speculative Store Bypass 4790 auto - Kernel detects whether the CPU model contains an 4791 implementation of Speculative Store Bypass and 4792 picks the most appropriate mitigation. If the 4793 CPU is not vulnerable, "off" is selected. If the 4794 CPU is vulnerable the default mitigation is 4795 architecture and Kconfig dependent. See below. 4796 prctl - Control Speculative Store Bypass per thread 4797 via prctl. Speculative Store Bypass is enabled 4798 for a process by default. The state of the control 4799 is inherited on fork. 4800 seccomp - Same as "prctl" above, but all seccomp threads 4801 will disable SSB unless they explicitly opt out. 4802 4803 Default mitigations: 4804 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 4805 4806 On powerpc the options are: 4807 4808 on,auto - On Power8 and Power9 insert a store-forwarding 4809 barrier on kernel entry and exit. On Power7 4810 perform a software flush on kernel entry and 4811 exit. 4812 off - No action. 4813 4814 Not specifying this option is equivalent to 4815 spec_store_bypass_disable=auto. 4816 4817 spia_io_base= [HW,MTD] 4818 spia_fio_base= 4819 spia_pedr= 4820 spia_peddr= 4821 4822 split_lock_detect= 4823 [X86] Enable split lock detection 4824 4825 When enabled (and if hardware support is present), atomic 4826 instructions that access data across cache line 4827 boundaries will result in an alignment check exception. 4828 4829 off - not enabled 4830 4831 warn - the kernel will emit rate limited warnings 4832 about applications triggering the #AC 4833 exception. This mode is the default on CPUs 4834 that supports split lock detection. 4835 4836 fatal - the kernel will send SIGBUS to applications 4837 that trigger the #AC exception. 4838 4839 If an #AC exception is hit in the kernel or in 4840 firmware (i.e. not while executing in user mode) 4841 the kernel will oops in either "warn" or "fatal" 4842 mode. 4843 4844 srbds= [X86,INTEL] 4845 Control the Special Register Buffer Data Sampling 4846 (SRBDS) mitigation. 4847 4848 Certain CPUs are vulnerable to an MDS-like 4849 exploit which can leak bits from the random 4850 number generator. 4851 4852 By default, this issue is mitigated by 4853 microcode. However, the microcode fix can cause 4854 the RDRAND and RDSEED instructions to become 4855 much slower. Among other effects, this will 4856 result in reduced throughput from /dev/urandom. 4857 4858 The microcode mitigation can be disabled with 4859 the following option: 4860 4861 off: Disable mitigation and remove 4862 performance impact to RDRAND and RDSEED 4863 4864 srcutree.counter_wrap_check [KNL] 4865 Specifies how frequently to check for 4866 grace-period sequence counter wrap for the 4867 srcu_data structure's ->srcu_gp_seq_needed field. 4868 The greater the number of bits set in this kernel 4869 parameter, the less frequently counter wrap will 4870 be checked for. Note that the bottom two bits 4871 are ignored. 4872 4873 srcutree.exp_holdoff [KNL] 4874 Specifies how many nanoseconds must elapse 4875 since the end of the last SRCU grace period for 4876 a given srcu_struct until the next normal SRCU 4877 grace period will be considered for automatic 4878 expediting. Set to zero to disable automatic 4879 expediting. 4880 4881 ssbd= [ARM64,HW] 4882 Speculative Store Bypass Disable control 4883 4884 On CPUs that are vulnerable to the Speculative 4885 Store Bypass vulnerability and offer a 4886 firmware based mitigation, this parameter 4887 indicates how the mitigation should be used: 4888 4889 force-on: Unconditionally enable mitigation for 4890 for both kernel and userspace 4891 force-off: Unconditionally disable mitigation for 4892 for both kernel and userspace 4893 kernel: Always enable mitigation in the 4894 kernel, and offer a prctl interface 4895 to allow userspace to register its 4896 interest in being mitigated too. 4897 4898 stack_guard_gap= [MM] 4899 override the default stack gap protection. The value 4900 is in page units and it defines how many pages prior 4901 to (for stacks growing down) resp. after (for stacks 4902 growing up) the main stack are reserved for no other 4903 mapping. Default value is 256 pages. 4904 4905 stacktrace [FTRACE] 4906 Enabled the stack tracer on boot up. 4907 4908 stacktrace_filter=[function-list] 4909 [FTRACE] Limit the functions that the stack tracer 4910 will trace at boot up. function-list is a comma separated 4911 list of functions. This list can be changed at run 4912 time by the stack_trace_filter file in the debugfs 4913 tracing directory. Note, this enables stack tracing 4914 and the stacktrace above is not needed. 