1 acpi= [HW,ACPI,X86,ARM64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force" 14 are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 { vendor | video | native | none } 26 If set to vendor, prefer vendor-specific driver 27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 28 of the ACPI video.ko driver. 29 If set to video, use the ACPI video.ko driver. 30 If set to native, use the device's native backlight mode. 31 If set to none, disable the ACPI backlight interface. 32 33 acpi_force_32bit_fadt_addr 34 force FADT to use 32 bit addresses rather than the 35 64 bit X_* addresses. Some firmware have broken 64 36 bit addresses for force ACPI ignore these and use 37 the older legacy 32 bit addresses. 38 39 acpica_no_return_repair [HW, ACPI] 40 Disable AML predefined validation mechanism 41 This mechanism can repair the evaluation result to make 42 the return objects more ACPI specification compliant. 43 This option is useful for developers to identify the 44 root cause of an AML interpreter issue when the issue 45 has something to do with the repair mechanism. 46 47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 49 Format: <int> 50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 51 debug output. Bits in debug_layer correspond to a 52 _COMPONENT in an ACPI source file, e.g., 53 #define _COMPONENT ACPI_EVENTS 54 Bits in debug_level correspond to a level in 55 ACPI_DEBUG_PRINT statements, e.g., 56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 57 The debug_level mask defaults to "info". See 58 Documentation/firmware-guide/acpi/debug.rst for more information about 59 debug layers and levels. 60 61 Enable processor driver info messages: 62 acpi.debug_layer=0x20000000 63 Enable AML "Debug" output, i.e., stores to the Debug 64 object while interpreting AML: 65 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 66 Enable all messages related to ACPI hardware: 67 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 68 69 Some values produce so much output that the system is 70 unusable. The "log_buf_len" parameter may be useful 71 if you need to capture more output. 72 73 acpi_enforce_resources= [ACPI] 74 { strict | lax | no } 75 Check for resource conflicts between native drivers 76 and ACPI OperationRegions (SystemIO and SystemMemory 77 only). IO ports and memory declared in ACPI might be 78 used by the ACPI subsystem in arbitrary AML code and 79 can interfere with legacy drivers. 80 strict (default): access to resources claimed by ACPI 81 is denied; legacy drivers trying to access reserved 82 resources will fail to bind to device using them. 83 lax: access to resources claimed by ACPI is allowed; 84 legacy drivers trying to access reserved resources 85 will bind successfully but a warning message is logged. 86 no: ACPI OperationRegions are not marked as reserved, 87 no further checks are performed. 88 89 acpi_force_table_verification [HW,ACPI] 90 Enable table checksum verification during early stage. 91 By default, this is disabled due to x86 early mapping 92 size limitation. 93 94 acpi_irq_balance [HW,ACPI] 95 ACPI will balance active IRQs 96 default in APIC mode 97 98 acpi_irq_nobalance [HW,ACPI] 99 ACPI will not move active IRQs (default) 100 default in PIC mode 101 102 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 103 Format: <irq>,<irq>... 104 105 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 106 use by PCI 107 Format: <irq>,<irq>... 108 109 acpi_mask_gpe= [HW,ACPI] 110 Due to the existence of _Lxx/_Exx, some GPEs triggered 111 by unsupported hardware/firmware features can result in 112 GPE floodings that cannot be automatically disabled by 113 the GPE dispatcher. 114 This facility can be used to prevent such uncontrolled 115 GPE floodings. 116 Format: <byte> 117 118 acpi_no_auto_serialize [HW,ACPI] 119 Disable auto-serialization of AML methods 120 AML control methods that contain the opcodes to create 121 named objects will be marked as "Serialized" by the 122 auto-serialization feature. 123 This feature is enabled by default. 124 This option allows to turn off the feature. 125 126 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 127 kernels. 128 129 acpi_no_static_ssdt [HW,ACPI] 130 Disable installation of static SSDTs at early boot time 131 By default, SSDTs contained in the RSDT/XSDT will be 132 installed automatically and they will appear under 133 /sys/firmware/acpi/tables. 134 This option turns off this feature. 135 Note that specifying this option does not affect 136 dynamic table installation which will install SSDT 137 tables to /sys/firmware/acpi/tables/dynamic. 138 139 acpi_no_watchdog [HW,ACPI,WDT] 140 Ignore the ACPI-based watchdog interface (WDAT) and let 141 a native driver control the watchdog device instead. 142 143 acpi_rsdp= [ACPI,EFI,KEXEC] 144 Pass the RSDP address to the kernel, mostly used 145 on machines running EFI runtime service to boot the 146 second kernel for kdump. 147 148 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 149 Format: To spoof as Windows 98: ="Microsoft Windows" 150 151 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 152 of 2 which is mandated by ACPI 6) as the supported ACPI 153 specification revision (when using this switch, it may 154 be necessary to carry out a cold reboot _twice_ in a 155 row to make it take effect on the platform firmware). 156 157 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 158 acpi_osi="string1" # add string1 159 acpi_osi="!string2" # remove string2 160 acpi_osi=!* # remove all strings 161 acpi_osi=! # disable all built-in OS vendor 162 strings 163 acpi_osi=!! # enable all built-in OS vendor 164 strings 165 acpi_osi= # disable all strings 166 167 'acpi_osi=!' can be used in combination with single or 168 multiple 'acpi_osi="string1"' to support specific OS 169 vendor string(s). Note that such command can only 170 affect the default state of the OS vendor strings, thus 171 it cannot affect the default state of the feature group 172 strings and the current state of the OS vendor strings, 173 specifying it multiple times through kernel command line 174 is meaningless. This command is useful when one do not 175 care about the state of the feature group strings which 176 should be controlled by the OSPM. 177 Examples: 178 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 179 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 180 can make '_OSI("Windows 2000")' TRUE. 181 182 'acpi_osi=' cannot be used in combination with other 183 'acpi_osi=' command lines, the _OSI method will not 184 exist in the ACPI namespace. NOTE that such command can 185 only affect the _OSI support state, thus specifying it 186 multiple times through kernel command line is also 187 meaningless. 188 Examples: 189 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 190 FALSE. 191 192 'acpi_osi=!*' can be used in combination with single or 193 multiple 'acpi_osi="string1"' to support specific 194 string(s). Note that such command can affect the 195 current state of both the OS vendor strings and the 196 feature group strings, thus specifying it multiple times 197 through kernel command line is meaningful. But it may 198 still not able to affect the final state of a string if 199 there are quirks related to this string. This command 200 is useful when one want to control the state of the 201 feature group strings to debug BIOS issues related to 202 the OSPM features. 203 Examples: 204 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 205 '_OSI("Module Device")' FALSE. 206 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 207 '_OSI("Module Device")' TRUE. 208 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 209 equivalent to 210 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 211 and 212 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 213 they all will make '_OSI("Windows 2000")' TRUE. 214 215 acpi_pm_good [X86] 216 Override the pmtimer bug detection: force the kernel 217 to assume that this machine's pmtimer latches its value 218 and always returns good values. 219 220 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 221 Format: { level | edge | high | low } 222 223 acpi_skip_timer_override [HW,ACPI] 224 Recognize and ignore IRQ0/pin2 Interrupt Override. 225 For broken nForce2 BIOS resulting in XT-PIC timer. 226 227 acpi_sleep= [HW,ACPI] Sleep options 228 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig, 229 old_ordering, nonvs, sci_force_enable, nobl } 230 See Documentation/power/video.rst for information on 231 s3_bios and s3_mode. 232 s3_beep is for debugging; it makes the PC's speaker beep 233 as soon as the kernel's real-mode entry point is called. 234 s4_nohwsig prevents ACPI hardware signature from being 235 used during resume from hibernation. 236 old_ordering causes the ACPI 1.0 ordering of the _PTS 237 control method, with respect to putting devices into 238 low power states, to be enforced (the ACPI 2.0 ordering 239 of _PTS is used by default). 240 nonvs prevents the kernel from saving/restoring the 241 ACPI NVS memory during suspend/hibernation and resume. 242 sci_force_enable causes the kernel to set SCI_EN directly 243 on resume from S1/S3 (which is against the ACPI spec, 244 but some broken systems don't work without it). 245 nobl causes the internal blacklist of systems known to 246 behave incorrectly in some ways with respect to system 247 suspend and resume to be ignored (use wisely). 248 249 acpi_use_timer_override [HW,ACPI] 250 Use timer override. For some broken Nvidia NF5 boards 251 that require a timer override, but don't have HPET 252 253 add_efi_memmap [EFI; X86] Include EFI memory map in 254 kernel's map of available physical RAM. 255 256 agp= [AGP] 257 { off | try_unsupported } 258 off: disable AGP support 259 try_unsupported: try to drive unsupported chipsets 260 (may crash computer or cause data corruption) 261 262 ALSA [HW,ALSA] 263 See Documentation/sound/alsa-configuration.rst 264 265 alignment= [KNL,ARM] 266 Allow the default userspace alignment fault handler 267 behaviour to be specified. Bit 0 enables warnings, 268 bit 1 enables fixups, and bit 2 sends a segfault. 269 270 align_va_addr= [X86-64] 271 Align virtual addresses by clearing slice [14:12] when 272 allocating a VMA at process creation time. This option 273 gives you up to 3% performance improvement on AMD F15h 274 machines (where it is enabled by default) for a 275 CPU-intensive style benchmark, and it can vary highly in 276 a microbenchmark depending on workload and compiler. 277 278 32: only for 32-bit processes 279 64: only for 64-bit processes 280 on: enable for both 32- and 64-bit processes 281 off: disable for both 32- and 64-bit processes 282 283 alloc_snapshot [FTRACE] 284 Allocate the ftrace snapshot buffer on boot up when the 285 main buffer is allocated. This is handy if debugging 286 and you need to use tracing_snapshot() on boot up, and 287 do not want to use tracing_snapshot_alloc() as it needs 288 to be done where GFP_KERNEL allocations are allowed. 289 290 amd_iommu= [HW,X86-64] 291 Pass parameters to the AMD IOMMU driver in the system. 292 Possible values are: 293 fullflush - enable flushing of IO/TLB entries when 294 they are unmapped. Otherwise they are 295 flushed before they will be reused, which 296 is a lot of faster 297 off - do not initialize any AMD IOMMU found in 298 the system 299 force_isolation - Force device isolation for all 300 devices. The IOMMU driver is not 301 allowed anymore to lift isolation 302 requirements as needed. This option 303 does not override iommu=pt 304 305 amd_iommu_dump= [HW,X86-64] 306 Enable AMD IOMMU driver option to dump the ACPI table 307 for AMD IOMMU. With this option enabled, AMD IOMMU 308 driver will print ACPI tables for AMD IOMMU during 309 IOMMU initialization. 310 311 amd_iommu_intr= [HW,X86-64] 312 Specifies one of the following AMD IOMMU interrupt 313 remapping modes: 314 legacy - Use legacy interrupt remapping mode. 315 vapic - Use virtual APIC mode, which allows IOMMU 316 to inject interrupts directly into guest. 317 This mode requires kvm-amd.avic=1. 318 (Default when IOMMU HW support is present.) 319 320 amijoy.map= [HW,JOY] Amiga joystick support 321 Map of devices attached to JOY0DAT and JOY1DAT 322 Format: <a>,<b> 323 See also Documentation/input/joydev/joystick.rst 324 325 analog.map= [HW,JOY] Analog joystick and gamepad support 326 Specifies type or capabilities of an analog joystick 327 connected to one of 16 gameports 328 Format: <type1>,<type2>,..<type16> 329 330 apc= [HW,SPARC] 331 Power management functions (SPARCstation-4/5 + deriv.) 332 Format: noidle 333 Disable APC CPU standby support. SPARCstation-Fox does 334 not play well with APC CPU idle - disable it if you have 335 APC and your system crashes randomly. 336 337 apic= [APIC,X86] Advanced Programmable Interrupt Controller 338 Change the output verbosity while booting 339 Format: { quiet (default) | verbose | debug } 340 Change the amount of debugging information output 341 when initialising the APIC and IO-APIC components. 342 For X86-32, this can also be used to specify an APIC 343 driver name. 344 Format: apic=driver_name 345 Examples: apic=bigsmp 346 347 apic_extnmi= [APIC,X86] External NMI delivery setting 348 Format: { bsp (default) | all | none } 349 bsp: External NMI is delivered only to CPU 0 350 all: External NMIs are broadcast to all CPUs as a 351 backup of CPU 0 352 none: External NMI is masked for all CPUs. This is 353 useful so that a dump capture kernel won't be 354 shot down by NMI 355 356 autoconf= [IPV6] 357 See Documentation/networking/ipv6.rst. 358 359 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 360 Limit apic dumping. The parameter defines the maximal 361 number of local apics being dumped. Also it is possible 362 to set it to "all" by meaning -- no limit here. 363 Format: { 1 (default) | 2 | ... | all }. 364 The parameter valid if only apic=debug or 365 apic=verbose is specified. 366 Example: apic=debug show_lapic=all 367 368 apm= [APM] Advanced Power Management 369 See header of arch/x86/kernel/apm_32.c. 370 371 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 372 Format: <io>,<irq>,<nodeID> 373 374 arm64.nobti [ARM64] Unconditionally disable Branch Target 375 Identification support 376 377 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 378 support 379 380 ataflop= [HW,M68k] 381 382 atarimouse= [HW,MOUSE] Atari Mouse 383 384 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 385 EzKey and similar keyboards 386 387 atkbd.reset= [HW] Reset keyboard during initialization 388 389 atkbd.set= [HW] Select keyboard code set 390 Format: <int> (2 = AT (default), 3 = PS/2) 391 392 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 393 keyboards 394 395 atkbd.softraw= [HW] Choose between synthetic and real raw mode 396 Format: <bool> (0 = real, 1 = synthetic (default)) 397 398 atkbd.softrepeat= [HW] 399 Use software keyboard repeat 400 401 audit= [KNL] Enable the audit sub-system 402 Format: { "0" | "1" | "off" | "on" } 403 0 | off - kernel audit is disabled and can not be 404 enabled until the next reboot 405 unset - kernel audit is initialized but disabled and 406 will be fully enabled by the userspace auditd. 407 1 | on - kernel audit is initialized and partially 408 enabled, storing at most audit_backlog_limit 409 messages in RAM until it is fully enabled by the 410 userspace auditd. 411 Default: unset 412 413 audit_backlog_limit= [KNL] Set the audit queue size limit. 414 Format: <int> (must be >=0) 415 Default: 64 416 417 bau= [X86_UV] Enable the BAU on SGI UV. The default 418 behavior is to disable the BAU (i.e. bau=0). 419 Format: { "0" | "1" } 420 0 - Disable the BAU. 421 1 - Enable the BAU. 422 unset - Disable the BAU. 423 424 baycom_epp= [HW,AX25] 425 Format: <io>,<mode> 426 427 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 428 Format: <io>,<mode> 429 See header of drivers/net/hamradio/baycom_par.c. 430 431 baycom_ser_fdx= [HW,AX25] 432 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 433 Format: <io>,<irq>,<mode>[,<baud>] 434 See header of drivers/net/hamradio/baycom_ser_fdx.c. 435 436 baycom_ser_hdx= [HW,AX25] 437 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 438 Format: <io>,<irq>,<mode> 439 See header of drivers/net/hamradio/baycom_ser_hdx.c. 440 441 blkdevparts= Manual partition parsing of block device(s) for 442 embedded devices based on command line input. 443 See Documentation/block/cmdline-partition.rst 444 445 boot_delay= Milliseconds to delay each printk during boot. 446 Values larger than 10 seconds (10000) are changed to 447 no delay (0). 448 Format: integer 449 450 bootconfig [KNL] 451 Extended command line options can be added to an initrd 452 and this will cause the kernel to look for it. 453 454 See Documentation/admin-guide/bootconfig.rst 455 456 bert_disable [ACPI] 457 Disable BERT OS support on buggy BIOSes. 458 459 bgrt_disable [ACPI][X86] 460 Disable BGRT to avoid flickering OEM logo. 461 462 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 463 bttv.radio= Most important insmod options are available as 464 kernel args too. 465 bttv.pll= See Documentation/admin-guide/media/bttv.rst 466 bttv.tuner= 467 468 bulk_remove=off [PPC] This parameter disables the use of the pSeries 469 firmware feature for flushing multiple hpte entries 470 at a time. 471 472 c101= [NET] Moxa C101 synchronous serial card 473 474 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 475 Sometimes CPU hardware bugs make them report the cache 476 size incorrectly. The kernel will attempt work arounds 477 to fix known problems, but for some CPUs it is not 478 possible to determine what the correct size should be. 479 This option provides an override for these situations. 480 481 carrier_timeout= 482 [NET] Specifies amount of time (in seconds) that 483 the kernel should wait for a network carrier. By default 484 it waits 120 seconds. 485 486 ca_keys= [KEYS] This parameter identifies a specific key(s) on 487 the system trusted keyring to be used for certificate 488 trust validation. 489 format: { id:<keyid> | builtin } 490 491 cca= [MIPS] Override the kernel pages' cache coherency 492 algorithm. Accepted values range from 0 to 7 493 inclusive. See arch/mips/include/asm/pgtable-bits.h 494 for platform specific values (SB1, Loongson3 and 495 others). 496 497 ccw_timeout_log [S390] 498 See Documentation/s390/common_io.rst for details. 499 500 cgroup_disable= [KNL] Disable a particular controller 501 Format: {name of the controller(s) to disable} 502 The effects of cgroup_disable=foo are: 503 - foo isn't auto-mounted if you mount all cgroups in 504 a single hierarchy 505 - foo isn't visible as an individually mountable 506 subsystem 507 {Currently only "memory" controller deal with this and 508 cut the overhead, others just disable the usage. So 509 only cgroup_disable=memory is actually worthy} 510 511 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 512 Format: { { controller | "all" | "named" } 513 [,{ controller | "all" | "named" }...] } 514 Like cgroup_disable, but only applies to cgroup v1; 515 the blacklisted controllers remain available in cgroup2. 516 "all" blacklists all controllers and "named" disables 517 named mounts. Specifying both "all" and "named" disables 518 all v1 hierarchies. 519 520 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 521 Format: <string> 522 nosocket -- Disable socket memory accounting. 523 nokmem -- Disable kernel memory accounting. 524 525 checkreqprot [SELINUX] Set initial checkreqprot flag value. 526 Format: { "0" | "1" } 527 See security/selinux/Kconfig help text. 528 0 -- check protection applied by kernel (includes 529 any implied execute protection). 530 1 -- check protection requested by application. 531 Default value is set via a kernel config option. 532 Value can be changed at runtime via 533 /sys/fs/selinux/checkreqprot. 534 Setting checkreqprot to 1 is deprecated. 535 536 cio_ignore= [S390] 537 See Documentation/s390/common_io.rst for details. 538 clk_ignore_unused 539 [CLK] 540 Prevents the clock framework from automatically gating 541 clocks that have not been explicitly enabled by a Linux 542 device driver but are enabled in hardware at reset or 543 by the bootloader/firmware. Note that this does not 544 force such clocks to be always-on nor does it reserve 545 those clocks in any way. This parameter is useful for 546 debug and development, but should not be needed on a 547 platform with proper driver support. For more 548 information, see Documentation/driver-api/clk.rst. 549 550 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 551 [Deprecated] 552 Forces specified clocksource (if available) to be used 553 when calculating gettimeofday(). If specified 554 clocksource is not available, it defaults to PIT. 555 Format: { pit | tsc | cyclone | pmtmr } 556 557 clocksource= Override the default clocksource 558 Format: <string> 559 Override the default clocksource and use the clocksource 560 with the name specified. 561 Some clocksource names to choose from, depending on 562 the platform: 563 [all] jiffies (this is the base, fallback clocksource) 564 [ACPI] acpi_pm 565 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 566 pxa_timer,timer3,32k_counter,timer0_1 567 [X86-32] pit,hpet,tsc; 568 scx200_hrt on Geode; cyclone on IBM x440 569 [MIPS] MIPS 570 [PARISC] cr16 571 [S390] tod 572 [SH] SuperH 573 [SPARC64] tick 574 [X86-64] hpet,tsc 575 576 clocksource.arm_arch_timer.evtstrm= 577 [ARM,ARM64] 578 Format: <bool> 579 Enable/disable the eventstream feature of the ARM 580 architected timer so that code using WFE-based polling 581 loops can be debugged more effectively on production 582 systems. 583 584 clearcpuid=BITNUM[,BITNUM...] [X86] 585 Disable CPUID feature X for the kernel. See 586 arch/x86/include/asm/cpufeatures.h for the valid bit 587 numbers. Note the Linux specific bits are not necessarily 588 stable over kernel options, but the vendor specific 589 ones should be. 590 Also note that user programs calling CPUID directly 591 or using the feature without checking anything 592 will still see it. This just prevents it from 593 being used by the kernel or shown in /proc/cpuinfo. 594 Also note the kernel might malfunction if you disable 595 some critical bits. 596 597 cma=nn[MG]@[start[MG][-end[MG]]] 598 [KNL,CMA] 599 Sets the size of kernel global memory area for 600 contiguous memory allocations and optionally the 601 placement constraint by the physical address range of 602 memory allocations. A value of 0 disables CMA 603 altogether. For more information, see 604 kernel/dma/contiguous.c 605 606 cma_pernuma=nn[MG] 607 [ARM64,KNL,CMA] 608 Sets the size of kernel per-numa memory area for 609 contiguous memory allocations. A value of 0 disables 610 per-numa CMA altogether. And If this option is not 611 specificed, the default value is 0. 612 With per-numa CMA enabled, DMA users on node nid will 613 first try to allocate buffer from the pernuma area 614 which is located in node nid, if the allocation fails, 615 they will fallback to the global default memory area. 616 617 cmo_free_hint= [PPC] Format: { yes | no } 618 Specify whether pages are marked as being inactive 619 when they are freed. This is used in CMO environments 620 to determine OS memory pressure for page stealing by 621 a hypervisor. 622 Default: yes 623 624 coherent_pool=nn[KMG] [ARM,KNL] 625 Sets the size of memory pool for coherent, atomic dma 626 allocations, by default set to 256K. 627 628 com20020= [HW,NET] ARCnet - COM20020 chipset 629 Format: 630 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 631 632 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 633 Format: <io>[,<irq>] 634 635 com90xx= [HW,NET] 636 ARCnet - COM90xx chipset (memory-mapped buffers) 637 Format: <io>[,<irq>[,<memstart>]] 638 639 condev= [HW,S390] console device 640 conmode= 641 642 console= [KNL] Output console device and options. 643 644 tty<n> Use the virtual console device <n>. 645 646 ttyS<n>[,options] 647 ttyUSB0[,options] 648 Use the specified serial port. The options are of 649 the form "bbbbpnf", where "bbbb" is the baud rate, 650 "p" is parity ("n", "o", or "e"), "n" is number of 651 bits, and "f" is flow control ("r" for RTS or 652 omit it). Default is "9600n8". 653 654 See Documentation/admin-guide/serial-console.rst for more 655 information. See 656 Documentation/networking/netconsole.rst for an 657 alternative. 658 659 uart[8250],io,<addr>[,options] 660 uart[8250],mmio,<addr>[,options] 661 uart[8250],mmio16,<addr>[,options] 662 uart[8250],mmio32,<addr>[,options] 663 uart[8250],0x<addr>[,options] 664 Start an early, polled-mode console on the 8250/16550 665 UART at the specified I/O port or MMIO address, 666 switching to the matching ttyS device later. 667 MMIO inter-register address stride is either 8-bit 668 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 669 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 670 to be equivalent to 'mmio'. 'options' are specified in 671 the same format described for ttyS above; if unspecified, 672 the h/w is not re-initialized. 673 674 hvc<n> Use the hypervisor console device <n>. This is for 675 both Xen and PowerPC hypervisors. 676 677 If the device connected to the port is not a TTY but a braille 678 device, prepend "brl," before the device type, for instance 679 console=brl,ttyS0 680 For now, only VisioBraille is supported. 681 682 console_msg_format= 683 [KNL] Change console messages format 684 default 685 By default we print messages on consoles in 686 "[time stamp] text\n" format (time stamp may not be 687 printed, depending on CONFIG_PRINTK_TIME or 688 `printk_time' param). 689 syslog 690 Switch to syslog format: "<%u>[time stamp] text\n" 691 IOW, each message will have a facility and loglevel 692 prefix. The format is similar to one used by syslog() 693 syscall, or to executing "dmesg -S --raw" or to reading 694 from /proc/kmsg. 695 696 consoleblank= [KNL] The console blank (screen saver) timeout in 697 seconds. A value of 0 disables the blank timer. 698 Defaults to 0. 699 700 coredump_filter= 701 [KNL] Change the default value for 702 /proc/<pid>/coredump_filter. 703 See also Documentation/filesystems/proc.rst. 704 705 coresight_cpu_debug.enable 706 [ARM,ARM64] 707 Format: <bool> 708 Enable/disable the CPU sampling based debugging. 709 0: default value, disable debugging 710 1: enable debugging at boot time 711 712 cpuidle.off=1 [CPU_IDLE] 713 disable the cpuidle sub-system 714 715 cpuidle.governor= 716 [CPU_IDLE] Name of the cpuidle governor to use. 717 718 cpufreq.off=1 [CPU_FREQ] 719 disable the cpufreq sub-system 720 721 cpufreq.default_governor= 722 [CPU_FREQ] Name of the default cpufreq governor or 723 policy to use. This governor must be registered in the 724 kernel before the cpufreq driver probes. 725 726 cpu_init_udelay=N 727 [X86] Delay for N microsec between assert and de-assert 728 of APIC INIT to start processors. This delay occurs 729 on every CPU online, such as boot, and resume from suspend. 730 Default: 10000 731 732 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 733 Format: 734 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 735 736 crashkernel=size[KMG][@offset[KMG]] 737 [KNL] Using kexec, Linux can switch to a 'crash kernel' 738 upon panic. This parameter reserves the physical 739 memory region [offset, offset + size] for that kernel 740 image. If '@offset' is omitted, then a suitable offset 741 is selected automatically. 742 [KNL, X86-64] Select a region under 4G first, and 743 fall back to reserve region above 4G when '@offset' 744 hasn't been specified. 745 See Documentation/admin-guide/kdump/kdump.rst for further details. 746 747 crashkernel=range1:size1[,range2:size2,...][@offset] 748 [KNL] Same as above, but depends on the memory 749 in the running system. The syntax of range is 750 start-[end] where start and end are both 751 a memory unit (amount[KMG]). See also 752 Documentation/admin-guide/kdump/kdump.rst for an example. 753 754 crashkernel=size[KMG],high 755 [KNL, X86-64] range could be above 4G. Allow kernel 756 to allocate physical memory region from top, so could 757 be above 4G if system have more than 4G ram installed. 758 Otherwise memory region will be allocated below 4G, if 759 available. 760 It will be ignored if crashkernel=X is specified. 761 crashkernel=size[KMG],low 762 [KNL, X86-64] range under 4G. When crashkernel=X,high 763 is passed, kernel could allocate physical memory region 764 above 4G, that cause second kernel crash on system 765 that require some amount of low memory, e.g. swiotlb 766 requires at least 64M+32K low memory, also enough extra 767 low memory is needed to make sure DMA buffers for 32-bit 768 devices won't run out. Kernel would try to allocate at 769 at least 256M below 4G automatically. 770 This one let user to specify own low range under 4G 771 for second kernel instead. 772 0: to disable low allocation. 773 It will be ignored when crashkernel=X,high is not used 774 or memory reserved is below 4G. 775 776 cryptomgr.notests 777 [KNL] Disable crypto self-tests 778 779 cs89x0_dma= [HW,NET] 780 Format: <dma> 781 782 cs89x0_media= [HW,NET] 783 Format: { rj45 | aui | bnc } 784 785 csdlock_debug= [KNL] Enable debug add-ons of cross-CPU function call 786 handling. When switched on, additional debug data is 787 printed to the console in case a hanging CPU is 788 detected, and that CPU is pinged again in order to try 789 to resolve the hang situation. 790 0: disable csdlock debugging (default) 791 1: enable basic csdlock debugging (minor impact) 792 ext: enable extended csdlock debugging (more impact, 793 but more data) 794 795 dasd= [HW,NET] 796 See header of drivers/s390/block/dasd_devmap.c. 797 798 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 799 (one device per port) 800 Format: <port#>,<type> 801 See also Documentation/input/devices/joystick-parport.rst 802 803 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 804 time. See 805 Documentation/admin-guide/dynamic-debug-howto.rst for 806 details. Deprecated, see dyndbg. 807 808 debug [KNL] Enable kernel debugging (events log level). 809 810 debug_boot_weak_hash 811 [KNL] Enable printing [hashed] pointers early in the 812 boot sequence. If enabled, we use a weak hash instead 813 of siphash to hash pointers. Use this option if you are 814 seeing instances of '(___ptrval___)') and need to see a 815 value (hashed pointer) instead. Cryptographically 816 insecure, please do not use on production kernels. 817 818 debug_locks_verbose= 819 [KNL] verbose locking self-tests 820 Format: <int> 821 Print debugging info while doing the locking API 822 self-tests. 823 Bitmask for the various LOCKTYPE_ tests. Defaults to 0 824 (no extra messages), setting it to -1 (all bits set) 825 will print _a_lot_ more information - normally only 826 useful to lockdep developers. 827 828 debug_objects [KNL] Enable object debugging 829 830 no_debug_objects 831 [KNL] Disable object debugging 832 833 debug_guardpage_minorder= 834 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 835 parameter allows control of the order of pages that will 836 be intentionally kept free (and hence protected) by the 837 buddy allocator. Bigger value increase the probability 838 of catching random memory corruption, but reduce the 839 amount of memory for normal system use. The maximum 840 possible value is MAX_ORDER/2. Setting this parameter 841 to 1 or 2 should be enough to identify most random 842 memory corruption problems caused by bugs in kernel or 843 driver code when a CPU writes to (or reads from) a 844 random memory location. Note that there exists a class 845 of memory corruptions problems caused by buggy H/W or 846 F/W or by drivers badly programing DMA (basically when 847 memory is written at bus level and the CPU MMU is 848 bypassed) which are not detectable by 849 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 850 tracking down these problems. 851 852 debug_pagealloc= 853 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 854 enables the feature at boot time. By default, it is 855 disabled and the system will work mostly the same as a 856 kernel built without CONFIG_DEBUG_PAGEALLOC. 857 Note: to get most of debug_pagealloc error reports, it's 858 useful to also enable the page_owner functionality. 859 on: enable the feature 860 861 debugfs= [KNL] This parameter enables what is exposed to userspace 862 and debugfs internal clients. 863 Format: { on, no-mount, off } 864 on: All functions are enabled. 865 no-mount: 866 Filesystem is not registered but kernel clients can 867 access APIs and a crashkernel can be used to read 868 its content. There is nothing to mount. 869 off: Filesystem is not registered and clients 870 get a -EPERM as result when trying to register files 871 or directories within debugfs. 872 This is equivalent of the runtime functionality if 873 debugfs was not enabled in the kernel at all. 874 Default value is set in build-time with a kernel configuration. 875 876 debugpat [X86] Enable PAT debugging 877 878 decnet.addr= [HW,NET] 879 Format: <area>[,<node>] 880 See also Documentation/networking/decnet.rst. 881 882 default_hugepagesz= 883 [HW] The size of the default HugeTLB page. This is 884 the size represented by the legacy /proc/ hugepages 885 APIs. In addition, this is the default hugetlb size 886 used for shmget(), mmap() and mounting hugetlbfs 887 filesystems. If not specified, defaults to the 888 architecture's default huge page size. Huge page 889 sizes are architecture dependent. See also 890 Documentation/admin-guide/mm/hugetlbpage.rst. 891 Format: size[KMG] 892 893 deferred_probe_timeout= 894 [KNL] Debugging option to set a timeout in seconds for 895 deferred probe to give up waiting on dependencies to 896 probe. Only specific dependencies (subsystems or 897 drivers) that have opted in will be ignored. A timeout of 0 898 will timeout at the end of initcalls. This option will also 899 dump out devices still on the deferred probe list after 900 retrying. 901 902 dfltcc= [HW,S390] 903 Format: { on | off | def_only | inf_only | always } 904 on: s390 zlib hardware support for compression on 905 level 1 and decompression (default) 906 off: No s390 zlib hardware support 907 def_only: s390 zlib hardware support for deflate 908 only (compression on level 1) 909 inf_only: s390 zlib hardware support for inflate 910 only (decompression) 911 always: Same as 'on' but ignores the selected compression 912 level always using hardware support (used for debugging) 913 914 dhash_entries= [KNL] 915 Set number of hash buckets for dentry cache. 916 917 disable_1tb_segments [PPC] 918 Disables the use of 1TB hash page table segments. This 919 causes the kernel to fall back to 256MB segments which 920 can be useful when debugging issues that require an SLB 921 miss to occur. 922 923 stress_slb [PPC] 924 Limits the number of kernel SLB entries, and flushes 925 them frequently to increase the rate of SLB faults 926 on kernel addresses. 927 928 disable= [IPV6] 929 See Documentation/networking/ipv6.rst. 930 931 hardened_usercopy= 932 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 933 hardening is enabled for this boot. Hardened 934 usercopy checking is used to protect the kernel 935 from reading or writing beyond known memory 936 allocation boundaries as a proactive defense 937 against bounds-checking flaws in the kernel's 938 copy_to_user()/copy_from_user() interface. 939 on Perform hardened usercopy checks (default). 940 off Disable hardened usercopy checks. 941 942 disable_radix [PPC] 943 Disable RADIX MMU mode on POWER9 944 945 radix_hcall_invalidate=on [PPC/PSERIES] 946 Disable RADIX GTSE feature and use hcall for TLB 947 invalidate. 948 949 disable_tlbie [PPC] 950 Disable TLBIE instruction. Currently does not work 951 with KVM, with HASH MMU, or with coherent accelerators. 952 953 disable_cpu_apicid= [X86,APIC,SMP] 954 Format: <int> 955 The number of initial APIC ID for the 956 corresponding CPU to be disabled at boot, 957 mostly used for the kdump 2nd kernel to 958 disable BSP to wake up multiple CPUs without 959 causing system reset or hang due to sending 960 INIT from AP to BSP. 961 962 disable_ddw [PPC/PSERIES] 963 Disable Dynamic DMA Window support. Use this 964 to workaround buggy firmware. 965 966 disable_ipv6= [IPV6] 967 See Documentation/networking/ipv6.rst. 968 969 disable_mtrr_cleanup [X86] 970 The kernel tries to adjust MTRR layout from continuous 971 to discrete, to make X server driver able to add WB 972 entry later. This parameter disables that. 973 974 disable_mtrr_trim [X86, Intel and AMD only] 975 By default the kernel will trim any uncacheable 976 memory out of your available memory pool based on 977 MTRR settings. This parameter disables that behavior, 978 possibly causing your machine to run very slowly. 979 980 disable_timer_pin_1 [X86] 981 Disable PIN 1 of APIC timer 982 Can be useful to work around chipset bugs. 983 984 dis_ucode_ldr [X86] Disable the microcode loader. 985 986 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 987 this option disables the debugging code at boot. 988 989 dma_debug_entries=<number> 990 This option allows to tune the number of preallocated 991 entries for DMA-API debugging code. One entry is 992 required per DMA-API allocation. Use this if the 993 DMA-API debugging code disables itself because the 994 architectural default is too low. 995 996 dma_debug_driver=<driver_name> 997 With this option the DMA-API debugging driver 998 filter feature can be enabled at boot time. Just 999 pass the driver to filter for as the parameter. 1000 The filter can be disabled or changed to another 1001 driver later using sysfs. 1002 1003 driver_async_probe= [KNL] 1004 List of driver names to be probed asynchronously. 1005 Format: <driver_name1>,<driver_name2>... 1006 1007 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1008 Broken monitors, graphic adapters, KVMs and EDIDless 1009 panels may send no or incorrect EDID data sets. 1010 This parameter allows to specify an EDID data sets 1011 in the /lib/firmware directory that are used instead. 1012 Generic built-in EDID data sets are used, if one of 1013 edid/1024x768.bin, edid/1280x1024.bin, 1014 edid/1680x1050.bin, or edid/1920x1080.bin is given 1015 and no file with the same name exists. Details and 1016 instructions how to build your own EDID data are 1017 available in Documentation/admin-guide/edid.rst. An EDID 1018 data set will only be used for a particular connector, 1019 if its name and a colon are prepended to the EDID 1020 name. Each connector may use a unique EDID data 1021 set by separating the files with a comma. An EDID 1022 data set with no connector name will be used for 1023 any connectors not explicitly specified. 1024 1025 dscc4.setup= [NET] 1026 1027 dt_cpu_ftrs= [PPC] 1028 Format: {"off" | "known"} 1029 Control how the dt_cpu_ftrs device-tree binding is 1030 used for CPU feature discovery and setup (if it 1031 exists). 1032 off: Do not use it, fall back to legacy cpu table. 1033 known: Do not pass through unknown features to guests 1034 or userspace, only those that the kernel is aware of. 1035 1036 dump_apple_properties [X86] 1037 Dump name and content of EFI device properties on 1038 x86 Macs. Useful for driver authors to determine 1039 what data is available or for reverse-engineering. 1040 1041 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1042 <module>.dyndbg[="val"] 1043 Enable debug messages at boot time. See 1044 Documentation/admin-guide/dynamic-debug-howto.rst 1045 for details. 1046 1047 nopku [X86] Disable Memory Protection Keys CPU feature found 1048 in some Intel CPUs. 1049 1050 <module>.async_probe [KNL] 1051 Enable asynchronous probe on this module. 1052 1053 early_ioremap_debug [KNL] 1054 Enable debug messages in early_ioremap support. This 1055 is useful for tracking down temporary early mappings 1056 which are not unmapped. 1057 1058 earlycon= [KNL] Output early console device and options. 1059 1060 When used with no options, the early console is 1061 determined by stdout-path property in device tree's 1062 chosen node or the ACPI SPCR table if supported by 1063 the platform. 1064 1065 cdns,<addr>[,options] 1066 Start an early, polled-mode console on a Cadence 1067 (xuartps) serial port at the specified address. Only 1068 supported option is baud rate. If baud rate is not 1069 specified, the serial port must already be setup and 1070 configured. 1071 1072 uart[8250],io,<addr>[,options] 1073 uart[8250],mmio,<addr>[,options] 1074 uart[8250],mmio32,<addr>[,options] 1075 uart[8250],mmio32be,<addr>[,options] 1076 uart[8250],0x<addr>[,options] 1077 Start an early, polled-mode console on the 8250/16550 1078 UART at the specified I/O port or MMIO address. 1079 MMIO inter-register address stride is either 8-bit 1080 (mmio) or 32-bit (mmio32 or mmio32be). 1081 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1082 to be equivalent to 'mmio'. 'options' are specified 1083 in the same format described for "console=ttyS<n>"; if 1084 unspecified, the h/w is not initialized. 1085 1086 pl011,<addr> 1087 pl011,mmio32,<addr> 1088 Start an early, polled-mode console on a pl011 serial 1089 port at the specified address. The pl011 serial port 1090 must already be setup and configured. Options are not 1091 yet supported. If 'mmio32' is specified, then only 1092 the driver will use only 32-bit accessors to read/write 1093 the device registers. 1094 1095 meson,<addr> 1096 Start an early, polled-mode console on a meson serial 1097 port at the specified address. The serial port must 1098 already be setup and configured. Options are not yet 1099 supported. 1100 1101 msm_serial,<addr> 1102 Start an early, polled-mode console on an msm serial 1103 port at the specified address. The serial port 1104 must already be setup and configured. Options are not 1105 yet supported. 1106 1107 msm_serial_dm,<addr> 1108 Start an early, polled-mode console on an msm serial 1109 dm port at the specified address. The serial port 1110 must already be setup and configured. Options are not 1111 yet supported. 1112 1113 owl,<addr> 1114 Start an early, polled-mode console on a serial port 1115 of an Actions Semi SoC, such as S500 or S900, at the 1116 specified address. The serial port must already be 1117 setup and configured. Options are not yet supported. 