4915 4916 sti= [PARISC,HW] 4917 Format: <num> 4918 Set the STI (builtin display/keyboard on the HP-PARISC 4919 machines) console (graphic card) which should be used 4920 as the initial boot-console. 4921 See also comment in drivers/video/console/sticore.c. 4922 4923 sti_font= [HW] 4924 See comment in drivers/video/console/sticore.c. 4925 4926 stifb= [HW] 4927 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 4928 4929 sunrpc.min_resvport= 4930 sunrpc.max_resvport= 4931 [NFS,SUNRPC] 4932 SunRPC servers often require that client requests 4933 originate from a privileged port (i.e. a port in the 4934 range 0 < portnr < 1024). 4935 An administrator who wishes to reserve some of these 4936 ports for other uses may adjust the range that the 4937 kernel's sunrpc client considers to be privileged 4938 using these two parameters to set the minimum and 4939 maximum port values. 4940 4941 sunrpc.svc_rpc_per_connection_limit= 4942 [NFS,SUNRPC] 4943 Limit the number of requests that the server will 4944 process in parallel from a single connection. 4945 The default value is 0 (no limit). 4946 4947 sunrpc.pool_mode= 4948 [NFS] 4949 Control how the NFS server code allocates CPUs to 4950 service thread pools. Depending on how many NICs 4951 you have and where their interrupts are bound, this 4952 option will affect which CPUs will do NFS serving. 4953 Note: this parameter cannot be changed while the 4954 NFS server is running. 4955 4956 auto the server chooses an appropriate mode 4957 automatically using heuristics 4958 global a single global pool contains all CPUs 4959 percpu one pool for each CPU 4960 pernode one pool for each NUMA node (equivalent 4961 to global on non-NUMA machines) 4962 4963 sunrpc.tcp_slot_table_entries= 4964 sunrpc.udp_slot_table_entries= 4965 [NFS,SUNRPC] 4966 Sets the upper limit on the number of simultaneous 4967 RPC calls that can be sent from the client to a 4968 server. Increasing these values may allow you to 4969 improve throughput, but will also increase the 4970 amount of memory reserved for use by the client. 4971 4972 suspend.pm_test_delay= 4973 [SUSPEND] 4974 Sets the number of seconds to remain in a suspend test 4975 mode before resuming the system (see 4976 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 4977 is set. Default value is 5. 4978 4979 svm= [PPC] 4980 Format: { on | off | y | n | 1 | 0 } 4981 This parameter controls use of the Protected 4982 Execution Facility on pSeries. 4983 4984 swapaccount=[0|1] 4985 [KNL] Enable accounting of swap in memory resource 4986 controller if no parameter or 1 is given or disable 4987 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 4988 4989 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 4990 Format: { <int> | force | noforce } 4991 <int> -- Number of I/O TLB slabs 4992 force -- force using of bounce buffers even if they 4993 wouldn't be automatically used by the kernel 4994 noforce -- Never use bounce buffers (for debugging) 4995 4996 switches= [HW,M68k] 4997 4998 sysctl.*= [KNL] 4999 Set a sysctl parameter, right before loading the init 5000 process, as if the value was written to the respective 5001 /proc/sys/... file. Both '.' and '/' are recognized as 5002 separators. Unrecognized parameters and invalid values 5003 are reported in the kernel log. Sysctls registered 5004 later by a loaded module cannot be set this way. 5005 Example: sysctl.vm.swappiness=40 5006 5007 sysfs.deprecated=0|1 [KNL] 5008 Enable/disable old style sysfs layout for old udev 5009 on older distributions. When this option is enabled 5010 very new udev will not work anymore. When this option 5011 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 5012 in older udev will not work anymore. 5013 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 5014 the kernel configuration. 5015 5016 sysrq_always_enabled 5017 [KNL] 5018 Ignore sysrq setting - this boot parameter will 5019 neutralize any effect of /proc/sys/kernel/sysrq. 5020 Useful for debugging. 5021 5022 tcpmhash_entries= [KNL,NET] 5023 Set the number of tcp_metrics_hash slots. 5024 Default value is 8192 or 16384 depending on total 5025 ram pages. This is used to specify the TCP metrics 5026 cache size. See Documentation/networking/ip-sysctl.rst 5027 "tcp_no_metrics_save" section for more details. 5028 5029 tdfx= [HW,DRM] 5030 5031 test_suspend= [SUSPEND][,N] 5032 Specify "mem" (for Suspend-to-RAM) or "standby" (for 5033 standby suspend) or "freeze" (for suspend type freeze) 5034 as the system sleep state during system startup with 5035 the optional capability to repeat N number of times. 5036 The system is woken from this state using a 5037 wakeup-capable RTC alarm. 