1118 1119 rda,<addr> 1120 Start an early, polled-mode console on a serial port 1121 of an RDA Micro SoC, such as RDA8810PL, at the 1122 specified address. The serial port must already be 1123 setup and configured. Options are not yet supported. 1124 1125 sbi 1126 Use RISC-V SBI (Supervisor Binary Interface) for early 1127 console. 1128 1129 smh Use ARM semihosting calls for early console. 1130 1131 s3c2410,<addr> 1132 s3c2412,<addr> 1133 s3c2440,<addr> 1134 s3c6400,<addr> 1135 s5pv210,<addr> 1136 exynos4210,<addr> 1137 Use early console provided by serial driver available 1138 on Samsung SoCs, requires selecting proper type and 1139 a correct base address of the selected UART port. The 1140 serial port must already be setup and configured. 1141 Options are not yet supported. 1142 1143 lantiq,<addr> 1144 Start an early, polled-mode console on a lantiq serial 1145 (lqasc) port at the specified address. The serial port 1146 must already be setup and configured. Options are not 1147 yet supported. 1148 1149 lpuart,<addr> 1150 lpuart32,<addr> 1151 Use early console provided by Freescale LP UART driver 1152 found on Freescale Vybrid and QorIQ LS1021A processors. 1153 A valid base address must be provided, and the serial 1154 port must already be setup and configured. 1155 1156 ec_imx21,<addr> 1157 ec_imx6q,<addr> 1158 Start an early, polled-mode, output-only console on the 1159 Freescale i.MX UART at the specified address. The UART 1160 must already be setup and configured. 1161 1162 ar3700_uart,<addr> 1163 Start an early, polled-mode console on the 1164 Armada 3700 serial port at the specified 1165 address. The serial port must already be setup 1166 and configured. Options are not yet supported. 1167 1168 qcom_geni,<addr> 1169 Start an early, polled-mode console on a Qualcomm 1170 Generic Interface (GENI) based serial port at the 1171 specified address. The serial port must already be 1172 setup and configured. Options are not yet supported. 1173 1174 efifb,[options] 1175 Start an early, unaccelerated console on the EFI 1176 memory mapped framebuffer (if available). On cache 1177 coherent non-x86 systems that use system memory for 1178 the framebuffer, pass the 'ram' option so that it is 1179 mapped with the correct attributes. 1180 1181 linflex,<addr> 1182 Use early console provided by Freescale LINFlexD UART 1183 serial driver for NXP S32V234 SoCs. A valid base 1184 address must be provided, and the serial port must 1185 already be setup and configured. 1186 1187 earlyprintk= [X86,SH,ARM,M68k,S390] 1188 earlyprintk=vga 1189 earlyprintk=sclp 1190 earlyprintk=xen 1191 earlyprintk=serial[,ttySn[,baudrate]] 1192 earlyprintk=serial[,0x...[,baudrate]] 1193 earlyprintk=ttySn[,baudrate] 1194 earlyprintk=dbgp[debugController#] 1195 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1196 earlyprintk=xdbc[xhciController#] 1197 1198 earlyprintk is useful when the kernel crashes before 1199 the normal console is initialized. It is not enabled by 1200 default because it has some cosmetic problems. 1201 1202 Append ",keep" to not disable it when the real console 1203 takes over. 1204 1205 Only one of vga, efi, serial, or usb debug port can 1206 be used at a time. 1207 1208 Currently only ttyS0 and ttyS1 may be specified by 1209 name. Other I/O ports may be explicitly specified 1210 on some architectures (x86 and arm at least) by 1211 replacing ttySn with an I/O port address, like this: 1212 earlyprintk=serial,0x1008,115200 1213 You can find the port for a given device in 1214 /proc/tty/driver/serial: 1215 2: uart:ST16650V2 port:00001008 irq:18 ... 1216 1217 Interaction with the standard serial driver is not 1218 very good. 1219 1220 The VGA and EFI output is eventually overwritten by 1221 the real console. 1222 1223 The xen output can only be used by Xen PV guests. 1224 1225 The sclp output can only be used on s390. 1226 1227 The optional "force" to "pciserial" enables use of a 1228 PCI device even when its classcode is not of the 1229 UART class. 1230 1231 edac_report= [HW,EDAC] Control how to report EDAC event 1232 Format: {"on" | "off" | "force"} 1233 on: enable EDAC to report H/W event. May be overridden 1234 by other higher priority error reporting module. 1235 off: disable H/W event reporting through EDAC. 1236 force: enforce the use of EDAC to report H/W event. 1237 default: on. 1238 1239 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1240 ekgdboc=kbd 1241 1242 This is designed to be used in conjunction with 1243 the boot argument: earlyprintk=vga 1244 1245 This parameter works in place of the kgdboc parameter 1246 but can only be used if the backing tty is available 1247 very early in the boot process. For early debugging 1248 via a serial port see kgdboc_earlycon instead. 1249 1250 edd= [EDD] 1251 Format: {"off" | "on" | "skip[mbr]"} 1252 1253 efi= [EFI] 1254 Format: { "debug", "disable_early_pci_dma", 1255 "nochunk", "noruntime", "nosoftreserve", 1256 "novamap", "no_disable_early_pci_dma" } 1257 debug: enable misc debug output. 1258 disable_early_pci_dma: disable the busmaster bit on all 1259 PCI bridges while in the EFI boot stub. 1260 nochunk: disable reading files in "chunks" in the EFI 1261 boot stub, as chunking can cause problems with some 1262 firmware implementations. 1263 noruntime : disable EFI runtime services support 1264 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1265 attribute may cause the kernel to reserve the 1266 memory range for a memory mapping driver to 1267 claim. Specify efi=nosoftreserve to disable this 1268 reservation and treat the memory by its base type 1269 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1270 novamap: do not call SetVirtualAddressMap(). 1271 no_disable_early_pci_dma: Leave the busmaster bit set 1272 on all PCI bridges while in the EFI boot stub 1273 1274 efi_no_storage_paranoia [EFI; X86] 1275 Using this parameter you can use more than 50% of 1276 your efi variable storage. Use this parameter only if 1277 you are really sure that your UEFI does sane gc and 1278 fulfills the spec otherwise your board may brick. 1279 1280 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1281 Add arbitrary attribute to specific memory range by 1282 updating original EFI memory map. 1283 Region of memory which aa attribute is added to is 1284 from ss to ss+nn. 1285 1286 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1287 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1288 attribute is added to range 0x100000000-0x180000000 and 1289 0x10a0000000-0x1120000000. 1290 1291 If efi_fake_mem=8G@9G:0x40000 is specified, the 1292 EFI_MEMORY_SP(0x40000) attribute is added to 1293 range 0x240000000-0x43fffffff. 1294 1295 Using this parameter you can do debugging of EFI memmap 1296 related features. For example, you can do debugging of 1297 Address Range Mirroring feature even if your box 1298 doesn't support it, or mark specific memory as 1299 "soft reserved". 1300 1301 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1302 that is to be dynamically loaded by Linux. If there are 1303 multiple variables with the same name but with different 1304 vendor GUIDs, all of them will be loaded. See 1305 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1306 1307 1308 eisa_irq_edge= [PARISC,HW] 1309 See header of drivers/parisc/eisa.c. 1310 1311 elanfreq= [X86-32] 1312 See comment before function elanfreq_setup() in 1313 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1314 1315 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1316 Specifies physical address of start of kernel core 1317 image elf header and optionally the size. Generally 1318 kexec loader will pass this option to capture kernel. 1319 See Documentation/admin-guide/kdump/kdump.rst for details. 1320 1321 enable_mtrr_cleanup [X86] 1322 The kernel tries to adjust MTRR layout from continuous 1323 to discrete, to make X server driver able to add WB 1324 entry later. This parameter enables that. 1325 1326 enable_timer_pin_1 [X86] 1327 Enable PIN 1 of APIC timer 1328 Can be useful to work around chipset bugs 1329 (in particular on some ATI chipsets). 1330 The kernel tries to set a reasonable default. 1331 1332 enforcing [SELINUX] Set initial enforcing status. 1333 Format: {"0" | "1"} 1334 See security/selinux/Kconfig help text. 1335 0 -- permissive (log only, no denials). 1336 1 -- enforcing (deny and log). 1337 Default value is 0. 1338 Value can be changed at runtime via 1339 /sys/fs/selinux/enforce. 1340 1341 erst_disable [ACPI] 1342 Disable Error Record Serialization Table (ERST) 1343 support. 1344 1345 ether= [HW,NET] Ethernet cards parameters 1346 This option is obsoleted by the "netdev=" option, which 1347 has equivalent usage. See its documentation for details. 1348 1349 evm= [EVM] 1350 Format: { "fix" } 1351 Permit 'security.evm' to be updated regardless of 1352 current integrity status. 1353 1354 failslab= 1355 fail_usercopy= 1356 fail_page_alloc= 1357 fail_make_request=[KNL] 1358 General fault injection mechanism. 1359 Format: <interval>,<probability>,<space>,<times> 1360 See also Documentation/fault-injection/. 1361 1362 fb_tunnels= [NET] 1363 Format: { initns | none } 1364 See Documentation/admin-guide/sysctl/net.rst for 1365 fb_tunnels_only_for_init_ns 1366 1367 floppy= [HW] 1368 See Documentation/admin-guide/blockdev/floppy.rst. 1369 1370 force_pal_cache_flush 1371 [IA-64] Avoid check_sal_cache_flush which may hang on 1372 buggy SAL_CACHE_FLUSH implementations. Using this 1373 parameter will force ia64_sal_cache_flush to call 1374 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1375 1376 forcepae [X86-32] 1377 Forcefully enable Physical Address Extension (PAE). 1378 Many Pentium M systems disable PAE but may have a 1379 functionally usable PAE implementation. 1380 Warning: use of this parameter will taint the kernel 1381 and may cause unknown problems. 1382 1383 ftrace=[tracer] 1384 [FTRACE] will set and start the specified tracer 1385 as early as possible in order to facilitate early 1386 boot debugging. 1387 1388 ftrace_dump_on_oops[=orig_cpu] 1389 [FTRACE] will dump the trace buffers on oops. 1390 If no parameter is passed, ftrace will dump 1391 buffers of all CPUs, but if you pass orig_cpu, it will 1392 dump only the buffer of the CPU that triggered the 1393 oops. 1394 1395 ftrace_filter=[function-list] 1396 [FTRACE] Limit the functions traced by the function 1397 tracer at boot up. function-list is a comma-separated 1398 list of functions. This list can be changed at run 1399 time by the set_ftrace_filter file in the debugfs 1400 tracing directory. 1401 1402 ftrace_notrace=[function-list] 1403 [FTRACE] Do not trace the functions specified in 1404 function-list. This list can be changed at run time 1405 by the set_ftrace_notrace file in the debugfs 1406 tracing directory. 1407 1408 ftrace_graph_filter=[function-list] 1409 [FTRACE] Limit the top level callers functions traced 1410 by the function graph tracer at boot up. 1411 function-list is a comma-separated list of functions 1412 that can be changed at run time by the 1413 set_graph_function file in the debugfs tracing directory. 1414 1415 ftrace_graph_notrace=[function-list] 1416 [FTRACE] Do not trace from the functions specified in 1417 function-list. This list is a comma-separated list of 1418 functions that can be changed at run time by the 1419 set_graph_notrace file in the debugfs tracing directory. 1420 1421 ftrace_graph_max_depth=<uint> 1422 [FTRACE] Used with the function graph tracer. This is 1423 the max depth it will trace into a function. This value 1424 can be changed at run time by the max_graph_depth file 1425 in the tracefs tracing directory. default: 0 (no limit) 1426 1427 fw_devlink= [KNL] Create device links between consumer and supplier 1428 devices by scanning the firmware to infer the 1429 consumer/supplier relationships. This feature is 1430 especially useful when drivers are loaded as modules as 1431 it ensures proper ordering of tasks like device probing 1432 (suppliers first, then consumers), supplier boot state 1433 clean up (only after all consumers have probed), 1434 suspend/resume & runtime PM (consumers first, then 1435 suppliers). 1436 Format: { off | permissive | on | rpm } 1437 off -- Don't create device links from firmware info. 1438 permissive -- Create device links from firmware info 1439 but use it only for ordering boot state clean 1440 up (sync_state() calls). 1441 on -- Create device links from firmware info and use it 1442 to enforce probe and suspend/resume ordering. 1443 rpm -- Like "on", but also use to order runtime PM. 1444 1445 fw_devlink.strict=<bool> 1446 [KNL] Treat all inferred dependencies as mandatory 1447 dependencies. This only applies for fw_devlink=on|rpm. 1448 Format: <bool> 1449 1450 gamecon.map[2|3]= 1451 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1452 support via parallel port (up to 5 devices per port) 1453 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1454 See also Documentation/input/devices/joystick-parport.rst 1455 1456 gamma= [HW,DRM] 1457 1458 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART 1459 Format: off | on 1460 default: on 1461 1462 gcov_persist= [GCOV] When non-zero (default), profiling data for 1463 kernel modules is saved and remains accessible via 1464 debugfs, even when the module is unloaded/reloaded. 1465 When zero, profiling data is discarded and associated 1466 debugfs files are removed at module unload time. 1467 1468 goldfish [X86] Enable the goldfish android emulator platform. 1469 Don't use this when you are not running on the 1470 android emulator 1471 1472 gpio-mockup.gpio_mockup_ranges 1473 [HW] Sets the ranges of gpiochip of for this device. 1474 Format: <start1>,<end1>,<start2>,<end2>... 1475 gpio-mockup.gpio_mockup_named_lines 1476 [HW] Let the driver know GPIO lines should be named. 1477 1478 gpt [EFI] Forces disk with valid GPT signature but 1479 invalid Protective MBR to be treated as GPT. If the 1480 primary GPT is corrupted, it enables the backup/alternate 1481 GPT to be used instead. 1482 1483 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1484 the "Enable 0" bit of the configuration register. 1485 Format: 0 | 1 1486 Default: 0 1487 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1488 the "Enable 0" bit of the configuration register. 1489 Format: 0 | 1 1490 Default: 0 1491 grcan.select= [HW] Select which physical interface to use. 1492 Format: 0 | 1 1493 Default: 0 1494 grcan.txsize= [HW] Sets the size of the tx buffer. 1495 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1496 Default: 1024 1497 grcan.rxsize= [HW] Sets the size of the rx buffer. 1498 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1499 Default: 1024 1500 1501 hardlockup_all_cpu_backtrace= 1502 [KNL] Should the hard-lockup detector generate 1503 backtraces on all cpus. 1504 Format: 0 | 1 1505 1506 hashdist= [KNL,NUMA] Large hashes allocated during boot 1507 are distributed across NUMA nodes. Defaults on 1508 for 64-bit NUMA, off otherwise. 1509 Format: 0 | 1 (for off | on) 1510 1511 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1512 1513 hd= [EIDE] (E)IDE hard drive subsystem geometry 1514 Format: <cyl>,<head>,<sect> 1515 1516 hest_disable [ACPI] 1517 Disable Hardware Error Source Table (HEST) support; 1518 corresponding firmware-first mode error processing 1519 logic will be disabled. 1520 1521 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1522 size of <nn>. This works even on boxes that have no 1523 highmem otherwise. This also works to reduce highmem 1524 size on bigger boxes. 1525 1526 highres= [KNL] Enable/disable high resolution timer mode. 1527 Valid parameters: "on", "off" 1528 Default: "on" 1529 1530 hlt [BUGS=ARM,SH] 1531 1532 hpet= [X86-32,HPET] option to control HPET usage 1533 Format: { enable (default) | disable | force | 1534 verbose } 1535 disable: disable HPET and use PIT instead 1536 force: allow force enabled of undocumented chips (ICH4, 1537 VIA, nVidia) 1538 verbose: show contents of HPET registers during setup 1539 1540 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1541 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1542 1543 hugetlb_cma= [HW,CMA] The size of a CMA area used for allocation 1544 of gigantic hugepages. 1545 Format: nn[KMGTPE] 1546 1547 Reserve a CMA area of given size and allocate gigantic 1548 hugepages using the CMA allocator. If enabled, the 1549 boot-time allocation of gigantic hugepages is skipped. 1550 1551 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1552 If this follows hugepagesz (below), it specifies 1553 the number of pages of hugepagesz to be allocated. 1554 If this is the first HugeTLB parameter on the command 1555 line, it specifies the number of pages to allocate for 1556 the default huge page size. See also 1557 Documentation/admin-guide/mm/hugetlbpage.rst. 1558 Format: <integer> 1559 1560 hugepagesz= 1561 [HW] The size of the HugeTLB pages. This is used in 1562 conjunction with hugepages (above) to allocate huge 1563 pages of a specific size at boot. The pair 1564 hugepagesz=X hugepages=Y can be specified once for 1565 each supported huge page size. Huge page sizes are 1566 architecture dependent. See also 1567 Documentation/admin-guide/mm/hugetlbpage.rst. 1568 Format: size[KMG] 1569 1570 hugetlb_free_vmemmap= 1571 [KNL] Reguires CONFIG_HUGETLB_PAGE_FREE_VMEMMAP 1572 enabled. 1573 Allows heavy hugetlb users to free up some more 1574 memory (6 * PAGE_SIZE for each 2MB hugetlb page). 1575 Format: { on | off (default) } 1576 1577 on: enable the feature 1578 off: disable the feature 1579 1580 Built with CONFIG_HUGETLB_PAGE_FREE_VMEMMAP_DEFAULT_ON=y, 1581 the default is on. 1582 1583 This is not compatible with memory_hotplug.memmap_on_memory. 1584 If both parameters are enabled, hugetlb_free_vmemmap takes 1585 precedence over memory_hotplug.memmap_on_memory. 1586 1587 hung_task_panic= 1588 [KNL] Should the hung task detector generate panics. 1589 Format: 0 | 1 1590 1591 A value of 1 instructs the kernel to panic when a 1592 hung task is detected. The default value is controlled 1593 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1594 option. The value selected by this boot parameter can 1595 be changed later by the kernel.hung_task_panic sysctl. 1596 1597 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1598 terminal devices. Valid values: 0..8 1599 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1600 If specified, z/VM IUCV HVC accepts connections 1601 from listed z/VM user IDs only. 1602 1603 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1604 which allow the hypervisor to 'idle' the 1605 guest on lock contention. 1606 1607 keep_bootcon [KNL] 1608 Do not unregister boot console at start. This is only 1609 useful for debugging when something happens in the window 1610 between unregistering the boot console and initializing 1611 the real console. 1612 1613 i2c_bus= [HW] Override the default board specific I2C bus speed 1614 or register an additional I2C bus that is not 1615 registered from board initialization code. 1616 Format: 1617 <bus_id>,<clkrate> 1618 1619 i8042.debug [HW] Toggle i8042 debug mode 1620 i8042.unmask_kbd_data 1621 [HW] Enable printing of interrupt data from the KBD port 1622 (disabled by default, and as a pre-condition 1623 requires that i8042.debug=1 be enabled) 1624 i8042.direct [HW] Put keyboard port into non-translated mode 1625 i8042.dumbkbd [HW] Pretend that controller can only read data from 1626 keyboard and cannot control its state 1627 (Don't attempt to blink the leds) 1628 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1629 i8042.nokbd [HW] Don't check/create keyboard port 1630 i8042.noloop [HW] Disable the AUX Loopback command while probing 1631 for the AUX port 1632 i8042.nomux [HW] Don't check presence of an active multiplexing 1633 controller 1634 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1635 controllers 1636 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1637 i8042.reset [HW] Reset the controller during init, cleanup and 1638 suspend-to-ram transitions, only during s2r 1639 transitions, or never reset 1640 Format: { 1 | Y | y | 0 | N | n } 1641 1, Y, y: always reset controller 1642 0, N, n: don't ever reset controller 1643 Default: only on s2r transitions on x86; most other 1644 architectures force reset to be always executed 1645 i8042.unlock [HW] Unlock (ignore) the keylock 1646 i8042.kbdreset [HW] Reset device connected to KBD port 1647 1648 i810= [HW,DRM] 1649 1650 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1651 indicates that the driver is running on unsupported 1652 hardware. 1653 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1654 does not match list of supported models. 1655 i8k.power_status 1656 [HW] Report power status in /proc/i8k 1657 (disabled by default) 1658 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1659 capability is set. 1660 1661 i915.invert_brightness= 1662 [DRM] Invert the sense of the variable that is used to 1663 set the brightness of the panel backlight. Normally a 1664 brightness value of 0 indicates backlight switched off, 1665 and the maximum of the brightness value sets the backlight 1666 to maximum brightness. If this parameter is set to 0 1667 (default) and the machine requires it, or this parameter 1668 is set to 1, a brightness value of 0 sets the backlight 1669 to maximum brightness, and the maximum of the brightness 1670 value switches the backlight off. 1671 -1 -- never invert brightness 1672 0 -- machine default 1673 1 -- force brightness inversion 1674 1675 icn= [HW,ISDN] 1676 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1677 1678 ide-core.nodma= [HW] (E)IDE subsystem 1679 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1680 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1681 .cdrom .chs .ignore_cable are additional options 1682 See Documentation/ide/ide.rst. 1683 1684 ide-generic.probe-mask= [HW] (E)IDE subsystem 1685 Format: <int> 1686 Probe mask for legacy ISA IDE ports. Depending on 1687 platform up to 6 ports are supported, enabled by 1688 setting corresponding bits in the mask to 1. The 1689 default value is 0x0, which has a special meaning. 1690 On systems that have PCI, it triggers scanning the 1691 PCI bus for the first and the second port, which 1692 are then probed. On systems without PCI the value 1693 of 0x0 enables probing the two first ports as if it 1694 was 0x3. 1695 1696 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1697 Claim all unknown PCI IDE storage controllers. 1698 1699 idle= [X86] 1700 Format: idle=poll, idle=halt, idle=nomwait 1701 Poll forces a polling idle loop that can slightly 1702 improve the performance of waking up a idle CPU, but 1703 will use a lot of power and make the system run hot. 1704 Not recommended. 1705 idle=halt: Halt is forced to be used for CPU idle. 1706 In such case C2/C3 won't be used again. 1707 idle=nomwait: Disable mwait for CPU C-states 1708 1709 idxd.sva= [HW] 1710 Format: <bool> 1711 Allow force disabling of Shared Virtual Memory (SVA) 1712 support for the idxd driver. By default it is set to 1713 true (1). 1714 1715 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1716 Format: { strict | legacy | 2008 | relaxed } 1717 Default: strict 1718 1719 Choose which programs will be accepted for execution 1720 based on the IEEE 754 NaN encoding(s) supported by 1721 the FPU and the NaN encoding requested with the value 1722 of an ELF file header flag individually set by each 1723 binary. Hardware implementations are permitted to 1724 support either or both of the legacy and the 2008 NaN 1725 encoding mode. 1726 1727 Available settings are as follows: 1728 strict accept binaries that request a NaN encoding 1729 supported by the FPU 1730 legacy only accept legacy-NaN binaries, if supported 1731 by the FPU 1732 2008 only accept 2008-NaN binaries, if supported 1733 by the FPU 1734 relaxed accept any binaries regardless of whether 1735 supported by the FPU 1736 1737 The FPU emulator is always able to support both NaN 1738 encodings, so if no FPU hardware is present or it has 1739 been disabled with 'nofpu', then the settings of 1740 'legacy' and '2008' strap the emulator accordingly, 1741 'relaxed' straps the emulator for both legacy-NaN and 1742 2008-NaN, whereas 'strict' enables legacy-NaN only on 1743 legacy processors and both NaN encodings on MIPS32 or 1744 MIPS64 CPUs. 1745 1746 The setting for ABS.fmt/NEG.fmt instruction execution 1747 mode generally follows that for the NaN encoding, 1748 except where unsupported by hardware. 1749 1750 ignore_loglevel [KNL] 1751 Ignore loglevel setting - this will print /all/ 1752 kernel messages to the console. Useful for debugging. 1753 We also add it as printk module parameter, so users 1754 could change it dynamically, usually by 1755 /sys/module/printk/parameters/ignore_loglevel. 1756 1757 ignore_rlimit_data 1758 Ignore RLIMIT_DATA setting for data mappings, 1759 print warning at first misuse. Can be changed via 1760 /sys/module/kernel/parameters/ignore_rlimit_data. 1761 1762 ihash_entries= [KNL] 1763 Set number of hash buckets for inode cache. 1764 1765 ima_appraise= [IMA] appraise integrity measurements 1766 Format: { "off" | "enforce" | "fix" | "log" } 1767 default: "enforce" 1768 1769 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1770 The builtin appraise policy appraises all files 1771 owned by uid=0. 1772 1773 ima_canonical_fmt [IMA] 1774 Use the canonical format for the binary runtime 1775 measurements, instead of host native format. 1776 1777 ima_hash= [IMA] 1778 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1779 | sha512 | ... } 1780 default: "sha1" 1781 1782 The list of supported hash algorithms is defined 1783 in crypto/hash_info.h. 1784 1785 ima_policy= [IMA] 1786 The builtin policies to load during IMA setup. 1787 Format: "tcb | appraise_tcb | secure_boot | 1788 fail_securely | critical_data" 1789 1790 The "tcb" policy measures all programs exec'd, files 1791 mmap'd for exec, and all files opened with the read 1792 mode bit set by either the effective uid (euid=0) or 1793 uid=0. 1794 1795 The "appraise_tcb" policy appraises the integrity of 1796 all files owned by root. 1797 1798 The "secure_boot" policy appraises the integrity 1799 of files (eg. kexec kernel image, kernel modules, 1800 firmware, policy, etc) based on file signatures. 1801 1802 The "fail_securely" policy forces file signature 1803 verification failure also on privileged mounted 1804 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1805 flag. 1806 1807 The "critical_data" policy measures kernel integrity 1808 critical data. 1809 1810 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1811 Load a policy which meets the needs of the Trusted 1812 Computing Base. This means IMA will measure all 1813 programs exec'd, files mmap'd for exec, and all files 1814 opened for read by uid=0. 1815 1816 ima_template= [IMA] 1817 Select one of defined IMA measurements template formats. 1818 Formats: { "ima" | "ima-ng" | "ima-sig" } 1819 Default: "ima-ng" 1820 1821 ima_template_fmt= 1822 [IMA] Define a custom template format. 1823 Format: { "field1|...|fieldN" } 1824 1825 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1826 Format: <min_file_size> 1827 Set the minimal file size for using asynchronous hash. 1828 If left unspecified, ahash usage is disabled. 1829 1830 ahash performance varies for different data sizes on 1831 different crypto accelerators. This option can be used 1832 to achieve the best performance for a particular HW. 1833 1834 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1835 Format: <bufsize> 1836 Set hashing buffer size. Default: 4k. 1837 1838 ahash performance varies for different chunk sizes on 1839 different crypto accelerators. This option can be used 1840 to achieve best performance for particular HW. 1841 1842 init= [KNL] 1843 Format: <full_path> 1844 Run specified binary instead of /sbin/init as init 1845 process. 1846 1847 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1848 for working out where the kernel is dying during 1849 startup. 1850 1851 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1852 initcall functions. Useful for debugging built-in 1853 modules and initcalls. 1854 1855 initramfs_async= [KNL] 1856 Format: <bool> 1857 Default: 1 1858 This parameter controls whether the initramfs 1859 image is unpacked asynchronously, concurrently 1860 with devices being probed and 1861 initialized. This should normally just work, 1862 but as a debugging aid, one can get the 1863 historical behaviour of the initramfs 1864 unpacking being completed before device_ and 1865 late_ initcalls. 1866 1867 initrd= [BOOT] Specify the location of the initial ramdisk 1868 1869 initrdmem= [KNL] Specify a physical address and size from which to 1870 load the initrd. If an initrd is compiled in or 1871 specified in the bootparams, it takes priority over this 1872 setting. 1873 Format: ss[KMG],nn[KMG] 1874 Default is 0, 0 1875 1876 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1877 zeroes. 1878 Format: 0 | 1 1879 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1880 1881 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1882 Format: 0 | 1 1883 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1884 1885 init_pkru= [X86] Specify the default memory protection keys rights 1886 register contents for all processes. 0x55555554 by 1887 default (disallow access to all but pkey 0). Can 1888 override in debugfs after boot. 1889 1890 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1891 Format: <irq> 1892 1893 int_pln_enable [X86] Enable power limit notification interrupt 1894 1895 integrity_audit=[IMA] 1896 Format: { "0" | "1" } 1897 0 -- basic integrity auditing messages. (Default) 1898 1 -- additional integrity auditing messages. 1899 1900 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1901 on 1902 Enable intel iommu driver. 1903 off 1904 Disable intel iommu driver. 1905 igfx_off [Default Off] 1906 By default, gfx is mapped as normal device. If a gfx 1907 device has a dedicated DMAR unit, the DMAR unit is 1908 bypassed by not enabling DMAR with this option. In 1909 this case, gfx device will use physical address for 1910 DMA. 1911 strict [Default Off] 1912 With this option on every unmap_single operation will 1913 result in a hardware IOTLB flush operation as opposed 1914 to batching them for performance. 1915 sp_off [Default Off] 1916 By default, super page will be supported if Intel IOMMU 1917 has the capability. With this option, super page will 1918 not be supported. 1919 sm_on [Default Off] 1920 By default, scalable mode will be disabled even if the 1921 hardware advertises that it has support for the scalable 1922 mode translation. With this option set, scalable mode 1923 will be used on hardware which claims to support it. 1924 tboot_noforce [Default Off] 1925 Do not force the Intel IOMMU enabled under tboot. 1926 By default, tboot will force Intel IOMMU on, which 1927 could harm performance of some high-throughput 1928 devices like 40GBit network cards, even if identity 1929 mapping is enabled. 1930 Note that using this option lowers the security 1931 provided by tboot because it makes the system 1932 vulnerable to DMA attacks. 1933 1934 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1935 0 disables intel_idle and fall back on acpi_idle. 1936 1 to 9 specify maximum depth of C-state. 1937 1938 intel_pstate= [X86] 1939 disable 1940 Do not enable intel_pstate as the default 1941 scaling driver for the supported processors 1942 passive 1943 Use intel_pstate as a scaling driver, but configure it 1944 to work with generic cpufreq governors (instead of 1945 enabling its internal governor). This mode cannot be 1946 used along with the hardware-managed P-states (HWP) 1947 feature. 1948 force 1949 Enable intel_pstate on systems that prohibit it by default 1950 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1951 instead of acpi-cpufreq may disable platform features, such 1952 as thermal controls and power capping, that rely on ACPI 1953 P-States information being indicated to OSPM and therefore 1954 should be used with caution. This option does not work with 1955 processors that aren't supported by the intel_pstate driver 1956 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1957 no_hwp 1958 Do not enable hardware P state control (HWP) 1959 if available. 1960 hwp_only 1961 Only load intel_pstate on systems which support 1962 hardware P state control (HWP) if available. 1963 support_acpi_ppc 1964 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1965 Description Table, specifies preferred power management 1966 profile as "Enterprise Server" or "Performance Server", 1967 then this feature is turned on by default. 1968 per_cpu_perf_limits 1969 Allow per-logical-CPU P-State performance control limits using 1970 cpufreq sysfs interface 1971 1972 intremap= [X86-64, Intel-IOMMU] 1973 on enable Interrupt Remapping (default) 1974 off disable Interrupt Remapping 1975 nosid disable Source ID checking 1976 no_x2apic_optout 1977 BIOS x2APIC opt-out request will be ignored 1978 nopost disable Interrupt Posting 1979 1980 iomem= Disable strict checking of access to MMIO memory 1981 strict regions from userspace. 1982 relaxed 1983 1984 iommu= [X86] 1985 off 1986 force 1987 noforce 1988 biomerge 1989 panic 1990 nopanic 1991 merge 1992 nomerge 1993 soft 1994 pt [X86] 1995 nopt [X86] 1996 nobypass [PPC/POWERNV] 1997 Disable IOMMU bypass, using IOMMU for PCI devices. 1998 1999 iommu.forcedac= [ARM64, X86] Control IOVA allocation for PCI devices. 2000 Format: { "0" | "1" } 2001 0 - Try to allocate a 32-bit DMA address first, before 2002 falling back to the full range if needed. 2003 1 - Allocate directly from the full usable range, 2004 forcing Dual Address Cycle for PCI cards supporting 2005 greater than 32-bit addressing. 2006 2007 iommu.strict= [ARM64] Configure TLB invalidation behaviour 2008 Format: { "0" | "1" } 2009 0 - Lazy mode. 2010 Request that DMA unmap operations use deferred 2011 invalidation of hardware TLBs, for increased 2012 throughput at the cost of reduced device isolation. 2013 Will fall back to strict mode if not supported by 2014 the relevant IOMMU driver. 2015 1 - Strict mode (default). 2016 DMA unmap operations invalidate IOMMU hardware TLBs 2017 synchronously. 2018 2019 iommu.passthrough= 2020 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 2021 Format: { "0" | "1" } 2022 0 - Use IOMMU translation for DMA. 2023 1 - Bypass the IOMMU for DMA. 2024 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2025 2026 io7= [HW] IO7 for Marvel-based Alpha systems 2027 See comment before marvel_specify_io7 in 2028 arch/alpha/kernel/core_marvel.c. 2029 2030 io_delay= [X86] I/O delay method 2031 0x80 2032 Standard port 0x80 based delay 2033 0xed 2034 Alternate port 0xed based delay (needed on some systems) 2035 udelay 2036 Simple two microseconds delay 2037 none 2038 No delay 2039 2040 ip= [IP_PNP] 2041 See Documentation/admin-guide/nfs/nfsroot.rst. 2042 2043 ipcmni_extend [KNL] Extend the maximum number of unique System V 2044 IPC identifiers from 32,768 to 16,777,216. 2045 2046 irqaffinity= [SMP] Set the default irq affinity mask 2047 The argument is a cpu list, as described above. 2048 2049 irqchip.gicv2_force_probe= 2050 [ARM, ARM64] 2051 Format: <bool> 2052 Force the kernel to look for the second 4kB page 2053 of a GICv2 controller even if the memory range 2054 exposed by the device tree is too small. 2055 2056 irqchip.gicv3_nolpi= 2057 [ARM, ARM64] 2058 Force the kernel to ignore the availability of 2059 LPIs (and by consequence ITSs). Intended for system 2060 that use the kernel as a bootloader, and thus want 2061 to let secondary kernels in charge of setting up 2062 LPIs. 2063 2064 irqchip.gicv3_pseudo_nmi= [ARM64] 2065 Enables support for pseudo-NMIs in the kernel. This 2066 requires the kernel to be built with 2067 CONFIG_ARM64_PSEUDO_NMI. 2068 2069 irqfixup [HW] 2070 When an interrupt is not handled search all handlers 2071 for it. Intended to get systems with badly broken 2072 firmware running. 2073 2074 irqpoll [HW] 2075 When an interrupt is not handled search all handlers 2076 for it. Also check all handlers each timer 2077 interrupt. Intended to get systems with badly broken 2078 firmware running. 2079 2080 isapnp= [ISAPNP] 2081 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2082 2083 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2084 [Deprecated - use cpusets instead] 2085 Format: [flag-list,]<cpu-list> 2086 2087 Specify one or more CPUs to isolate from disturbances 2088 specified in the flag list (default: domain): 2089 2090 nohz 2091 Disable the tick when a single task runs. 2092 2093 A residual 1Hz tick is offloaded to workqueues, which you 2094 need to affine to housekeeping through the global 2095 workqueue's affinity configured via the 2096 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2097 by using the 'domain' flag described below. 2098 2099 NOTE: by default the global workqueue runs on all CPUs, 2100 so to protect individual CPUs the 'cpumask' file has to 2101 be configured manually after bootup. 2102 2103 domain 2104 Isolate from the general SMP balancing and scheduling 2105 algorithms. Note that performing domain isolation this way 2106 is irreversible: it's not possible to bring back a CPU to 2107 the domains once isolated through isolcpus. It's strongly 2108 advised to use cpusets instead to disable scheduler load 2109 balancing through the "cpuset.sched_load_balance" file. 2110 It offers a much more flexible interface where CPUs can 2111 move in and out of an isolated set anytime. 2112 2113 You can move a process onto or off an "isolated" CPU via 2114 the CPU affinity syscalls or cpuset. 2115 <cpu number> begins at 0 and the maximum value is 2116 "number of CPUs in system - 1". 2117 2118 managed_irq 2119 2120 Isolate from being targeted by managed interrupts 2121 which have an interrupt mask containing isolated 2122 CPUs. The affinity of managed interrupts is 2123 handled by the kernel and cannot be changed via 2124 the /proc/irq/* interfaces. 2125 2126 This isolation is best effort and only effective 2127 if the automatically assigned interrupt mask of a 2128 device queue contains isolated and housekeeping 2129 CPUs. If housekeeping CPUs are online then such 2130 interrupts are directed to the housekeeping CPU 2131 so that IO submitted on the housekeeping CPU 2132 cannot disturb the isolated CPU. 2133 2134 If a queue's affinity mask contains only isolated 2135 CPUs then this parameter has no effect on the 2136 interrupt routing decision, though interrupts are 2137 only delivered when tasks running on those 2138 isolated CPUs submit IO. IO submitted on 2139 housekeeping CPUs has no influence on those 2140 queues. 2141 2142 The format of <cpu-list> is described above. 2143 2144 iucv= [HW,NET] 2145 2146 ivrs_ioapic [HW,X86-64] 2147 Provide an override to the IOAPIC-ID<->DEVICE-ID 2148 mapping provided in the IVRS ACPI table. For 2149 example, to map IOAPIC-ID decimal 10 to 2150 PCI device 00:14.0 write the parameter as: 2151 ivrs_ioapic[10]=00:14.0 2152 2153 ivrs_hpet [HW,X86-64] 2154 Provide an override to the HPET-ID<->DEVICE-ID 2155 mapping provided in the IVRS ACPI table. For 2156 example, to map HPET-ID decimal 0 to 2157 PCI device 00:14.0 write the parameter as: 2158 ivrs_hpet[0]=00:14.0 2159 2160 ivrs_acpihid [HW,X86-64] 2161 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2162 mapping provided in the IVRS ACPI table. For 2163 example, to map UART-HID:UID AMD0020:0 to 2164 PCI device 00:14.5 write the parameter as: 2165 ivrs_acpihid[00:14.5]=AMD0020:0 2166 2167 js= [HW,JOY] Analog joystick 2168 See Documentation/input/joydev/joystick.rst. 2169 2170 nokaslr [KNL] 2171 When CONFIG_RANDOMIZE_BASE is set, this disables 2172 kernel and module base offset ASLR (Address Space 2173 Layout Randomization). 2174 2175 kasan_multi_shot 2176 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2177 report on every invalid memory access. Without this 2178 parameter KASAN will print report only for the first 2179 invalid access. 2180 2181 keepinitrd [HW,ARM] 2182 2183 kernelcore= [KNL,X86,IA-64,PPC] 2184 Format: nn[KMGTPE] | nn% | "mirror" 2185 This parameter specifies the amount of memory usable by 2186 the kernel for non-movable allocations. The requested 2187 amount is spread evenly throughout all nodes in the 2188 system as ZONE_NORMAL. The remaining memory is used for 2189 movable memory in its own zone, ZONE_MOVABLE. In the 2190 event, a node is too small to have both ZONE_NORMAL and 2191 ZONE_MOVABLE, kernelcore memory will take priority and 2192 other nodes will have a larger ZONE_MOVABLE. 2193 2194 ZONE_MOVABLE is used for the allocation of pages that 2195 may be reclaimed or moved by the page migration 2196 subsystem. Note that allocations like PTEs-from-HighMem 2197 still use the HighMem zone if it exists, and the Normal 2198 zone if it does not. 2199 2200 It is possible to specify the exact amount of memory in 2201 the form of "nn[KMGTPE]", a percentage of total system 2202 memory in the form of "nn%", or "mirror". If "mirror" 2203 option is specified, mirrored (reliable) memory is used 2204 for non-movable allocations and remaining memory is used 2205 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2206 are exclusive, so you cannot specify multiple forms. 2207 2208 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2209 Format: <Controller#>[,poll interval] 2210 The controller # is the number of the ehci usb debug 2211 port as it is probed via PCI. The poll interval is 2212 optional and is the number seconds in between 2213 each poll cycle to the debug port in case you need 2214 the functionality for interrupting the kernel with 2215 gdb or control-c on the dbgp connection. When 2216 not using this parameter you use sysrq-g to break into 2217 the kernel debugger. 2218 2219 kgdboc= [KGDB,HW] kgdb over consoles. 2220 Requires a tty driver that supports console polling, 2221 or a supported polling keyboard driver (non-usb). 