5038 5039 thash_entries= [KNL,NET] 5040 Set number of hash buckets for TCP connection 5041 5042 thermal.act= [HW,ACPI] 5043 -1: disable all active trip points in all thermal zones 5044 <degrees C>: override all lowest active trip points 5045 5046 thermal.crt= [HW,ACPI] 5047 -1: disable all critical trip points in all thermal zones 5048 <degrees C>: override all critical trip points 5049 5050 thermal.nocrt= [HW,ACPI] 5051 Set to disable actions on ACPI thermal zone 5052 critical and hot trip points. 5053 5054 thermal.off= [HW,ACPI] 5055 1: disable ACPI thermal control 5056 5057 thermal.psv= [HW,ACPI] 5058 -1: disable all passive trip points 5059 <degrees C>: override all passive trip points to this 5060 value 5061 5062 thermal.tzp= [HW,ACPI] 5063 Specify global default ACPI thermal zone polling rate 5064 <deci-seconds>: poll all this frequency 5065 0: no polling (default) 5066 5067 threadirqs [KNL] 5068 Force threading of all interrupt handlers except those 5069 marked explicitly IRQF_NO_THREAD. 5070 5071 topology= [S390] 5072 Format: {off | on} 5073 Specify if the kernel should make use of the cpu 5074 topology information if the hardware supports this. 5075 The scheduler will make use of this information and 5076 e.g. base its process migration decisions on it. 5077 Default is on. 5078 5079 topology_updates= [KNL, PPC, NUMA] 5080 Format: {off} 5081 Specify if the kernel should ignore (off) 5082 topology updates sent by the hypervisor to this 5083 LPAR. 5084 5085 torture.disable_onoff_at_boot= [KNL] 5086 Prevent the CPU-hotplug component of torturing 5087 until after init has spawned. 5088 5089 tp720= [HW,PS2] 5090 5091 tpm_suspend_pcr=[HW,TPM] 5092 Format: integer pcr id 5093 Specify that at suspend time, the tpm driver 5094 should extend the specified pcr with zeros, 5095 as a workaround for some chips which fail to 5096 flush the last written pcr on TPM_SaveState. 5097 This will guarantee that all the other pcrs 5098 are saved. 5099 5100 trace_buf_size=nn[KMG] 5101 [FTRACE] will set tracing buffer size on each cpu. 5102 5103 trace_event=[event-list] 5104 [FTRACE] Set and start specified trace events in order 5105 to facilitate early boot debugging. The event-list is a 5106 comma separated list of trace events to enable. See 5107 also Documentation/trace/events.rst 5108 5109 trace_options=[option-list] 5110 [FTRACE] Enable or disable tracer options at boot. 5111 The option-list is a comma delimited list of options 5112 that can be enabled or disabled just as if you were 5113 to echo the option name into 5114 5115 /sys/kernel/debug/tracing/trace_options 5116 5117 For example, to enable stacktrace option (to dump the 5118 stack trace of each event), add to the command line: 5119 5120 trace_options=stacktrace 5121 5122 See also Documentation/trace/ftrace.rst "trace options" 5123 section. 5124 5125 tp_printk[FTRACE] 5126 Have the tracepoints sent to printk as well as the 5127 tracing ring buffer. This is useful for early boot up 5128 where the system hangs or reboots and does not give the 5129 option for reading the tracing buffer or performing a 5130 ftrace_dump_on_oops. 5131 5132 To turn off having tracepoints sent to printk, 5133 echo 0 > /proc/sys/kernel/tracepoint_printk 5134 Note, echoing 1 into this file without the 5135 tracepoint_printk kernel cmdline option has no effect. 5136 5137 ** CAUTION ** 5138 5139 Having tracepoints sent to printk() and activating high 5140 frequency tracepoints such as irq or sched, can cause 5141 the system to live lock. 5142 5143 traceoff_on_warning 5144 [FTRACE] enable this option to disable tracing when a 5145 warning is hit. This turns off "tracing_on". Tracing can 5146 be enabled again by echoing '1' into the "tracing_on" 5147 file located in /sys/kernel/debug/tracing/ 5148 5149 This option is useful, as it disables the trace before 5150 the WARNING dump is called, which prevents the trace to 5151 be filled with content caused by the warning output. 5152 5153 This option can also be set at run time via the sysctl 5154 option: kernel/traceoff_on_warning 5155 5156 transparent_hugepage= 5157 [KNL] 5158 Format: [always|madvise|never] 5159 Can be used to control the default behavior of the system 5160 with respect to transparent hugepages. 5161 See Documentation/admin-guide/mm/transhuge.rst 5162 for more details. 5163 5164 tsc= Disable clocksource stability checks for TSC. 5165 Format: <string> 5166 [x86] reliable: mark tsc clocksource as reliable, this 5167 disables clocksource verification at runtime, as well 5168 as the stability checks done at bootup. Used to enable 5169 high-resolution timer mode on older hardware, and in 5170 virtualized environment. 