2222 Serial only format: <serial_device>[,baud] 2223 keyboard only format: kbd 2224 keyboard and serial format: kbd,<serial_device>[,baud] 2225 Optional Kernel mode setting: 2226 kms, kbd format: kms,kbd 2227 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2228 2229 kgdboc_earlycon= [KGDB,HW] 2230 If the boot console provides the ability to read 2231 characters and can work in polling mode, you can use 2232 this parameter to tell kgdb to use it as a backend 2233 until the normal console is registered. Intended to 2234 be used together with the kgdboc parameter which 2235 specifies the normal console to transition to. 2236 2237 The name of the early console should be specified 2238 as the value of this parameter. Note that the name of 2239 the early console might be different than the tty 2240 name passed to kgdboc. It's OK to leave the value 2241 blank and the first boot console that implements 2242 read() will be picked. 2243 2244 kgdbwait [KGDB] Stop kernel execution and enter the 2245 kernel debugger at the earliest opportunity. 2246 2247 kmac= [MIPS] Korina ethernet MAC address. 2248 Configure the RouterBoard 532 series on-chip 2249 Ethernet adapter MAC address. 2250 2251 kmemleak= [KNL] Boot-time kmemleak enable/disable 2252 Valid arguments: on, off 2253 Default: on 2254 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2255 the default is off. 2256 2257 kprobe_event=[probe-list] 2258 [FTRACE] Add kprobe events and enable at boot time. 2259 The probe-list is a semicolon delimited list of probe 2260 definitions. Each definition is same as kprobe_events 2261 interface, but the parameters are comma delimited. 2262 For example, to add a kprobe event on vfs_read with 2263 arg1 and arg2, add to the command line; 2264 2265 kprobe_event=p,vfs_read,$arg1,$arg2 2266 2267 See also Documentation/trace/kprobetrace.rst "Kernel 2268 Boot Parameter" section. 2269 2270 kpti= [ARM64] Control page table isolation of user 2271 and kernel address spaces. 2272 Default: enabled on cores which need mitigation. 2273 0: force disabled 2274 1: force enabled 2275 2276 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2277 Default is 0 (don't ignore, but inject #GP) 2278 2279 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2280 Default is false (don't support). 2281 2282 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2283 KVM MMU at runtime. 2284 Default is 0 (off) 2285 2286 kvm.nx_huge_pages= 2287 [KVM] Controls the software workaround for the 2288 X86_BUG_ITLB_MULTIHIT bug. 2289 force : Always deploy workaround. 2290 off : Never deploy workaround. 2291 auto : Deploy workaround based on the presence of 2292 X86_BUG_ITLB_MULTIHIT. 2293 2294 Default is 'auto'. 2295 2296 If the software workaround is enabled for the host, 2297 guests do need not to enable it for nested guests. 2298 2299 kvm.nx_huge_pages_recovery_ratio= 2300 [KVM] Controls how many 4KiB pages are periodically zapped 2301 back to huge pages. 0 disables the recovery, otherwise if 2302 the value is N KVM will zap 1/Nth of the 4KiB pages every 2303 minute. The default is 60. 2304 2305 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2306 Default is 1 (enabled) 2307 2308 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2309 for all guests. 2310 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2311 2312 kvm-arm.mode= 2313 [KVM,ARM] Select one of KVM/arm64's modes of operation. 2314 2315 nvhe: Standard nVHE-based mode, without support for 2316 protected guests. 2317 2318 protected: nVHE-based mode with support for guests whose 2319 state is kept private from the host. 2320 Not valid if the kernel is running in EL2. 2321 2322 Defaults to VHE/nVHE based on hardware support. 2323 2324 kvm-arm.vgic_v3_group0_trap= 2325 [KVM,ARM] Trap guest accesses to GICv3 group-0 2326 system registers 2327 2328 kvm-arm.vgic_v3_group1_trap= 2329 [KVM,ARM] Trap guest accesses to GICv3 group-1 2330 system registers 2331 2332 kvm-arm.vgic_v3_common_trap= 2333 [KVM,ARM] Trap guest accesses to GICv3 common 2334 system registers 2335 2336 kvm-arm.vgic_v4_enable= 2337 [KVM,ARM] Allow use of GICv4 for direct injection of 2338 LPIs. 2339 2340 kvm_cma_resv_ratio=n [PPC] 2341 Reserves given percentage from system memory area for 2342 contiguous memory allocation for KVM hash pagetable 2343 allocation. 2344 By default it reserves 5% of total system memory. 2345 Format: <integer> 2346 Default: 5 2347 2348 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2349 (virtualized MMU) support on capable Intel chips. 2350 Default is 1 (enabled) 2351 2352 kvm-intel.emulate_invalid_guest_state= 2353 [KVM,Intel] Enable emulation of invalid guest states 2354 Default is 0 (disabled) 2355 2356 kvm-intel.flexpriority= 2357 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2358 Default is 1 (enabled) 2359 2360 kvm-intel.nested= 2361 [KVM,Intel] Enable VMX nesting (nVMX). 2362 Default is 0 (disabled) 2363 2364 kvm-intel.unrestricted_guest= 2365 [KVM,Intel] Disable unrestricted guest feature 2366 (virtualized real and unpaged mode) on capable 2367 Intel chips. Default is 1 (enabled) 2368 2369 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2370 CVE-2018-3620. 2371 2372 Valid arguments: never, cond, always 2373 2374 always: L1D cache flush on every VMENTER. 2375 cond: Flush L1D on VMENTER only when the code between 2376 VMEXIT and VMENTER can leak host memory. 2377 never: Disables the mitigation 2378 2379 Default is cond (do L1 cache flush in specific instances) 2380 2381 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2382 feature (tagged TLBs) on capable Intel chips. 2383 Default is 1 (enabled) 2384 2385 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2386 affected CPUs 2387 2388 The kernel PTE inversion protection is unconditionally 2389 enabled and cannot be disabled. 2390 2391 full 2392 Provides all available mitigations for the 2393 L1TF vulnerability. Disables SMT and 2394 enables all mitigations in the 2395 hypervisors, i.e. unconditional L1D flush. 2396 2397 SMT control and L1D flush control via the 2398 sysfs interface is still possible after 2399 boot. Hypervisors will issue a warning 2400 when the first VM is started in a 2401 potentially insecure configuration, 2402 i.e. SMT enabled or L1D flush disabled. 2403 2404 full,force 2405 Same as 'full', but disables SMT and L1D 2406 flush runtime control. Implies the 2407 'nosmt=force' command line option. 2408 (i.e. sysfs control of SMT is disabled.) 2409 2410 flush 2411 Leaves SMT enabled and enables the default 2412 hypervisor mitigation, i.e. conditional 2413 L1D flush. 2414 2415 SMT control and L1D flush control via the 2416 sysfs interface is still possible after 2417 boot. Hypervisors will issue a warning 2418 when the first VM is started in a 2419 potentially insecure configuration, 2420 i.e. SMT enabled or L1D flush disabled. 2421 2422 flush,nosmt 2423 2424 Disables SMT and enables the default 2425 hypervisor mitigation. 2426 2427 SMT control and L1D flush control via the 2428 sysfs interface is still possible after 2429 boot. Hypervisors will issue a warning 2430 when the first VM is started in a 2431 potentially insecure configuration, 2432 i.e. SMT enabled or L1D flush disabled. 2433 2434 flush,nowarn 2435 Same as 'flush', but hypervisors will not 2436 warn when a VM is started in a potentially 2437 insecure configuration. 2438 2439 off 2440 Disables hypervisor mitigations and doesn't 2441 emit any warnings. 2442 It also drops the swap size and available 2443 RAM limit restriction on both hypervisor and 2444 bare metal. 2445 2446 Default is 'flush'. 2447 2448 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2449 2450 l2cr= [PPC] 2451 2452 l3cr= [PPC] 2453 2454 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2455 disabled it. 2456 2457 lapic= [X86,APIC] Do not use TSC deadline 2458 value for LAPIC timer one-shot implementation. Default 2459 back to the programmable timer unit in the LAPIC. 2460 Format: notscdeadline 2461 2462 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2463 in C2 power state. 2464 2465 libata.dma= [LIBATA] DMA control 2466 libata.dma=0 Disable all PATA and SATA DMA 2467 libata.dma=1 PATA and SATA Disk DMA only 2468 libata.dma=2 ATAPI (CDROM) DMA only 2469 libata.dma=4 Compact Flash DMA only 2470 Combinations also work, so libata.dma=3 enables DMA 2471 for disks and CDROMs, but not CFs. 2472 2473 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2474 libata.ignore_hpa=0 keep BIOS limits (default) 2475 libata.ignore_hpa=1 ignore limits, using full disk 2476 2477 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2478 when set. 2479 Format: <int> 2480 2481 libata.force= [LIBATA] Force configurations. The format is comma- 2482 separated list of "[ID:]VAL" where ID is 2483 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2484 matching port, link or device. Basically, it matches 2485 the ATA ID string printed on console by libata. If 2486 the whole ID part is omitted, the last PORT and DEVICE 2487 values are used. If ID hasn't been specified yet, the 2488 configuration applies to all ports, links and devices. 2489 2490 If only DEVICE is omitted, the parameter applies to 2491 the port and all links and devices behind it. DEVICE 2492 number of 0 either selects the first device or the 2493 first fan-out link behind PMP device. It does not 2494 select the host link. DEVICE number of 15 selects the 2495 host link and device attached to it. 2496 2497 The VAL specifies the configuration to force. As long 2498 as there's no ambiguity shortcut notation is allowed. 2499 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2500 The following configurations can be forced. 2501 2502 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2503 Any ID with matching PORT is used. 2504 2505 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2506 2507 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2508 udma[/][16,25,33,44,66,100,133] notation is also 2509 allowed. 2510 2511 * [no]ncq: Turn on or off NCQ. 2512 2513 * [no]ncqtrim: Turn off queued DSM TRIM. 2514 2515 * nohrst, nosrst, norst: suppress hard, soft 2516 and both resets. 2517 2518 * rstonce: only attempt one reset during 2519 hot-unplug link recovery 2520 2521 * dump_id: dump IDENTIFY data. 2522 2523 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2524 2525 * disable: Disable this device. 2526 2527 If there are multiple matching configurations changing 2528 the same attribute, the last one is used. 2529 2530 memblock=debug [KNL] Enable memblock debug messages. 2531 2532 load_ramdisk= [RAM] [Deprecated] 2533 2534 lockd.nlm_grace_period=P [NFS] Assign grace period. 2535 Format: <integer> 2536 2537 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2538 Format: <integer> 2539 2540 lockd.nlm_timeout=T [NFS] Assign timeout value. 2541 Format: <integer> 2542 2543 lockd.nlm_udpport=M [NFS] Assign UDP port. 2544 Format: <integer> 2545 2546 lockdown= [SECURITY] 2547 { integrity | confidentiality } 2548 Enable the kernel lockdown feature. If set to 2549 integrity, kernel features that allow userland to 2550 modify the running kernel are disabled. If set to 2551 confidentiality, kernel features that allow userland 2552 to extract confidential information from the kernel 2553 are also disabled. 2554 2555 locktorture.nreaders_stress= [KNL] 2556 Set the number of locking read-acquisition kthreads. 2557 Defaults to being automatically set based on the 2558 number of online CPUs. 2559 2560 locktorture.nwriters_stress= [KNL] 2561 Set the number of locking write-acquisition kthreads. 2562 2563 locktorture.onoff_holdoff= [KNL] 2564 Set time (s) after boot for CPU-hotplug testing. 2565 2566 locktorture.onoff_interval= [KNL] 2567 Set time (s) between CPU-hotplug operations, or 2568 zero to disable CPU-hotplug testing. 2569 2570 locktorture.shuffle_interval= [KNL] 2571 Set task-shuffle interval (jiffies). Shuffling 2572 tasks allows some CPUs to go into dyntick-idle 2573 mode during the locktorture test. 2574 2575 locktorture.shutdown_secs= [KNL] 2576 Set time (s) after boot system shutdown. This 2577 is useful for hands-off automated testing. 2578 2579 locktorture.stat_interval= [KNL] 2580 Time (s) between statistics printk()s. 2581 2582 locktorture.stutter= [KNL] 2583 Time (s) to stutter testing, for example, 2584 specifying five seconds causes the test to run for 2585 five seconds, wait for five seconds, and so on. 2586 This tests the locking primitive's ability to 2587 transition abruptly to and from idle. 2588 2589 locktorture.torture_type= [KNL] 2590 Specify the locking implementation to test. 2591 2592 locktorture.verbose= [KNL] 2593 Enable additional printk() statements. 2594 2595 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2596 Format: <irq> 2597 2598 loglevel= All Kernel Messages with a loglevel smaller than the 2599 console loglevel will be printed to the console. It can 2600 also be changed with klogd or other programs. The 2601 loglevels are defined as follows: 2602 2603 0 (KERN_EMERG) system is unusable 2604 1 (KERN_ALERT) action must be taken immediately 2605 2 (KERN_CRIT) critical conditions 2606 3 (KERN_ERR) error conditions 2607 4 (KERN_WARNING) warning conditions 2608 5 (KERN_NOTICE) normal but significant condition 2609 6 (KERN_INFO) informational 2610 7 (KERN_DEBUG) debug-level messages 2611 2612 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2613 in bytes. n must be a power of two and greater 2614 than the minimal size. The minimal size is defined 2615 by LOG_BUF_SHIFT kernel config parameter. There is 2616 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2617 that allows to increase the default size depending on 2618 the number of CPUs. See init/Kconfig for more details. 2619 2620 logo.nologo [FB] Disables display of the built-in Linux logo. 2621 This may be used to provide more screen space for 2622 kernel log messages and is useful when debugging 2623 kernel boot problems. 2624 2625 lp=0 [LP] Specify parallel ports to use, e.g, 2626 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2627 lp=reset first parallel port). 'lp=0' disables the 2628 lp=auto printer driver. 'lp=reset' (which can be 2629 specified in addition to the ports) causes 2630 attached printers to be reset. Using 2631 lp=port1,port2,... specifies the parallel ports 2632 to associate lp devices with, starting with 2633 lp0. A port specification may be 'none' to skip 2634 that lp device, or a parport name such as 2635 'parport0'. Specifying 'lp=auto' instead of a 2636 port specification list means that device IDs 2637 from each port should be examined, to see if 2638 an IEEE 1284-compliant printer is attached; if 2639 so, the driver will manage that printer. 2640 See also header of drivers/char/lp.c. 2641 2642 lpj=n [KNL] 2643 Sets loops_per_jiffy to given constant, thus avoiding 2644 time-consuming boot-time autodetection (up to 250 ms per 2645 CPU). 0 enables autodetection (default). To determine 2646 the correct value for your kernel, boot with normal 2647 autodetection and see what value is printed. Note that 2648 on SMP systems the preset will be applied to all CPUs, 2649 which is likely to cause problems if your CPUs need 2650 significantly divergent settings. An incorrect value 2651 will cause delays in the kernel to be wrong, leading to 2652 unpredictable I/O errors and other breakage. Although 2653 unlikely, in the extreme case this might damage your 2654 hardware. 2655 2656 ltpc= [NET] 2657 Format: <io>,<irq>,<dma> 2658 2659 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2660 2661 lsm=lsm1,...,lsmN 2662 [SECURITY] Choose order of LSM initialization. This 2663 overrides CONFIG_LSM, and the "security=" parameter. 2664 2665 machvec= [IA-64] Force the use of a particular machine-vector 2666 (machvec) in a generic kernel. 2667 Example: machvec=hpzx1 2668 2669 machtype= [Loongson] Share the same kernel image file between 2670 different yeeloong laptops. 2671 Example: machtype=lemote-yeeloong-2f-7inch 2672 2673 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2674 than or equal to this physical address is ignored. 2675 2676 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2677 will bring up during bootup. maxcpus=n : n >= 0 limits 2678 the kernel to bring up 'n' processors. Surely after 2679 bootup you can bring up the other plugged cpu by executing 2680 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2681 only takes effect during system bootup. 2682 While n=0 is a special case, it is equivalent to "nosmp", 2683 which also disables the IO APIC. 2684 2685 max_loop= [LOOP] The number of loop block devices that get 2686 (loop.max_loop) unconditionally pre-created at init time. The default 2687 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2688 of statically allocating a predefined number, loop 2689 devices can be requested on-demand with the 2690 /dev/loop-control interface. 2691 2692 mce [X86-32] Machine Check Exception 2693 2694 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2695 2696 md= [HW] RAID subsystems devices and level 2697 See Documentation/admin-guide/md.rst. 2698 2699 mdacon= [MDA] 2700 Format: <first>,<last> 2701 Specifies range of consoles to be captured by the MDA. 2702 2703 mds= [X86,INTEL] 2704 Control mitigation for the Micro-architectural Data 2705 Sampling (MDS) vulnerability. 2706 2707 Certain CPUs are vulnerable to an exploit against CPU 2708 internal buffers which can forward information to a 2709 disclosure gadget under certain conditions. 2710 2711 In vulnerable processors, the speculatively 2712 forwarded data can be used in a cache side channel 2713 attack, to access data to which the attacker does 2714 not have direct access. 2715 2716 This parameter controls the MDS mitigation. The 2717 options are: 2718 2719 full - Enable MDS mitigation on vulnerable CPUs 2720 full,nosmt - Enable MDS mitigation and disable 2721 SMT on vulnerable CPUs 2722 off - Unconditionally disable MDS mitigation 2723 2724 On TAA-affected machines, mds=off can be prevented by 2725 an active TAA mitigation as both vulnerabilities are 2726 mitigated with the same mechanism so in order to disable 2727 this mitigation, you need to specify tsx_async_abort=off 2728 too. 2729 2730 Not specifying this option is equivalent to 2731 mds=full. 2732 2733 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2734 2735 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2736 Amount of memory to be used in cases as follows: 2737 2738 1 for test; 2739 2 when the kernel is not able to see the whole system memory; 2740 3 memory that lies after 'mem=' boundary is excluded from 2741 the hypervisor, then assigned to KVM guests. 2742 2743 [X86] Work as limiting max address. Use together 2744 with memmap= to avoid physical address space collisions. 2745 Without memmap= PCI devices could be placed at addresses 2746 belonging to unused RAM. 2747 2748 Note that this only takes effects during boot time since 2749 in above case 3, memory may need be hot added after boot 2750 if system memory of hypervisor is not sufficient. 2751 2752 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2753 memory. 2754 2755 memchunk=nn[KMG] 2756 [KNL,SH] Allow user to override the default size for 2757 per-device physically contiguous DMA buffers. 2758 2759 memhp_default_state=online/offline 2760 [KNL] Set the initial state for the memory hotplug 2761 onlining policy. If not specified, the default value is 2762 set according to the 2763 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2764 option. 2765 See Documentation/admin-guide/mm/memory-hotplug.rst. 2766 2767 memmap=exactmap [KNL,X86] Enable setting of an exact 2768 E820 memory map, as specified by the user. 2769 Such memmap=exactmap lines can be constructed based on 2770 BIOS output or other requirements. See the memmap=nn@ss 2771 option description. 2772 2773 memmap=nn[KMG]@ss[KMG] 2774 [KNL, X86, MIPS, XTENSA] Force usage of a specific region of memory. 2775 Region of memory to be used is from ss to ss+nn. 2776 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2777 which limits max address to nn[KMG]. 2778 Multiple different regions can be specified, 2779 comma delimited. 2780 Example: 2781 memmap=100M@2G,100M#3G,1G!1024G 2782 2783 memmap=nn[KMG]#ss[KMG] 2784 [KNL,ACPI] Mark specific memory as ACPI data. 2785 Region of memory to be marked is from ss to ss+nn. 2786 2787 memmap=nn[KMG]$ss[KMG] 2788 [KNL,ACPI] Mark specific memory as reserved. 2789 Region of memory to be reserved is from ss to ss+nn. 2790 Example: Exclude memory from 0x18690000-0x1869ffff 2791 memmap=64K$0x18690000 2792 or 2793 memmap=0x10000$0x18690000 2794 Some bootloaders may need an escape character before '$', 2795 like Grub2, otherwise '$' and the following number 2796 will be eaten. 2797 2798 memmap=nn[KMG]!ss[KMG] 2799 [KNL,X86] Mark specific memory as protected. 2800 Region of memory to be used, from ss to ss+nn. 2801 The memory region may be marked as e820 type 12 (0xc) 2802 and is NVDIMM or ADR memory. 