5171 [x86] noirqtime: Do not use TSC to do irq accounting. 5172 Used to run time disable IRQ_TIME_ACCOUNTING on any 5173 platforms where RDTSC is slow and this accounting 5174 can add overhead. 5175 [x86] unstable: mark the TSC clocksource as unstable, this 5176 marks the TSC unconditionally unstable at bootup and 5177 avoids any further wobbles once the TSC watchdog notices. 5178 [x86] nowatchdog: disable clocksource watchdog. Used 5179 in situations with strict latency requirements (where 5180 interruptions from clocksource watchdog are not 5181 acceptable). 5182 5183 tsc_early_khz= [X86] Skip early TSC calibration and use the given 5184 value instead. Useful when the early TSC frequency discovery 5185 procedure is not reliable, such as on overclocked systems 5186 with CPUID.16h support and partial CPUID.15h support. 5187 Format: <unsigned int> 5188 5189 tsx= [X86] Control Transactional Synchronization 5190 Extensions (TSX) feature in Intel processors that 5191 support TSX control. 5192 5193 This parameter controls the TSX feature. The options are: 5194 5195 on - Enable TSX on the system. Although there are 5196 mitigations for all known security vulnerabilities, 5197 TSX has been known to be an accelerator for 5198 several previous speculation-related CVEs, and 5199 so there may be unknown security risks associated 5200 with leaving it enabled. 5201 5202 off - Disable TSX on the system. (Note that this 5203 option takes effect only on newer CPUs which are 5204 not vulnerable to MDS, i.e., have 5205 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 5206 the new IA32_TSX_CTRL MSR through a microcode 5207 update. This new MSR allows for the reliable 5208 deactivation of the TSX functionality.) 5209 5210 auto - Disable TSX if X86_BUG_TAA is present, 5211 otherwise enable TSX on the system. 5212 5213 Not specifying this option is equivalent to tsx=off. 5214 5215 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5216 for more details. 5217 5218 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 5219 Abort (TAA) vulnerability. 5220 5221 Similar to Micro-architectural Data Sampling (MDS) 5222 certain CPUs that support Transactional 5223 Synchronization Extensions (TSX) are vulnerable to an 5224 exploit against CPU internal buffers which can forward 5225 information to a disclosure gadget under certain 5226 conditions. 5227 5228 In vulnerable processors, the speculatively forwarded 5229 data can be used in a cache side channel attack, to 5230 access data to which the attacker does not have direct 5231 access. 5232 5233 This parameter controls the TAA mitigation. The 5234 options are: 5235 5236 full - Enable TAA mitigation on vulnerable CPUs 5237 if TSX is enabled. 5238 5239 full,nosmt - Enable TAA mitigation and disable SMT on 5240 vulnerable CPUs. If TSX is disabled, SMT 5241 is not disabled because CPU is not 5242 vulnerable to cross-thread TAA attacks. 5243 off - Unconditionally disable TAA mitigation 5244 5245 On MDS-affected machines, tsx_async_abort=off can be 5246 prevented by an active MDS mitigation as both vulnerabilities 5247 are mitigated with the same mechanism so in order to disable 5248 this mitigation, you need to specify mds=off too. 5249 5250 Not specifying this option is equivalent to 5251 tsx_async_abort=full. On CPUs which are MDS affected 5252 and deploy MDS mitigation, TAA mitigation is not 5253 required and doesn't provide any additional 5254 mitigation. 5255 5256 For details see: 5257 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5258 5259 turbografx.map[2|3]= [HW,JOY] 5260 TurboGraFX parallel port interface 5261 Format: 5262 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 5263 See also Documentation/input/devices/joystick-parport.rst 5264 5265 udbg-immortal [PPC] When debugging early kernel crashes that 5266 happen after console_init() and before a proper 5267 console driver takes over, this boot options might 5268 help "seeing" what's going on. 5269 5270 uhash_entries= [KNL,NET] 5271 Set number of hash buckets for UDP/UDP-Lite connections 5272 5273 uhci-hcd.ignore_oc= 5274 [USB] Ignore overcurrent events (default N). 5275 Some badly-designed motherboards generate lots of 5276 bogus events, for ports that aren't wired to 5277 anything. Set this parameter to avoid log spamming. 5278 Note that genuine overcurrent events won't be 5279 reported either. 