2803 2804 memmap=<size>%<offset>-<oldtype>+<newtype> 2805 [KNL,ACPI] Convert memory within the specified region 2806 from <oldtype> to <newtype>. If "-<oldtype>" is left 2807 out, the whole region will be marked as <newtype>, 2808 even if previously unavailable. If "+<newtype>" is left 2809 out, matching memory will be removed. Types are 2810 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2811 3 = ACPI, 12 = PRAM. 2812 2813 memory_corruption_check=0/1 [X86] 2814 Some BIOSes seem to corrupt the first 64k of 2815 memory when doing things like suspend/resume. 2816 Setting this option will scan the memory 2817 looking for corruption. Enabling this will 2818 both detect corruption and prevent the kernel 2819 from using the memory being corrupted. 2820 However, its intended as a diagnostic tool; if 2821 repeatable BIOS-originated corruption always 2822 affects the same memory, you can use memmap= 2823 to prevent the kernel from using that memory. 2824 2825 memory_corruption_check_size=size [X86] 2826 By default it checks for corruption in the low 2827 64k, making this memory unavailable for normal 2828 use. Use this parameter to scan for 2829 corruption in more or less memory. 2830 2831 memory_corruption_check_period=seconds [X86] 2832 By default it checks for corruption every 60 2833 seconds. Use this parameter to check at some 2834 other rate. 0 disables periodic checking. 2835 2836 memory_hotplug.memmap_on_memory 2837 [KNL,X86,ARM] Boolean flag to enable this feature. 2838 Format: {on | off (default)} 2839 When enabled, runtime hotplugged memory will 2840 allocate its internal metadata (struct pages) 2841 from the hotadded memory which will allow to 2842 hotadd a lot of memory without requiring 2843 additional memory to do so. 2844 This feature is disabled by default because it 2845 has some implication on large (e.g. GB) 2846 allocations in some configurations (e.g. small 2847 memory blocks). 2848 The state of the flag can be read in 2849 /sys/module/memory_hotplug/parameters/memmap_on_memory. 2850 Note that even when enabled, there are a few cases where 2851 the feature is not effective. 2852 2853 This is not compatible with hugetlb_free_vmemmap. If 2854 both parameters are enabled, hugetlb_free_vmemmap takes 2855 precedence over memory_hotplug.memmap_on_memory. 2856 2857 memtest= [KNL,X86,ARM,PPC,RISCV] Enable memtest 2858 Format: <integer> 2859 default : 0 <disable> 2860 Specifies the number of memtest passes to be 2861 performed. Each pass selects another test 2862 pattern from a given set of patterns. Memtest 2863 fills the memory with this pattern, validates 2864 memory contents and reserves bad memory 2865 regions that are detected. 2866 2867 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2868 Valid arguments: on, off 2869 Default (depends on kernel configuration option): 2870 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2871 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2872 mem_encrypt=on: Activate SME 2873 mem_encrypt=off: Do not activate SME 2874 2875 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2876 for details on when memory encryption can be activated. 2877 2878 mem_sleep_default= [SUSPEND] Default system suspend mode: 2879 s2idle - Suspend-To-Idle 2880 shallow - Power-On Suspend or equivalent (if supported) 2881 deep - Suspend-To-RAM or equivalent (if supported) 2882 See Documentation/admin-guide/pm/sleep-states.rst. 2883 2884 meye.*= [HW] Set MotionEye Camera parameters 2885 See Documentation/admin-guide/media/meye.rst. 2886 2887 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2888 Multi-Function General Purpose Timers on AMD Geode 2889 platforms. 2890 2891 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2892 the BIOS has incorrectly applied a workaround. TinyBIOS 2893 version 0.98 is known to be affected, 0.99 fixes the 2894 problem by letting the user disable the workaround. 2895 2896 mga= [HW,DRM] 2897 2898 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2899 physical address is ignored. 2900 2901 mini2440= [ARM,HW,KNL] 2902 Format:[0..2][b][c][t] 2903 Default: "0tb" 2904 MINI2440 configuration specification: 2905 0 - The attached screen is the 3.5" TFT 2906 1 - The attached screen is the 7" TFT 2907 2 - The VGA Shield is attached (1024x768) 2908 Leaving out the screen size parameter will not load 2909 the TFT driver, and the framebuffer will be left 2910 unconfigured. 2911 b - Enable backlight. The TFT backlight pin will be 2912 linked to the kernel VESA blanking code and a GPIO 2913 LED. This parameter is not necessary when using the 2914 VGA shield. 2915 c - Enable the s3c camera interface. 2916 t - Reserved for enabling touchscreen support. The 2917 touchscreen support is not enabled in the mainstream 2918 kernel as of 2.6.30, a preliminary port can be found 2919 in the "bleeding edge" mini2440 support kernel at 2920 https://repo.or.cz/w/linux-2.6/mini2440.git 2921 2922 mitigations= 2923 [X86,PPC,S390,ARM64] Control optional mitigations for 2924 CPU vulnerabilities. This is a set of curated, 2925 arch-independent options, each of which is an 2926 aggregation of existing arch-specific options. 2927 2928 off 2929 Disable all optional CPU mitigations. This 2930 improves system performance, but it may also 2931 expose users to several CPU vulnerabilities. 2932 Equivalent to: nopti [X86,PPC] 2933 kpti=0 [ARM64] 2934 nospectre_v1 [X86,PPC] 2935 nobp=0 [S390] 2936 nospectre_v2 [X86,PPC,S390,ARM64] 2937 spectre_v2_user=off [X86] 2938 spec_store_bypass_disable=off [X86,PPC] 2939 ssbd=force-off [ARM64] 2940 l1tf=off [X86] 2941 mds=off [X86] 2942 tsx_async_abort=off [X86] 2943 kvm.nx_huge_pages=off [X86] 2944 no_entry_flush [PPC] 2945 no_uaccess_flush [PPC] 2946 2947 Exceptions: 2948 This does not have any effect on 2949 kvm.nx_huge_pages when 2950 kvm.nx_huge_pages=force. 2951 2952 auto (default) 2953 Mitigate all CPU vulnerabilities, but leave SMT 2954 enabled, even if it's vulnerable. This is for 2955 users who don't want to be surprised by SMT 2956 getting disabled across kernel upgrades, or who 2957 have other ways of avoiding SMT-based attacks. 2958 Equivalent to: (default behavior) 2959 2960 auto,nosmt 2961 Mitigate all CPU vulnerabilities, disabling SMT 2962 if needed. This is for users who always want to 2963 be fully mitigated, even if it means losing SMT. 2964 Equivalent to: l1tf=flush,nosmt [X86] 2965 mds=full,nosmt [X86] 2966 tsx_async_abort=full,nosmt [X86] 2967 2968 mminit_loglevel= 2969 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2970 parameter allows control of the logging verbosity for 2971 the additional memory initialisation checks. A value 2972 of 0 disables mminit logging and a level of 4 will 2973 log everything. Information is printed at KERN_DEBUG 2974 so loglevel=8 may also need to be specified. 2975 2976 module.sig_enforce 2977 [KNL] When CONFIG_MODULE_SIG is set, this means that 2978 modules without (valid) signatures will fail to load. 2979 Note that if CONFIG_MODULE_SIG_FORCE is set, that 2980 is always true, so this option does nothing. 2981 2982 module_blacklist= [KNL] Do not load a comma-separated list of 2983 modules. Useful for debugging problem modules. 2984 2985 mousedev.tap_time= 2986 [MOUSE] Maximum time between finger touching and 2987 leaving touchpad surface for touch to be considered 2988 a tap and be reported as a left button click (for 2989 touchpads working in absolute mode only). 2990 Format: <msecs> 2991 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 2992 reporting absolute coordinates, such as tablets 2993 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 2994 reporting absolute coordinates, such as tablets 2995 2996 movablecore= [KNL,X86,IA-64,PPC] 2997 Format: nn[KMGTPE] | nn% 2998 This parameter is the complement to kernelcore=, it 2999 specifies the amount of memory used for migratable 3000 allocations. If both kernelcore and movablecore is 3001 specified, then kernelcore will be at *least* the 3002 specified value but may be more. If movablecore on its 3003 own is specified, the administrator must be careful 3004 that the amount of memory usable for all allocations 3005 is not too small. 3006 3007 movable_node [KNL] Boot-time switch to make hotplugable memory 3008 NUMA nodes to be movable. This means that the memory 3009 of such nodes will be usable only for movable 3010 allocations which rules out almost all kernel 3011 allocations. Use with caution! 3012 3013 MTD_Partition= [MTD] 3014 Format: <name>,<region-number>,<size>,<offset> 3015 3016 MTD_Region= [MTD] Format: 3017 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 3018 3019 mtdparts= [MTD] 3020 See drivers/mtd/parsers/cmdlinepart.c 3021 3022 multitce=off [PPC] This parameter disables the use of the pSeries 3023 firmware feature for updating multiple TCE entries 3024 at a time. 3025 3026 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 3027 3028 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 3029 3030 boundary - index of last SLC block on Flex-OneNAND. 3031 The remaining blocks are configured as MLC blocks. 3032 lock - Configure if Flex-OneNAND boundary should be locked. 3033 Once locked, the boundary cannot be changed. 3034 1 indicates lock status, 0 indicates unlock status. 3035 3036 mtdset= [ARM] 3037 ARM/S3C2412 JIVE boot control 3038 3039 See arch/arm/mach-s3c/mach-jive.c 3040 3041 mtouchusb.raw_coordinates= 3042 [HW] Make the MicroTouch USB driver use raw coordinates 3043 ('y', default) or cooked coordinates ('n') 3044 3045 mtrr_chunk_size=nn[KMG] [X86] 3046 used for mtrr cleanup. It is largest continuous chunk 3047 that could hold holes aka. UC entries. 3048 3049 mtrr_gran_size=nn[KMG] [X86] 3050 Used for mtrr cleanup. It is granularity of mtrr block. 3051 Default is 1. 3052 Large value could prevent small alignment from 3053 using up MTRRs. 3054 3055 mtrr_spare_reg_nr=n [X86] 3056 Format: <integer> 3057 Range: 0,7 : spare reg number 3058 Default : 1 3059 Used for mtrr cleanup. It is spare mtrr entries number. 3060 Set to 2 or more if your graphical card needs more. 3061 3062 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 3063 3064 netdev= [NET] Network devices parameters 3065 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 3066 Note that mem_start is often overloaded to mean 3067 something different and driver-specific. 3068 This usage is only documented in each driver source 3069 file if at all. 3070 3071 nf_conntrack.acct= 3072 [NETFILTER] Enable connection tracking flow accounting 3073 0 to disable accounting 3074 1 to enable accounting 3075 Default value is 0. 3076 3077 nfsaddrs= [NFS] Deprecated. Use ip= instead. 3078 See Documentation/admin-guide/nfs/nfsroot.rst. 3079 3080 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 3081 See Documentation/admin-guide/nfs/nfsroot.rst. 3082 3083 nfsrootdebug [NFS] enable nfsroot debugging messages. 3084 See Documentation/admin-guide/nfs/nfsroot.rst. 3085 3086 nfs.callback_nr_threads= 3087 [NFSv4] set the total number of threads that the 3088 NFS client will assign to service NFSv4 callback 3089 requests. 3090 3091 nfs.callback_tcpport= 3092 [NFS] set the TCP port on which the NFSv4 callback 3093 channel should listen. 3094 3095 nfs.cache_getent= 3096 [NFS] sets the pathname to the program which is used 3097 to update the NFS client cache entries. 3098 3099 nfs.cache_getent_timeout= 3100 [NFS] sets the timeout after which an attempt to 3101 update a cache entry is deemed to have failed. 3102 3103 nfs.idmap_cache_timeout= 3104 [NFS] set the maximum lifetime for idmapper cache 3105 entries. 3106 3107 nfs.enable_ino64= 3108 [NFS] enable 64-bit inode numbers. 3109 If zero, the NFS client will fake up a 32-bit inode 3110 number for the readdir() and stat() syscalls instead 3111 of returning the full 64-bit number. 3112 The default is to return 64-bit inode numbers. 3113 3114 nfs.max_session_cb_slots= 3115 [NFSv4.1] Sets the maximum number of session 3116 slots the client will assign to the callback 3117 channel. This determines the maximum number of 3118 callbacks the client will process in parallel for 3119 a particular server. 3120 3121 nfs.max_session_slots= 3122 [NFSv4.1] Sets the maximum number of session slots 3123 the client will attempt to negotiate with the server. 3124 This limits the number of simultaneous RPC requests 3125 that the client can send to the NFSv4.1 server. 3126 Note that there is little point in setting this 3127 value higher than the max_tcp_slot_table_limit. 3128 3129 nfs.nfs4_disable_idmapping= 3130 [NFSv4] When set to the default of '1', this option 3131 ensures that both the RPC level authentication 3132 scheme and the NFS level operations agree to use 3133 numeric uids/gids if the mount is using the 3134 'sec=sys' security flavour. In effect it is 3135 disabling idmapping, which can make migration from 3136 legacy NFSv2/v3 systems to NFSv4 easier. 3137 Servers that do not support this mode of operation 3138 will be autodetected by the client, and it will fall 3139 back to using the idmapper. 3140 To turn off this behaviour, set the value to '0'. 3141 nfs.nfs4_unique_id= 3142 [NFS4] Specify an additional fixed unique ident- 3143 ification string that NFSv4 clients can insert into 3144 their nfs_client_id4 string. This is typically a 3145 UUID that is generated at system install time. 3146 3147 nfs.send_implementation_id = 3148 [NFSv4.1] Send client implementation identification 3149 information in exchange_id requests. 3150 If zero, no implementation identification information 3151 will be sent. 3152 The default is to send the implementation identification 3153 information. 3154 3155 nfs.recover_lost_locks = 3156 [NFSv4] Attempt to recover locks that were lost due 3157 to a lease timeout on the server. Please note that 3158 doing this risks data corruption, since there are 3159 no guarantees that the file will remain unchanged 3160 after the locks are lost. 3161 If you want to enable the kernel legacy behaviour of 3162 attempting to recover these locks, then set this 3163 parameter to '1'. 3164 The default parameter value of '0' causes the kernel 3165 not to attempt recovery of lost locks. 3166 3167 nfs4.layoutstats_timer = 3168 [NFSv4.2] Change the rate at which the kernel sends 3169 layoutstats to the pNFS metadata server. 3170 3171 Setting this to value to 0 causes the kernel to use 3172 whatever value is the default set by the layout 3173 driver. A non-zero value sets the minimum interval 3174 in seconds between layoutstats transmissions. 3175 3176 nfsd.nfs4_disable_idmapping= 3177 [NFSv4] When set to the default of '1', the NFSv4 3178 server will return only numeric uids and gids to 3179 clients using auth_sys, and will accept numeric uids 3180 and gids from such clients. This is intended to ease 3181 migration from NFSv2/v3. 3182 3183 nmi_backtrace.backtrace_idle [KNL] 3184 Dump stacks even of idle CPUs in response to an 3185 NMI stack-backtrace request. 3186 3187 nmi_debug= [KNL,SH] Specify one or more actions to take 3188 when a NMI is triggered. 3189 Format: [state][,regs][,debounce][,die] 3190 3191 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3192 Format: [panic,][nopanic,][num] 3193 Valid num: 0 or 1 3194 0 - turn hardlockup detector in nmi_watchdog off 3195 1 - turn hardlockup detector in nmi_watchdog on 3196 When panic is specified, panic when an NMI watchdog 3197 timeout occurs (or 'nopanic' to not panic on an NMI 3198 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3199 To disable both hard and soft lockup detectors, 3200 please see 'nowatchdog'. 3201 This is useful when you use a panic=... timeout and 3202 need the box quickly up again. 3203 3204 These settings can be accessed at runtime via 3205 the nmi_watchdog and hardlockup_panic sysctls. 3206 3207 netpoll.carrier_timeout= 3208 [NET] Specifies amount of time (in seconds) that 3209 netpoll should wait for a carrier. By default netpoll 3210 waits 4 seconds. 3211 3212 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3213 emulation library even if a 387 maths coprocessor 3214 is present. 3215 3216 no5lvl [X86-64] Disable 5-level paging mode. Forces 3217 kernel to use 4-level paging instead. 3218 3219 nofsgsbase [X86] Disables FSGSBASE instructions. 3220 3221 no_console_suspend 3222 [HW] Never suspend the console 3223 Disable suspending of consoles during suspend and 3224 hibernate operations. Once disabled, debugging 3225 messages can reach various consoles while the rest 3226 of the system is being put to sleep (ie, while 3227 debugging driver suspend/resume hooks). This may 3228 not work reliably with all consoles, but is known 3229 to work with serial and VGA consoles. 3230 To facilitate more flexible debugging, we also add 3231 console_suspend, a printk module parameter to control 3232 it. Users could use console_suspend (usually 3233 /sys/module/printk/parameters/console_suspend) to 3234 turn on/off it dynamically. 3235 3236 novmcoredd [KNL,KDUMP] 3237 Disable device dump. Device dump allows drivers to 3238 append dump data to vmcore so you can collect driver 3239 specified debug info. Drivers can append the data 3240 without any limit and this data is stored in memory, 3241 so this may cause significant memory stress. Disabling 3242 device dump can help save memory but the driver debug 3243 data will be no longer available. This parameter 3244 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3245 is set. 3246 3247 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3248 caches in the slab allocator. Saves per-node memory, 3249 but will impact performance. 3250 3251 noalign [KNL,ARM] 3252 3253 noaltinstr [S390] Disables alternative instructions patching 3254 (CPU alternatives feature). 3255 3256 noapic [SMP,APIC] Tells the kernel to not make use of any 3257 IOAPICs that may be present in the system. 3258 3259 noautogroup Disable scheduler automatic task group creation. 3260 3261 nobats [PPC] Do not use BATs for mapping kernel lowmem 3262 on "Classic" PPC cores. 3263 3264 nocache [ARM] 3265 3266 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 3267 3268 nodelayacct [KNL] Disable per-task delay accounting 3269 3270 nodsp [SH] Disable hardware DSP at boot time. 3271 3272 noefi Disable EFI runtime services support. 3273 3274 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel. 3275 3276 noexec [IA-64] 3277 3278 noexec [X86] 3279 On X86-32 available only on PAE configured kernels. 3280 noexec=on: enable non-executable mappings (default) 3281 noexec=off: disable non-executable mappings 3282 3283 nosmap [X86,PPC] 3284 Disable SMAP (Supervisor Mode Access Prevention) 3285 even if it is supported by processor. 3286 3287 nosmep [X86,PPC] 3288 Disable SMEP (Supervisor Mode Execution Prevention) 3289 even if it is supported by processor. 3290 3291 noexec32 [X86-64] 3292 This affects only 32-bit executables. 3293 noexec32=on: enable non-executable mappings (default) 3294 read doesn't imply executable mappings 3295 noexec32=off: disable non-executable mappings 3296 read implies executable mappings 3297 3298 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3299 3300 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3301 register save and restore. The kernel will only save 3302 legacy floating-point registers on task switch. 3303 3304 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings. 3305 3306 nohugevmalloc [PPC] Disable kernel huge vmalloc mappings. 3307 3308 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3309 Equivalent to smt=1. 3310 3311 [KNL,X86] Disable symmetric multithreading (SMT). 3312 nosmt=force: Force disable SMT, cannot be undone 3313 via the sysfs control file. 3314 3315 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3316 (bounds check bypass). With this option data leaks are 3317 possible in the system. 3318 3319 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3320 the Spectre variant 2 (indirect branch prediction) 3321 vulnerability. System may allow data leaks with this 3322 option. 3323 3324 nospec_store_bypass_disable 3325 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3326 3327 no_uaccess_flush 3328 [PPC] Don't flush the L1-D cache after accessing user data. 3329 3330 noxsave [BUGS=X86] Disables x86 extended register state save 3331 and restore using xsave. The kernel will fallback to 3332 enabling legacy floating-point and sse state. 3333 3334 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3335 register states. The kernel will fall back to use 3336 xsave to save the states. By using this parameter, 3337 performance of saving the states is degraded because 3338 xsave doesn't support modified optimization while 3339 xsaveopt supports it on xsaveopt enabled systems. 3340 3341 noxsaves [X86] Disables xsaves and xrstors used in saving and 3342 restoring x86 extended register state in compacted 3343 form of xsave area. The kernel will fall back to use 3344 xsaveopt and xrstor to save and restore the states 3345 in standard form of xsave area. By using this 3346 parameter, xsave area per process might occupy more 3347 memory on xsaves enabled systems. 3348 3349 nohlt [ARM,ARM64,MICROBLAZE,SH] Forces the kernel to busy wait 3350 in do_idle() and not use the arch_cpu_idle() 3351 implementation; requires CONFIG_GENERIC_IDLE_POLL_SETUP 3352 to be effective. This is useful on platforms where the 3353 sleep(SH) or wfi(ARM,ARM64) instructions do not work 3354 correctly or when doing power measurements to evalute 3355 the impact of the sleep instructions. This is also 3356 useful when using JTAG debugger. 