5280 5281 unknown_nmi_panic 5282 [X86] Cause panic on unknown NMI. 5283 5284 usbcore.authorized_default= 5285 [USB] Default USB device authorization: 5286 (default -1 = authorized except for wireless USB, 5287 0 = not authorized, 1 = authorized, 2 = authorized 5288 if device connected to internal port) 5289 5290 usbcore.autosuspend= 5291 [USB] The autosuspend time delay (in seconds) used 5292 for newly-detected USB devices (default 2). This 5293 is the time required before an idle device will be 5294 autosuspended. Devices for which the delay is set 5295 to a negative value won't be autosuspended at all. 5296 5297 usbcore.usbfs_snoop= 5298 [USB] Set to log all usbfs traffic (default 0 = off). 5299 5300 usbcore.usbfs_snoop_max= 5301 [USB] Maximum number of bytes to snoop in each URB 5302 (default = 65536). 5303 5304 usbcore.blinkenlights= 5305 [USB] Set to cycle leds on hubs (default 0 = off). 5306 5307 usbcore.old_scheme_first= 5308 [USB] Start with the old device initialization 5309 scheme (default 0 = off). 5310 5311 usbcore.usbfs_memory_mb= 5312 [USB] Memory limit (in MB) for buffers allocated by 5313 usbfs (default = 16, 0 = max = 2047). 5314 5315 usbcore.use_both_schemes= 5316 [USB] Try the other device initialization scheme 5317 if the first one fails (default 1 = enabled). 5318 5319 usbcore.initial_descriptor_timeout= 5320 [USB] Specifies timeout for the initial 64-byte 5321 USB_REQ_GET_DESCRIPTOR request in milliseconds 5322 (default 5000 = 5.0 seconds). 5323 5324 usbcore.nousb [USB] Disable the USB subsystem 5325 5326 usbcore.quirks= 5327 [USB] A list of quirk entries to augment the built-in 5328 usb core quirk list. List entries are separated by 5329 commas. Each entry has the form 5330 VendorID:ProductID:Flags. The IDs are 4-digit hex 5331 numbers and Flags is a set of letters. Each letter 5332 will change the built-in quirk; setting it if it is 5333 clear and clearing it if it is set. The letters have 5334 the following meanings: 5335 a = USB_QUIRK_STRING_FETCH_255 (string 5336 descriptors must not be fetched using 5337 a 255-byte read); 5338 b = USB_QUIRK_RESET_RESUME (device can't resume 5339 correctly so reset it instead); 5340 c = USB_QUIRK_NO_SET_INTF (device can't handle 5341 Set-Interface requests); 5342 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5343 handle its Configuration or Interface 5344 strings); 5345 e = USB_QUIRK_RESET (device can't be reset 5346 (e.g morph devices), don't use reset); 5347 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5348 more interface descriptions than the 5349 bNumInterfaces count, and can't handle 5350 talking to these interfaces); 5351 g = USB_QUIRK_DELAY_INIT (device needs a pause 5352 during initialization, after we read 5353 the device descriptor); 5354 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5355 high speed and super speed interrupt 5356 endpoints, the USB 2.0 and USB 3.0 spec 5357 require the interval in microframes (1 5358 microframe = 125 microseconds) to be 5359 calculated as interval = 2 ^ 5360 (bInterval-1). 5361 Devices with this quirk report their 5362 bInterval as the result of this 5363 calculation instead of the exponent 5364 variable used in the calculation); 5365 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5366 handle device_qualifier descriptor 5367 requests); 5368 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5369 generates spurious wakeup, ignore 5370 remote wakeup capability); 5371 k = USB_QUIRK_NO_LPM (device can't handle Link 5372 Power Management); 5373 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5374 (Device reports its bInterval as linear 5375 frames instead of the USB 2.0 5376 calculation); 5377 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5378 to be disconnected before suspend to 5379 prevent spurious wakeup); 5380 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5381 pause after every control message); 5382 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5383 delay after resetting its port); 5384 Example: quirks=0781:5580:bk,0a5c:5834:gij 5385 5386 usbhid.mousepoll= 5387 [USBHID] The interval which mice are to be polled at. 5388 5389 usbhid.jspoll= 5390 [USBHID] The interval which joysticks are to be polled at. 5391 5392 usbhid.kbpoll= 5393 [USBHID] The interval which keyboards are to be polled at. 5394 5395 usb-storage.delay_use= 5396 [UMS] The delay in seconds before a new device is 5397 scanned for Logical Units (default 1). 