3357 3358 no_file_caps Tells the kernel not to honor file capabilities. The 3359 only way then for a file to be executed with privilege 3360 is to be setuid root or executed by root. 3361 3362 nohalt [IA-64] Tells the kernel not to use the power saving 3363 function PAL_HALT_LIGHT when idle. This increases 3364 power-consumption. On the positive side, it reduces 3365 interrupt wake-up latency, which may improve performance 3366 in certain environments such as networked servers or 3367 real-time systems. 3368 3369 no_hash_pointers 3370 Force pointers printed to the console or buffers to be 3371 unhashed. By default, when a pointer is printed via %p 3372 format string, that pointer is "hashed", i.e. obscured 3373 by hashing the pointer value. This is a security feature 3374 that hides actual kernel addresses from unprivileged 3375 users, but it also makes debugging the kernel more 3376 difficult since unequal pointers can no longer be 3377 compared. However, if this command-line option is 3378 specified, then all normal pointers will have their true 3379 value printed. Pointers printed via %pK may still be 3380 hashed. This option should only be specified when 3381 debugging the kernel. Please do not use on production 3382 kernels. 3383 3384 nohibernate [HIBERNATION] Disable hibernation and resume. 3385 3386 nohz= [KNL] Boottime enable/disable dynamic ticks 3387 Valid arguments: on, off 3388 Default: on 3389 3390 nohz_full= [KNL,BOOT,SMP,ISOL] 3391 The argument is a cpu list, as described above. 3392 In kernels built with CONFIG_NO_HZ_FULL=y, set 3393 the specified list of CPUs whose tick will be stopped 3394 whenever possible. The boot CPU will be forced outside 3395 the range to maintain the timekeeping. Any CPUs 3396 in this list will have their RCU callbacks offloaded, 3397 just as if they had also been called out in the 3398 rcu_nocbs= boot parameter. 3399 3400 noiotrap [SH] Disables trapped I/O port accesses. 3401 3402 noirqdebug [X86-32] Disables the code which attempts to detect and 3403 disable unhandled interrupt sources. 3404 3405 no_timer_check [X86,APIC] Disables the code which tests for 3406 broken timer IRQ sources. 3407 3408 noisapnp [ISAPNP] Disables ISA PnP code. 3409 3410 noinitrd [RAM] Tells the kernel not to load any configured 3411 initial RAM disk. 3412 3413 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3414 remapping. 3415 [Deprecated - use intremap=off] 3416 3417 nointroute [IA-64] 3418 3419 noinvpcid [X86] Disable the INVPCID cpu feature. 3420 3421 nojitter [IA-64] Disables jitter checking for ITC timers. 3422 3423 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3424 3425 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3426 fault handling. 3427 3428 no-vmw-sched-clock 3429 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3430 clock and use the default one. 3431 3432 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time 3433 accounting. steal time is computed, but won't 3434 influence scheduler behaviour 3435 3436 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3437 3438 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3439 3440 noltlbs [PPC] Do not use large page/tlb entries for kernel 3441 lowmem mapping on PPC40x and PPC8xx 3442 3443 nomca [IA-64] Disable machine check abort handling 3444 3445 nomce [X86-32] Disable Machine Check Exception 3446 3447 nomfgpt [X86-32] Disable Multi-Function General Purpose 3448 Timer usage (for AMD Geode machines). 3449 3450 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3451 shutdown the other cpus. Instead use the REBOOT_VECTOR 3452 irq. 3453 3454 nomodule Disable module load 3455 3456 nopat [X86] Disable PAT (page attribute table extension of 3457 pagetables) support. 3458 3459 nopcid [X86-64] Disable the PCID cpu feature. 3460 3461 norandmaps Don't use address space randomization. Equivalent to 3462 echo 0 > /proc/sys/kernel/randomize_va_space 3463 3464 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3465 with UP alternatives 3466 3467 nordrand [X86] Disable kernel use of the RDRAND and 3468 RDSEED instructions even if they are supported 3469 by the processor. RDRAND and RDSEED are still 3470 available to user space applications. 3471 3472 noresume [SWSUSP] Disables resume and restores original swap 3473 space. 3474 3475 no-scroll [VGA] Disables scrollback. 3476 This is required for the Braillex ib80-piezo Braille 3477 reader made by F.H. Papenmeier (Germany). 3478 3479 nosbagart [IA-64] 3480 3481 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3482 3483 nosgx [X86-64,SGX] Disables Intel SGX kernel support. 3484 3485 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3486 and disable the IO APIC. legacy for "maxcpus=0". 3487 3488 nosoftlockup [KNL] Disable the soft-lockup detector. 3489 3490 nosync [HW,M68K] Disables sync negotiation for all devices. 3491 3492 nowatchdog [KNL] Disable both lockup detectors, i.e. 3493 soft-lockup and NMI watchdog (hard-lockup). 3494 3495 nowb [ARM] 3496 3497 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3498 3499 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3500 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3501 Some features depend on CPU0. Known dependencies are: 3502 1. Resume from suspend/hibernate depends on CPU0. 3503 Suspend/hibernate will fail if CPU0 is offline and you 3504 need to online CPU0 before suspend/hibernate. 3505 2. PIC interrupts also depend on CPU0. CPU0 can't be 3506 removed if a PIC interrupt is detected. 3507 It's said poweroff/reboot may depend on CPU0 on some 3508 machines although I haven't seen such issues so far 3509 after CPU0 is offline on a few tested machines. 3510 If the dependencies are under your control, you can 3511 turn on cpu0_hotplug. 3512 3513 nps_mtm_hs_ctr= [KNL,ARC] 3514 This parameter sets the maximum duration, in 3515 cycles, each HW thread of the CTOP can run 3516 without interruptions, before HW switches it. 3517 The actual maximum duration is 16 times this 3518 parameter's value. 3519 Format: integer between 1 and 255 3520 Default: 255 3521 3522 nptcg= [IA-64] Override max number of concurrent global TLB 3523 purges which is reported from either PAL_VM_SUMMARY or 3524 SAL PALO. 3525 3526 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3527 could support. nr_cpus=n : n >= 1 limits the kernel to 3528 support 'n' processors. It could be larger than the 3529 number of already plugged CPU during bootup, later in 3530 runtime you can physically add extra cpu until it reaches 3531 n. So during boot up some boot time memory for per-cpu 3532 variables need be pre-allocated for later physical cpu 3533 hot plugging. 3534 3535 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3536 3537 numa_balancing= [KNL,ARM64,PPC,RISCV,S390,X86] Enable or disable automatic 3538 NUMA balancing. 3539 Allowed values are enable and disable 3540 3541 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3542 'node', 'default' can be specified 3543 This can be set from sysctl after boot. 3544 See Documentation/admin-guide/sysctl/vm.rst for details. 3545 3546 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3547 See Documentation/core-api/debugging-via-ohci1394.rst for more 3548 info. 3549 3550 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3551 Rather than timing out after 20 ms if an EC 3552 command is not properly ACKed, override the length 3553 of the timeout. We have interrupts disabled while 3554 waiting for the ACK, so if this is set too high 3555 interrupts *may* be lost! 3556 3557 omap_mux= [OMAP] Override bootloader pin multiplexing. 3558 Format: <mux_mode0.mode_name=value>... 3559 For example, to override I2C bus2: 3560 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3561 3562 oops=panic Always panic on oopses. Default is to just kill the 3563 process, but there is a small probability of 3564 deadlocking the machine. 3565 This will also cause panics on machine check exceptions. 3566 Useful together with panic=30 to trigger a reboot. 3567 3568 page_alloc.shuffle= 3569 [KNL] Boolean flag to control whether the page allocator 3570 should randomize its free lists. The randomization may 3571 be automatically enabled if the kernel detects it is 3572 running on a platform with a direct-mapped memory-side 3573 cache, and this parameter can be used to 3574 override/disable that behavior. The state of the flag 3575 can be read from sysfs at: 3576 /sys/module/page_alloc/parameters/shuffle. 3577 3578 page_owner= [KNL] Boot-time page_owner enabling option. 3579 Storage of the information about who allocated 3580 each page is disabled in default. With this switch, 3581 we can turn it on. 3582 on: enable the feature 3583 3584 page_poison= [KNL] Boot-time parameter changing the state of 3585 poisoning on the buddy allocator, available with 3586 CONFIG_PAGE_POISONING=y. 3587 off: turn off poisoning (default) 3588 on: turn on poisoning 3589 3590 page_reporting.page_reporting_order= 3591 [KNL] Minimal page reporting order 3592 Format: <integer> 3593 Adjust the minimal page reporting order. The page 3594 reporting is disabled when it exceeds (MAX_ORDER-1). 3595 3596 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3597 timeout > 0: seconds before rebooting 3598 timeout = 0: wait forever 3599 timeout < 0: reboot immediately 3600 Format: <timeout> 3601 3602 panic_print= Bitmask for printing system info when panic happens. 3603 User can chose combination of the following bits: 3604 bit 0: print all tasks info 3605 bit 1: print system memory info 3606 bit 2: print timer info 3607 bit 3: print locks info if CONFIG_LOCKDEP is on 3608 bit 4: print ftrace buffer 3609 bit 5: print all printk messages in buffer 3610 3611 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 3612 Format: <hex>[,nousertaint] 3613 Hexadecimal bitmask representing the set of TAINT flags 3614 that will cause the kernel to panic when add_taint() is 3615 called with any of the flags in this set. 3616 The optional switch "nousertaint" can be utilized to 3617 prevent userspace forced crashes by writing to sysctl 3618 /proc/sys/kernel/tainted any flagset matching with the 3619 bitmask set on panic_on_taint. 3620 See Documentation/admin-guide/tainted-kernels.rst for 3621 extra details on the taint flags that users can pick 3622 to compose the bitmask to assign to panic_on_taint. 3623 3624 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3625 on a WARN(). 3626 3627 crash_kexec_post_notifiers 3628 Run kdump after running panic-notifiers and dumping 3629 kmsg. This only for the users who doubt kdump always 3630 succeeds in any situation. 3631 Note that this also increases risks of kdump failure, 3632 because some panic notifiers can make the crashed 3633 kernel more unstable. 3634 3635 parkbd.port= [HW] Parallel port number the keyboard adapter is 3636 connected to, default is 0. 3637 Format: <parport#> 3638 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3639 0 for XT, 1 for AT (default is AT). 3640 Format: <mode> 3641 3642 parport= [HW,PPT] Specify parallel ports. 0 disables. 3643 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3644 Use 'auto' to force the driver to use any 3645 IRQ/DMA settings detected (the default is to 3646 ignore detected IRQ/DMA settings because of 3647 possible conflicts). You can specify the base 3648 address, IRQ, and DMA settings; IRQ and DMA 3649 should be numbers, or 'auto' (for using detected 3650 settings on that particular port), or 'nofifo' 3651 (to avoid using a FIFO even if it is detected). 3652 Parallel ports are assigned in the order they 3653 are specified on the command line, starting 3654 with parport0. 3655 3656 parport_init_mode= [HW,PPT] 3657 Configure VIA parallel port to operate in 3658 a specific mode. This is necessary on Pegasos 3659 computer where firmware has no options for setting 3660 up parallel port mode and sets it to spp. 3661 Currently this function knows 686a and 8231 chips. 3662 Format: [spp|ps2|epp|ecp|ecpepp] 3663 3664 pata_legacy.all= [HW,LIBATA] 3665 Format: <int> 3666 Set to non-zero to probe primary and secondary ISA 3667 port ranges on PCI systems where no PCI PATA device 3668 has been found at either range. Disabled by default. 3669 3670 pata_legacy.autospeed= [HW,LIBATA] 3671 Format: <int> 3672 Set to non-zero if a chip is present that snoops speed 3673 changes. Disabled by default. 3674 3675 pata_legacy.ht6560a= [HW,LIBATA] 3676 Format: <int> 3677 Set to 1, 2, or 3 for HT 6560A on the primary channel, 3678 the secondary channel, or both channels respectively. 3679 Disabled by default. 3680 3681 pata_legacy.ht6560b= [HW,LIBATA] 3682 Format: <int> 3683 Set to 1, 2, or 3 for HT 6560B on the primary channel, 3684 the secondary channel, or both channels respectively. 3685 Disabled by default. 3686 3687 pata_legacy.iordy_mask= [HW,LIBATA] 3688 Format: <int> 3689 IORDY enable mask. Set individual bits to allow IORDY 3690 for the respective channel. Bit 0 is for the first 3691 legacy channel handled by this driver, bit 1 is for 3692 the second channel, and so on. The sequence will often 3693 correspond to the primary legacy channel, the secondary 3694 legacy channel, and so on, but the handling of a PCI 3695 bus and the use of other driver options may interfere 3696 with the sequence. By default IORDY is allowed across 3697 all channels. 3698 3699 pata_legacy.opti82c46x= [HW,LIBATA] 3700 Format: <int> 3701 Set to 1, 2, or 3 for Opti 82c611A on the primary 3702 channel, the secondary channel, or both channels 3703 respectively. Disabled by default. 3704 3705 pata_legacy.opti82c611a= [HW,LIBATA] 3706 Format: <int> 3707 Set to 1, 2, or 3 for Opti 82c465MV on the primary 3708 channel, the secondary channel, or both channels 3709 respectively. Disabled by default. 3710 3711 pata_legacy.pio_mask= [HW,LIBATA] 3712 Format: <int> 3713 PIO mode mask for autospeed devices. Set individual 3714 bits to allow the use of the respective PIO modes. 3715 Bit 0 is for mode 0, bit 1 is for mode 1, and so on. 3716 All modes allowed by default. 3717 3718 pata_legacy.probe_all= [HW,LIBATA] 3719 Format: <int> 3720 Set to non-zero to probe tertiary and further ISA 3721 port ranges on PCI systems. Disabled by default. 3722 3723 pata_legacy.probe_mask= [HW,LIBATA] 3724 Format: <int> 3725 Probe mask for legacy ISA PATA ports. Depending on 3726 platform configuration and the use of other driver 3727 options up to 6 legacy ports are supported: 0x1f0, 3728 0x170, 0x1e8, 0x168, 0x1e0, 0x160, however probing 3729 of individual ports can be disabled by setting the 3730 corresponding bits in the mask to 1. Bit 0 is for 3731 the first port in the list above (0x1f0), and so on. 3732 By default all supported ports are probed. 3733 3734 pata_legacy.qdi= [HW,LIBATA] 3735 Format: <int> 3736 Set to non-zero to probe QDI controllers. By default 3737 set to 1 if CONFIG_PATA_QDI_MODULE, 0 otherwise. 3738 3739 pata_legacy.winbond= [HW,LIBATA] 3740 Format: <int> 3741 Set to non-zero to probe Winbond controllers. Use 3742 the standard I/O port (0x130) if 1, otherwise the 3743 value given is the I/O port to use (typically 0x1b0). 3744 By default set to 1 if CONFIG_PATA_WINBOND_VLB_MODULE, 3745 0 otherwise. 3746 3747 pata_platform.pio_mask= [HW,LIBATA] 3748 Format: <int> 3749 Supported PIO mode mask. Set individual bits to allow 3750 the use of the respective PIO modes. Bit 0 is for 3751 mode 0, bit 1 is for mode 1, and so on. Mode 0 only 3752 allowed by default. 3753 3754 pause_on_oops= 3755 Halt all CPUs after the first oops has been printed for 3756 the specified number of seconds. This is to be used if 3757 your oopses keep scrolling off the screen. 3758 3759 pcbit= [HW,ISDN] 3760 3761 pcd. [PARIDE] 3762 See header of drivers/block/paride/pcd.c. 3763 See also Documentation/admin-guide/blockdev/paride.rst. 3764 3765 pci=option[,option...] [PCI] various PCI subsystem options. 3766 3767 Some options herein operate on a specific device 3768 or a set of devices (<pci_dev>). These are 3769 specified in one of the following formats: 3770 3771 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3772 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3773 3774 Note: the first format specifies a PCI 3775 bus/device/function address which may change 3776 if new hardware is inserted, if motherboard 3777 firmware changes, or due to changes caused 3778 by other kernel parameters. If the 3779 domain is left unspecified, it is 3780 taken to be zero. Optionally, a path 3781 to a device through multiple device/function 3782 addresses can be specified after the base 3783 address (this is more robust against 3784 renumbering issues). The second format 3785 selects devices using IDs from the 3786 configuration space which may match multiple 3787 devices in the system. 3788 3789 earlydump dump PCI config space before the kernel 3790 changes anything 3791 off [X86] don't probe for the PCI bus 3792 bios [X86-32] force use of PCI BIOS, don't access 3793 the hardware directly. Use this if your machine 3794 has a non-standard PCI host bridge. 3795 nobios [X86-32] disallow use of PCI BIOS, only direct 3796 hardware access methods are allowed. Use this 3797 if you experience crashes upon bootup and you 3798 suspect they are caused by the BIOS. 3799 conf1 [X86] Force use of PCI Configuration Access 3800 Mechanism 1 (config address in IO port 0xCF8, 3801 data in IO port 0xCFC, both 32-bit). 3802 conf2 [X86] Force use of PCI Configuration Access 3803 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3804 the function, IO port 0xCFA, also 8-bit, sets 3805 bus number. The config space is then accessed 3806 through ports 0xC000-0xCFFF). 3807 See http://wiki.osdev.org/PCI for more info 3808 on the configuration access mechanisms. 3809 noaer [PCIE] If the PCIEAER kernel config parameter is 3810 enabled, this kernel boot option can be used to 3811 disable the use of PCIE advanced error reporting. 3812 nodomains [PCI] Disable support for multiple PCI 3813 root domains (aka PCI segments, in ACPI-speak). 3814 nommconf [X86] Disable use of MMCONFIG for PCI 3815 Configuration 3816 check_enable_amd_mmconf [X86] check for and enable 3817 properly configured MMIO access to PCI 3818 config space on AMD family 10h CPU 3819 nomsi [MSI] If the PCI_MSI kernel config parameter is 3820 enabled, this kernel boot option can be used to 3821 disable the use of MSI interrupts system-wide. 3822 noioapicquirk [APIC] Disable all boot interrupt quirks. 3823 Safety option to keep boot IRQs enabled. This 3824 should never be necessary. 3825 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3826 primary IO-APIC for bridges that cannot disable 3827 boot IRQs. This fixes a source of spurious IRQs 3828 when the system masks IRQs. 3829 noioapicreroute [APIC] Disable workaround that uses the 3830 boot IRQ equivalent of an IRQ that connects to 3831 a chipset where boot IRQs cannot be disabled. 3832 The opposite of ioapicreroute. 3833 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3834 routing table. These calls are known to be buggy 3835 on several machines and they hang the machine 3836 when used, but on other computers it's the only 3837 way to get the interrupt routing table. Try 3838 this option if the kernel is unable to allocate 3839 IRQs or discover secondary PCI buses on your 3840 motherboard. 3841 rom [X86] Assign address space to expansion ROMs. 3842 Use with caution as certain devices share 3843 address decoders between ROMs and other 3844 resources. 3845 norom [X86] Do not assign address space to 3846 expansion ROMs that do not already have 3847 BIOS assigned address ranges. 3848 nobar [X86] Do not assign address space to the 3849 BARs that weren't assigned by the BIOS. 3850 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3851 assigned automatically to PCI devices. You can 3852 make the kernel exclude IRQs of your ISA cards 3853 this way. 3854 pirqaddr=0xAAAAA [X86] Specify the physical address 3855 of the PIRQ table (normally generated 3856 by the BIOS) if it is outside the 3857 F0000h-100000h range. 3858 lastbus=N [X86] Scan all buses thru bus #N. Can be 3859 useful if the kernel is unable to find your 3860 secondary buses and you want to tell it 3861 explicitly which ones they are. 3862 assign-busses [X86] Always assign all PCI bus 3863 numbers ourselves, overriding 3864 whatever the firmware may have done. 3865 usepirqmask [X86] Honor the possible IRQ mask stored 3866 in the BIOS $PIR table. This is needed on 3867 some systems with broken BIOSes, notably 3868 some HP Pavilion N5400 and Omnibook XE3 3869 notebooks. This will have no effect if ACPI 3870 IRQ routing is enabled. 3871 noacpi [X86] Do not use ACPI for IRQ routing 3872 or for PCI scanning. 3873 use_crs [X86] Use PCI host bridge window information 3874 from ACPI. On BIOSes from 2008 or later, this 3875 is enabled by default. If you need to use this, 3876 please report a bug. 3877 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3878 If you need to use this, please report a bug. 3879 routeirq Do IRQ routing for all PCI devices. 