5398 5399 usb-storage.quirks= 5400 [UMS] A list of quirks entries to supplement or 5401 override the built-in unusual_devs list. List 5402 entries are separated by commas. Each entry has 5403 the form VID:PID:Flags where VID and PID are Vendor 5404 and Product ID values (4-digit hex numbers) and 5405 Flags is a set of characters, each corresponding 5406 to a common usb-storage quirk flag as follows: 5407 a = SANE_SENSE (collect more than 18 bytes 5408 of sense data, not on uas); 5409 b = BAD_SENSE (don't collect more than 18 5410 bytes of sense data, not on uas); 5411 c = FIX_CAPACITY (decrease the reported 5412 device capacity by one sector); 5413 d = NO_READ_DISC_INFO (don't use 5414 READ_DISC_INFO command, not on uas); 5415 e = NO_READ_CAPACITY_16 (don't use 5416 READ_CAPACITY_16 command); 5417 f = NO_REPORT_OPCODES (don't use report opcodes 5418 command, uas only); 5419 g = MAX_SECTORS_240 (don't transfer more than 5420 240 sectors at a time, uas only); 5421 h = CAPACITY_HEURISTICS (decrease the 5422 reported device capacity by one 5423 sector if the number is odd); 5424 i = IGNORE_DEVICE (don't bind to this 5425 device); 5426 j = NO_REPORT_LUNS (don't use report luns 5427 command, uas only); 5428 l = NOT_LOCKABLE (don't try to lock and 5429 unlock ejectable media, not on uas); 5430 m = MAX_SECTORS_64 (don't transfer more 5431 than 64 sectors = 32 KB at a time, 5432 not on uas); 5433 n = INITIAL_READ10 (force a retry of the 5434 initial READ(10) command, not on uas); 5435 o = CAPACITY_OK (accept the capacity 5436 reported by the device, not on uas); 5437 p = WRITE_CACHE (the device cache is ON 5438 by default, not on uas); 5439 r = IGNORE_RESIDUE (the device reports 5440 bogus residue values, not on uas); 5441 s = SINGLE_LUN (the device has only one 5442 Logical Unit); 5443 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5444 commands, uas only); 5445 u = IGNORE_UAS (don't bind to the uas driver); 5446 w = NO_WP_DETECT (don't test whether the 5447 medium is write-protected). 5448 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5449 even if the device claims no cache, 5450 not on uas) 5451 Example: quirks=0419:aaf5:rl,0421:0433:rc 5452 5453 user_debug= [KNL,ARM] 5454 Format: <int> 5455 See arch/arm/Kconfig.debug help text. 5456 1 - undefined instruction events 5457 2 - system calls 5458 4 - invalid data aborts 5459 8 - SIGSEGV faults 5460 16 - SIGBUS faults 5461 Example: user_debug=31 5462 5463 userpte= 5464 [X86] Flags controlling user PTE allocations. 5465 5466 nohigh = do not allocate PTE pages in 5467 HIGHMEM regardless of setting 5468 of CONFIG_HIGHPTE. 5469 5470 vdso= [X86,SH] 5471 On X86_32, this is an alias for vdso32=. Otherwise: 5472 5473 vdso=1: enable VDSO (the default) 5474 vdso=0: disable VDSO mapping 5475 5476 vdso32= [X86] Control the 32-bit vDSO 5477 vdso32=1: enable 32-bit VDSO 5478 vdso32=0 or vdso32=2: disable 32-bit VDSO 5479 5480 See the help text for CONFIG_COMPAT_VDSO for more 5481 details. If CONFIG_COMPAT_VDSO is set, the default is 5482 vdso32=0; otherwise, the default is vdso32=1. 5483 5484 For compatibility with older kernels, vdso32=2 is an 5485 alias for vdso32=0. 5486 5487 Try vdso32=0 if you encounter an error that says: 5488 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 5489 5490 vector= [IA-64,SMP] 5491 vector=percpu: enable percpu vector domain 5492 5493 video= [FB] Frame buffer configuration 5494 See Documentation/fb/modedb.rst. 5495 5496 video.brightness_switch_enabled= [0,1] 5497 If set to 1, on receiving an ACPI notify event 5498 generated by hotkey, video driver will adjust brightness 5499 level and then send out the event to user space through 5500 the allocated input device; If set to 0, video driver 5501 will only send out the event without touching backlight 5502 brightness level. 5503 default: 1 5504 5505 virtio_mmio.device= 5506 [VMMIO] Memory mapped virtio (platform) device. 5507 5508 <size>@<baseaddr>:<irq>[:<id>] 5509 where: 5510 <size> := size (can use standard suffixes 5511 like K, M and G) 5512 <baseaddr> := physical base address 5513 <irq> := interrupt number (as passed to 5514 request_irq()) 5515 <id> := (optional) platform device id 5516 example: 5517 virtio_mmio.device=1K@0x100b0000:48:7 5518 5519 Can be used multiple times for multiple devices. 5520 5521 vga= [BOOT,X86-32] Select a particular video mode 5522 See Documentation/x86/boot.rst and 5523 Documentation/admin-guide/svga.rst. 5524 Use vga=ask for menu. 5525 This is actually a boot loader parameter; the value is 5526 passed to the kernel using a special protocol. 5527 5528 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 5529 May slow down system boot speed, especially when 5530 enabled on systems with a large amount of memory. 5531 All options are enabled by default, and this 5532 interface is meant to allow for selectively 5533 enabling or disabling specific virtual memory 5534 debugging features. 5535 5536 Available options are: 5537 P Enable page structure init time poisoning 5538 - Disable all of the above options 5539 5540 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 5541 size of <nn>. This can be used to increase the 5542 minimum size (128MB on x86). It can also be used to 5543 decrease the size and leave more room for directly 5544 mapped kernel RAM. 5545 5546 vmcp_cma=nn[MG] [KNL,S390] 5547 Sets the memory size reserved for contiguous memory 5548 allocations for the vmcp device driver. 5549 5550 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 5551 Format: <command> 5552 5553 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 5554 Format: <command> 5555 5556 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 5557 Format: <command> 5558 5559 vsyscall= [X86-64] 5560 Controls the behavior of vsyscalls (i.e. calls to 5561 fixed addresses of 0xffffffffff600x00 from legacy 5562 code). Most statically-linked binaries and older 5563 versions of glibc use these calls. Because these 5564 functions are at fixed addresses, they make nice 5565 targets for exploits that can control RIP. 5566 5567 emulate [default] Vsyscalls turn into traps and are 5568 emulated reasonably safely. The vsyscall 5569 page is readable. 5570 5571 xonly Vsyscalls turn into traps and are 5572 emulated reasonably safely. The vsyscall 5573 page is not readable. 5574 5575 none Vsyscalls don't work at all. This makes 5576 them quite hard to use for exploits but 5577 might break your system. 5578 5579 vt.color= [VT] Default text color. 5580 Format: 0xYX, X = foreground, Y = background. 5581 Default: 0x07 = light gray on black. 5582 5583 vt.cur_default= [VT] Default cursor shape. 5584 Format: 0xCCBBAA, where AA, BB, and CC are the same as 5585 the parameters of the <Esc>[?A;B;Cc escape sequence; 5586 see VGA-softcursor.txt. Default: 2 = underline. 5587 5588 vt.default_blu= [VT] 5589 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 5590 Change the default blue palette of the console. 5591 This is a 16-member array composed of values 5592 ranging from 0-255. 5593 5594 vt.default_grn= [VT] 5595 Format: <green0>,<green1>,<green2>,...,<green15> 5596 Change the default green palette of the console. 5597 This is a 16-member array composed of values 5598 ranging from 0-255. 5599 5600 vt.default_red= [VT] 5601 Format: <red0>,<red1>,<red2>,...,<red15> 5602 Change the default red palette of the console. 5603 This is a 16-member array composed of values 5604 ranging from 0-255. 5605 5606 vt.default_utf8= 5607 [VT] 5608 Format=<0|1> 5609 Set system-wide default UTF-8 mode for all tty's. 5610 Default is 1, i.e. UTF-8 mode is enabled for all 5611 newly opened terminals. 5612 5613 vt.global_cursor_default= 5614 [VT] 5615 Format=<-1|0|1> 5616 Set system-wide default for whether a cursor 5617 is shown on new VTs. Default is -1, 5618 i.e. cursors will be created by default unless 5619 overridden by individual drivers. 0 will hide 5620 cursors, 1 will display them. 5621 5622 vt.italic= [VT] Default color for italic text; 0-15. 5623 Default: 2 = green. 5624 5625 vt.underline= [VT] Default color for underlined text; 0-15. 5626 Default: 3 = cyan. 5627 5628 watchdog timers [HW,WDT] For information on watchdog timers, 5629 see Documentation/watchdog/watchdog-parameters.rst 5630 or other driver-specific files in the 5631 Documentation/watchdog/ directory. 5632 5633 watchdog_thresh= 5634 [KNL] 5635 Set the hard lockup detector stall duration 5636 threshold in seconds. The soft lockup detector 5637 threshold is set to twice the value. A value of 0 5638 disables both lockup detectors. Default is 10 5639 seconds. 5640 5641 workqueue.watchdog_thresh= 5642 If CONFIG_WQ_WATCHDOG is configured, workqueue can 5643 warn stall conditions and dump internal state to 5644 help debugging. 0 disables workqueue stall 5645 detection; otherwise, it's the stall threshold 5646 duration in seconds. The default value is 30 and 5647 it can be updated at runtime by writing to the 5648 corresponding sysfs file. 5649 5650 workqueue.disable_numa 5651 By default, all work items queued to unbound 5652 workqueues are affine to the NUMA nodes they're 5653 issued on, which results in better behavior in 5654 general. If NUMA affinity needs to be disabled for 5655 whatever reason, this option can be used. Note 5656 that this also can be controlled per-workqueue for 5657 workqueues visible under /sys/bus/workqueue/. 5658 5659 workqueue.power_efficient 5660 Per-cpu workqueues are generally preferred because 5661 they show better performance thanks to cache 5662 locality; unfortunately, per-cpu workqueues tend to 5663 be more power hungry than unbound workqueues. 5664 5665 Enabling this makes the per-cpu workqueues which 5666 were observed to contribute significantly to power 5667 consumption unbound, leading to measurably lower 5668 power usage at the cost of small performance 5669 overhead. 5670 5671 The default value of this parameter is determined by 5672 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 5673 5674 workqueue.debug_force_rr_cpu 5675 Workqueue used to implicitly guarantee that work 5676 items queued without explicit CPU specified are put 5677 on the local CPU. This guarantee is no longer true 5678 and while local CPU is still preferred work items 5679 may be put on foreign CPUs. This debug option 5680 forces round-robin CPU selection to flush out 5681 usages which depend on the now broken guarantee. 5682 When enabled, memory and cache locality will be 5683 impacted. 5684 5685 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 5686 default x2apic cluster mode on platforms 5687 supporting x2apic. 5688 5689 x86_intel_mid_timer= [X86-32,APBT] 5690 Choose timer option for x86 Intel MID platform. 5691 Two valid options are apbt timer only and lapic timer 5692 plus one apbt timer for broadcast timer. 5693 x86_intel_mid_timer=apbt_only | lapic_and_apbt 5694 5695 xen_512gb_limit [KNL,X86-64,XEN] 5696 Restricts the kernel running paravirtualized under Xen 5697 to use only up to 512 GB of RAM. The reason to do so is 5698 crash analysis tools and Xen tools for doing domain 5699 save/restore/migration must be enabled to handle larger 5700 domains. 5701 5702 xen_emul_unplug= [HW,X86,XEN] 5703 Unplug Xen emulated devices 5704 Format: [unplug0,][unplug1] 5705 ide-disks -- unplug primary master IDE devices 5706 aux-ide-disks -- unplug non-primary-master IDE devices 5707 nics -- unplug network devices 5708 all -- unplug all emulated devices (NICs and IDE disks) 5709 unnecessary -- unplugging emulated devices is 5710 unnecessary even if the host did not respond to 5711 the unplug protocol 5712 never -- do not unplug even if version check succeeds 5713 5714 xen_legacy_crash [X86,XEN] 5715 Crash from Xen panic notifier, without executing late 5716 panic() code such as dumping handler. 5717 5718 xen_nopvspin [X86,XEN] 5719 Disables the ticketlock slowpath using Xen PV 5720 optimizations. 5721 5722 xen_nopv [X86] 5723 Disables the PV optimizations forcing the HVM guest to 5724 run as generic HVM guest with no PV drivers. 5725 This option is obsoleted by the "nopv" option, which 5726 has equivalent effect for XEN platform. 5727 5728 xen_scrub_pages= [XEN] 5729 Boolean option to control scrubbing pages before giving them back 5730 to Xen, for use by other domains. Can be also changed at runtime 5731 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 5732 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 5733 5734 xen_timer_slop= [X86-64,XEN] 5735 Set the timer slop (in nanoseconds) for the virtual Xen 5736 timers (default is 100000). This adjusts the minimum 5737 delta of virtualized Xen timers, where lower values 5738 improve timer resolution at the expense of processing 5739 more timer interrupts. 5740 5741 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 5742 Disables the PV optimizations forcing the guest to run 5743 as generic guest with no PV drivers. Currently support 5744 XEN HVM, KVM, HYPER_V and VMWARE guest. 5745 5746 xirc2ps_cs= [NET,PCMCIA] 5747 Format: 5748 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 5749 5750 xive= [PPC] 5751 By default on POWER9 and above, the kernel will 5752 natively use the XIVE interrupt controller. This option 5753 allows the fallback firmware mode to be used: 5754 5755 off Fallback to firmware control of XIVE interrupt 5756 controller on both pseries and powernv 5757 platforms. Only useful on POWER9 and above. 5758 5759 xhci-hcd.quirks [USB,KNL] 5760 A hex value specifying bitmask with supplemental xhci 5761 host controller quirks. Meaning of each bit can be 5762 consulted in header drivers/usb/host/xhci.h. 5763 5764 xmon [PPC] 5765 Format: { early | on | rw | ro | off } 5766 Controls if xmon debugger is enabled. Default is off. 5767 Passing only "xmon" is equivalent to "xmon=early". 5768 early Call xmon as early as possible on boot; xmon 5769 debugger is called from setup_arch(). 5770 on xmon debugger hooks will be installed so xmon 5771 is only called on a kernel crash. Default mode, 5772 i.e. either "ro" or "rw" mode, is controlled 5773 with CONFIG_XMON_DEFAULT_RO_MODE. 5774 rw xmon debugger hooks will be installed so xmon 5775 is called only on a kernel crash, mode is write, 5776 meaning SPR registers, memory and, other data 5777 can be written using xmon commands. 5778 ro same as "rw" option above but SPR registers, 5779 memory, and other data can't be written using 5780 xmon commands. 5781 off xmon is disabled. 5782