3880 This is normally done in pci_enable_device(), 3881 so this option is a temporary workaround 3882 for broken drivers that don't call it. 3883 skip_isa_align [X86] do not align io start addr, so can 3884 handle more pci cards 3885 noearly [X86] Don't do any early type 1 scanning. 3886 This might help on some broken boards which 3887 machine check when some devices' config space 3888 is read. But various workarounds are disabled 3889 and some IOMMU drivers will not work. 3890 bfsort Sort PCI devices into breadth-first order. 3891 This sorting is done to get a device 3892 order compatible with older (<= 2.4) kernels. 3893 nobfsort Don't sort PCI devices into breadth-first order. 3894 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3895 tuning and use the BIOS-configured MPS defaults. 3896 pcie_bus_safe Set every device's MPS to the largest value 3897 supported by all devices below the root complex. 3898 pcie_bus_perf Set device MPS to the largest allowable MPS 3899 based on its parent bus. Also set MRRS (Max 3900 Read Request Size) to the largest supported 3901 value (no larger than the MPS that the device 3902 or bus can support) for best performance. 3903 pcie_bus_peer2peer Set every device's MPS to 128B, which 3904 every device is guaranteed to support. This 3905 configuration allows peer-to-peer DMA between 3906 any pair of devices, possibly at the cost of 3907 reduced performance. This also guarantees 3908 that hot-added devices will work. 3909 cbiosize=nn[KMG] The fixed amount of bus space which is 3910 reserved for the CardBus bridge's IO window. 3911 The default value is 256 bytes. 3912 cbmemsize=nn[KMG] The fixed amount of bus space which is 3913 reserved for the CardBus bridge's memory 3914 window. The default value is 64 megabytes. 3915 resource_alignment= 3916 Format: 3917 [<order of align>@]<pci_dev>[; ...] 3918 Specifies alignment and device to reassign 3919 aligned memory resources. How to 3920 specify the device is described above. 3921 If <order of align> is not specified, 3922 PAGE_SIZE is used as alignment. 3923 A PCI-PCI bridge can be specified if resource 3924 windows need to be expanded. 3925 To specify the alignment for several 3926 instances of a device, the PCI vendor, 3927 device, subvendor, and subdevice may be 3928 specified, e.g., 12@pci:8086:9c22:103c:198f 3929 for 4096-byte alignment. 3930 ecrc= Enable/disable PCIe ECRC (transaction layer 3931 end-to-end CRC checking). 3932 bios: Use BIOS/firmware settings. This is the 3933 the default. 3934 off: Turn ECRC off 3935 on: Turn ECRC on. 3936 hpiosize=nn[KMG] The fixed amount of bus space which is 3937 reserved for hotplug bridge's IO window. 3938 Default size is 256 bytes. 3939 hpmmiosize=nn[KMG] The fixed amount of bus space which is 3940 reserved for hotplug bridge's MMIO window. 3941 Default size is 2 megabytes. 3942 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 3943 reserved for hotplug bridge's MMIO_PREF window. 3944 Default size is 2 megabytes. 3945 hpmemsize=nn[KMG] The fixed amount of bus space which is 3946 reserved for hotplug bridge's MMIO and 3947 MMIO_PREF window. 3948 Default size is 2 megabytes. 3949 hpbussize=nn The minimum amount of additional bus numbers 3950 reserved for buses below a hotplug bridge. 3951 Default is 1. 3952 realloc= Enable/disable reallocating PCI bridge resources 3953 if allocations done by BIOS are too small to 3954 accommodate resources required by all child 3955 devices. 3956 off: Turn realloc off 3957 on: Turn realloc on 3958 realloc same as realloc=on 3959 noari do not use PCIe ARI. 3960 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3961 do not use PCIe ATS (and IOMMU device IOTLB). 3962 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3963 only look for one device below a PCIe downstream 3964 port. 3965 big_root_window Try to add a big 64bit memory window to the PCIe 3966 root complex on AMD CPUs. Some GFX hardware 3967 can resize a BAR to allow access to all VRAM. 3968 Adding the window is slightly risky (it may 3969 conflict with unreported devices), so this 3970 taints the kernel. 3971 disable_acs_redir=<pci_dev>[; ...] 3972 Specify one or more PCI devices (in the format 3973 specified above) separated by semicolons. 3974 Each device specified will have the PCI ACS 3975 redirect capabilities forced off which will 3976 allow P2P traffic between devices through 3977 bridges without forcing it upstream. Note: 3978 this removes isolation between devices and 3979 may put more devices in an IOMMU group. 3980 force_floating [S390] Force usage of floating interrupts. 3981 nomio [S390] Do not use MIO instructions. 3982 norid [S390] ignore the RID field and force use of 3983 one PCI domain per PCI function 3984 3985 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3986 Management. 3987 off Disable ASPM. 3988 force Enable ASPM even on devices that claim not to support it. 3989 WARNING: Forcing ASPM on may cause system lockups. 3990 3991 pcie_ports= [PCIE] PCIe port services handling: 3992 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3993 even if the platform doesn't give the OS permission to 3994 use them. This may cause conflicts if the platform 3995 also tries to use these services. 3996 dpc-native Use native PCIe service for DPC only. May 3997 cause conflicts if firmware uses AER or DPC. 3998 compat Disable native PCIe services (PME, AER, DPC, PCIe 3999 hotplug). 4000 4001 pcie_port_pm= [PCIE] PCIe port power management handling: 4002 off Disable power management of all PCIe ports 4003 force Forcibly enable power management of all PCIe ports 4004 4005 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 4006 nomsi Do not use MSI for native PCIe PME signaling (this makes 4007 all PCIe root ports use INTx for all services). 4008 4009 pcmv= [HW,PCMCIA] BadgePAD 4 4010 4011 pd_ignore_unused 4012 [PM] 4013 Keep all power-domains already enabled by bootloader on, 4014 even if no driver has claimed them. This is useful 4015 for debug and development, but should not be 4016 needed on a platform with proper driver support. 4017 4018 pd. [PARIDE] 4019 See Documentation/admin-guide/blockdev/paride.rst. 4020 4021 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 4022 boot time. 4023 Format: { 0 | 1 } 4024 See arch/parisc/kernel/pdc_chassis.c 4025 4026 percpu_alloc= Select which percpu first chunk allocator to use. 4027 Currently supported values are "embed" and "page". 4028 Archs may support subset or none of the selections. 4029 See comments in mm/percpu.c for details on each 4030 allocator. This parameter is primarily for debugging 4031 and performance comparison. 4032 4033 pf. [PARIDE] 4034 See Documentation/admin-guide/blockdev/paride.rst. 4035 4036 pg. [PARIDE] 4037 See Documentation/admin-guide/blockdev/paride.rst. 4038 4039 pirq= [SMP,APIC] Manual mp-table setup 4040 See Documentation/x86/i386/IO-APIC.rst. 4041 4042 plip= [PPT,NET] Parallel port network link 4043 Format: { parport<nr> | timid | 0 } 4044 See also Documentation/admin-guide/parport.rst. 4045 4046 pmtmr= [X86] Manual setup of pmtmr I/O Port. 4047 Override pmtimer IOPort with a hex value. 4048 e.g. pmtmr=0x508 4049 4050 pm_debug_messages [SUSPEND,KNL] 4051 Enable suspend/resume debug messages during boot up. 4052 4053 pnp.debug=1 [PNP] 4054 Enable PNP debug messages (depends on the 4055 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 4056 via /sys/module/pnp/parameters/debug. We always show 4057 current resource usage; turning this on also shows 4058 possible settings and some assignment information. 4059 4060 pnpacpi= [ACPI] 4061 { off } 4062 4063 pnpbios= [ISAPNP] 4064 { on | off | curr | res | no-curr | no-res } 4065 4066 pnp_reserve_irq= 4067 [ISAPNP] Exclude IRQs for the autoconfiguration 4068 4069 pnp_reserve_dma= 4070 [ISAPNP] Exclude DMAs for the autoconfiguration 4071 4072 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 4073 Ranges are in pairs (I/O port base and size). 4074 4075 pnp_reserve_mem= 4076 [ISAPNP] Exclude memory regions for the 4077 autoconfiguration. 4078 Ranges are in pairs (memory base and size). 4079 4080 ports= [IP_VS_FTP] IPVS ftp helper module 4081 Default is 21. 4082 Up to 8 (IP_VS_APP_MAX_PORTS) ports 4083 may be specified. 4084 Format: <port>,<port>.... 4085 4086 powersave=off [PPC] This option disables power saving features. 4087 It specifically disables cpuidle and sets the 4088 platform machine description specific power_save 4089 function to NULL. On Idle the CPU just reduces 4090 execution priority. 4091 4092 ppc_strict_facility_enable 4093 [PPC] This option catches any kernel floating point, 4094 Altivec, VSX and SPE outside of regions specifically 4095 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 4096 There is some performance impact when enabling this. 4097 4098 ppc_tm= [PPC] 4099 Format: {"off"} 4100 Disable Hardware Transactional Memory 4101 4102 preempt= [KNL] 4103 Select preemption mode if you have CONFIG_PREEMPT_DYNAMIC 4104 none - Limited to cond_resched() calls 4105 voluntary - Limited to cond_resched() and might_sleep() calls 4106 full - Any section that isn't explicitly preempt disabled 4107 can be preempted anytime. 4108 4109 print-fatal-signals= 4110 [KNL] debug: print fatal signals 4111 4112 If enabled, warn about various signal handling 4113 related application anomalies: too many signals, 4114 too many POSIX.1 timers, fatal signals causing a 4115 coredump - etc. 4116 4117 If you hit the warning due to signal overflow, 4118 you might want to try "ulimit -i unlimited". 4119 4120 default: off. 4121 4122 printk.always_kmsg_dump= 4123 Trigger kmsg_dump for cases other than kernel oops or 4124 panics 4125 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4126 default: disabled 4127 4128 printk.devkmsg={on,off,ratelimit} 4129 Control writing to /dev/kmsg. 4130 on - unlimited logging to /dev/kmsg from userspace 4131 off - logging to /dev/kmsg disabled 4132 ratelimit - ratelimit the logging 4133 Default: ratelimit 4134 4135 printk.time= Show timing data prefixed to each printk message line 4136 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4137 4138 processor.max_cstate= [HW,ACPI] 4139 Limit processor to maximum C-state 4140 max_cstate=9 overrides any DMI blacklist limit. 4141 4142 processor.nocst [HW,ACPI] 4143 Ignore the _CST method to determine C-states, 4144 instead using the legacy FADT method 4145 4146 profile= [KNL] Enable kernel profiling via /proc/profile 4147 Format: [<profiletype>,]<number> 4148 Param: <profiletype>: "schedule", "sleep", or "kvm" 4149 [defaults to kernel profiling] 4150 Param: "schedule" - profile schedule points. 4151 Param: "sleep" - profile D-state sleeping (millisecs). 4152 Requires CONFIG_SCHEDSTATS 4153 Param: "kvm" - profile VM exits. 4154 Param: <number> - step/bucket size as a power of 2 for 4155 statistical time based profiling. 4156 4157 prompt_ramdisk= [RAM] [Deprecated] 4158 4159 prot_virt= [S390] enable hosting protected virtual machines 4160 isolated from the hypervisor (if hardware supports 4161 that). 4162 Format: <bool> 4163 4164 psi= [KNL] Enable or disable pressure stall information 4165 tracking. 4166 Format: <bool> 4167 4168 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 4169 probe for; one of (bare|imps|exps|lifebook|any). 4170 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 4171 per second. 4172 psmouse.resetafter= [HW,MOUSE] 4173 Try to reset the device after so many bad packets 4174 (0 = never). 4175 psmouse.resolution= 4176 [HW,MOUSE] Set desired mouse resolution, in dpi. 4177 psmouse.smartscroll= 4178 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 4179 0 = disabled, 1 = enabled (default). 4180 4181 pstore.backend= Specify the name of the pstore backend to use 4182 4183 pt. [PARIDE] 4184 See Documentation/admin-guide/blockdev/paride.rst. 4185 4186 pti= [X86-64] Control Page Table Isolation of user and 4187 kernel address spaces. Disabling this feature 4188 removes hardening, but improves performance of 4189 system calls and interrupts. 4190 4191 on - unconditionally enable 4192 off - unconditionally disable 4193 auto - kernel detects whether your CPU model is 4194 vulnerable to issues that PTI mitigates 4195 4196 Not specifying this option is equivalent to pti=auto. 4197 4198 nopti [X86-64] 4199 Equivalent to pti=off 4200 4201 pty.legacy_count= 4202 [KNL] Number of legacy pty's. Overwrites compiled-in 4203 default number. 4204 4205 quiet [KNL] Disable most log messages 4206 4207 r128= [HW,DRM] 4208 4209 raid= [HW,RAID] 4210 See Documentation/admin-guide/md.rst. 4211 4212 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 4213 See Documentation/admin-guide/blockdev/ramdisk.rst. 4214 4215 ramdisk_start= [RAM] RAM disk image start address 4216 4217 random.trust_cpu={on,off} 4218 [KNL] Enable or disable trusting the use of the 4219 CPU's random number generator (if available) to 4220 fully seed the kernel's CRNG. Default is controlled 4221 by CONFIG_RANDOM_TRUST_CPU. 4222 4223 randomize_kstack_offset= 4224 [KNL] Enable or disable kernel stack offset 4225 randomization, which provides roughly 5 bits of 4226 entropy, frustrating memory corruption attacks 4227 that depend on stack address determinism or 4228 cross-syscall address exposures. This is only 4229 available on architectures that have defined 4230 CONFIG_HAVE_ARCH_RANDOMIZE_KSTACK_OFFSET. 4231 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4232 Default is CONFIG_RANDOMIZE_KSTACK_OFFSET_DEFAULT. 4233 4234 ras=option[,option,...] [KNL] RAS-specific options 4235 4236 cec_disable [X86] 4237 Disable the Correctable Errors Collector, 4238 see CONFIG_RAS_CEC help text. 4239 4240 rcu_nocbs= [KNL] 4241 The argument is a cpu list, as described above. 4242 4243 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 4244 the specified list of CPUs to be no-callback CPUs. 4245 Invocation of these CPUs' RCU callbacks will be 4246 offloaded to "rcuox/N" kthreads created for that 4247 purpose, where "x" is "p" for RCU-preempt, and 4248 "s" for RCU-sched, and "N" is the CPU number. 4249 This reduces OS jitter on the offloaded CPUs, 4250 which can be useful for HPC and real-time 4251 workloads. It can also improve energy efficiency 4252 for asymmetric multiprocessors. 4253 4254 rcu_nocb_poll [KNL] 4255 Rather than requiring that offloaded CPUs 4256 (specified by rcu_nocbs= above) explicitly 4257 awaken the corresponding "rcuoN" kthreads, 4258 make these kthreads poll for callbacks. 4259 This improves the real-time response for the 4260 offloaded CPUs by relieving them of the need to 4261 wake up the corresponding kthread, but degrades 4262 energy efficiency by requiring that the kthreads 4263 periodically wake up to do the polling. 4264 4265 rcutree.blimit= [KNL] 4266 Set maximum number of finished RCU callbacks to 4267 process in one batch. 4268 4269 rcutree.dump_tree= [KNL] 4270 Dump the structure of the rcu_node combining tree 4271 out at early boot. This is used for diagnostic 4272 purposes, to verify correct tree setup. 4273 4274 rcutree.gp_cleanup_delay= [KNL] 4275 Set the number of jiffies to delay each step of 4276 RCU grace-period cleanup. 4277 4278 rcutree.gp_init_delay= [KNL] 4279 Set the number of jiffies to delay each step of 4280 RCU grace-period initialization. 4281 4282 rcutree.gp_preinit_delay= [KNL] 4283 Set the number of jiffies to delay each step of 4284 RCU grace-period pre-initialization, that is, 4285 the propagation of recent CPU-hotplug changes up 4286 the rcu_node combining tree. 4287 4288 rcutree.use_softirq= [KNL] 4289 If set to zero, move all RCU_SOFTIRQ processing to 4290 per-CPU rcuc kthreads. Defaults to a non-zero 4291 value, meaning that RCU_SOFTIRQ is used by default. 4292 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4293 4294 But note that CONFIG_PREEMPT_RT=y kernels disable 4295 this kernel boot parameter, forcibly setting it 4296 to zero. 4297 4298 rcutree.rcu_fanout_exact= [KNL] 4299 Disable autobalancing of the rcu_node combining 4300 tree. This is used by rcutorture, and might 4301 possibly be useful for architectures having high 4302 cache-to-cache transfer latencies. 4303 4304 rcutree.rcu_fanout_leaf= [KNL] 4305 Change the number of CPUs assigned to each 4306 leaf rcu_node structure. Useful for very 4307 large systems, which will choose the value 64, 4308 and for NUMA systems with large remote-access 4309 latencies, which will choose a value aligned 4310 with the appropriate hardware boundaries. 4311 4312 rcutree.rcu_min_cached_objs= [KNL] 4313 Minimum number of objects which are cached and 4314 maintained per one CPU. Object size is equal 4315 to PAGE_SIZE. The cache allows to reduce the 4316 pressure to page allocator, also it makes the 4317 whole algorithm to behave better in low memory 4318 condition. 4319 4320 rcutree.jiffies_till_first_fqs= [KNL] 4321 Set delay from grace-period initialization to 4322 first attempt to force quiescent states. 4323 Units are jiffies, minimum value is zero, 4324 and maximum value is HZ. 4325 4326 rcutree.jiffies_till_next_fqs= [KNL] 4327 Set delay between subsequent attempts to force 4328 quiescent states. Units are jiffies, minimum 4329 value is one, and maximum value is HZ. 4330 4331 rcutree.jiffies_till_sched_qs= [KNL] 4332 Set required age in jiffies for a 4333 given grace period before RCU starts 4334 soliciting quiescent-state help from 4335 rcu_note_context_switch() and cond_resched(). 4336 If not specified, the kernel will calculate 4337 a value based on the most recent settings 4338 of rcutree.jiffies_till_first_fqs 4339 and rcutree.jiffies_till_next_fqs. 4340 This calculated value may be viewed in 4341 rcutree.jiffies_to_sched_qs. Any attempt to set 4342 rcutree.jiffies_to_sched_qs will be cheerfully 4343 overwritten. 4344 4345 rcutree.kthread_prio= [KNL,BOOT] 4346 Set the SCHED_FIFO priority of the RCU per-CPU 4347 kthreads (rcuc/N). This value is also used for 4348 the priority of the RCU boost threads (rcub/N) 4349 and for the RCU grace-period kthreads (rcu_bh, 4350 rcu_preempt, and rcu_sched). If RCU_BOOST is 4351 set, valid values are 1-99 and the default is 1 4352 (the least-favored priority). Otherwise, when 4353 RCU_BOOST is not set, valid values are 0-99 and 4354 the default is zero (non-realtime operation). 4355 4356 rcutree.rcu_nocb_gp_stride= [KNL] 4357 Set the number of NOCB callback kthreads in 4358 each group, which defaults to the square root 4359 of the number of CPUs. Larger numbers reduce 4360 the wakeup overhead on the global grace-period 4361 kthread, but increases that same overhead on 4362 each group's NOCB grace-period kthread. 4363 4364 rcutree.qhimark= [KNL] 4365 Set threshold of queued RCU callbacks beyond which 4366 batch limiting is disabled. 4367 4368 rcutree.qlowmark= [KNL] 4369 Set threshold of queued RCU callbacks below which 4370 batch limiting is re-enabled. 4371 4372 rcutree.qovld= [KNL] 4373 Set threshold of queued RCU callbacks beyond which 4374 RCU's force-quiescent-state scan will aggressively 4375 enlist help from cond_resched() and sched IPIs to 4376 help CPUs more quickly reach quiescent states. 4377 Set to less than zero to make this be set based 4378 on rcutree.qhimark at boot time and to zero to 4379 disable more aggressive help enlistment. 4380 4381 rcutree.rcu_idle_gp_delay= [KNL] 4382 Set wakeup interval for idle CPUs that have 4383 RCU callbacks (RCU_FAST_NO_HZ=y). 4384 4385 rcutree.rcu_kick_kthreads= [KNL] 4386 Cause the grace-period kthread to get an extra 4387 wake_up() if it sleeps three times longer than 4388 it should at force-quiescent-state time. 4389 This wake_up() will be accompanied by a 4390 WARN_ONCE() splat and an ftrace_dump(). 4391 4392 rcutree.rcu_unlock_delay= [KNL] 4393 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 4394 this specifies an rcu_read_unlock()-time delay 4395 in microseconds. This defaults to zero. 4396 Larger delays increase the probability of 4397 catching RCU pointer leaks, that is, buggy use 4398 of RCU-protected pointers after the relevant 4399 rcu_read_unlock() has completed. 4400 4401 rcutree.sysrq_rcu= [KNL] 4402 Commandeer a sysrq key to dump out Tree RCU's 4403 rcu_node tree with an eye towards determining 4404 why a new grace period has not yet started. 4405 4406 rcuscale.gp_async= [KNL] 4407 Measure performance of asynchronous 4408 grace-period primitives such as call_rcu(). 4409 4410 rcuscale.gp_async_max= [KNL] 4411 Specify the maximum number of outstanding 4412 callbacks per writer thread. When a writer 4413 thread exceeds this limit, it invokes the 4414 corresponding flavor of rcu_barrier() to allow 4415 previously posted callbacks to drain. 4416 4417 rcuscale.gp_exp= [KNL] 4418 Measure performance of expedited synchronous 4419 grace-period primitives. 4420 4421 rcuscale.holdoff= [KNL] 4422 Set test-start holdoff period. The purpose of 4423 this parameter is to delay the start of the 4424 test until boot completes in order to avoid 4425 interference. 4426 4427 rcuscale.kfree_rcu_test= [KNL] 4428 Set to measure performance of kfree_rcu() flooding. 4429 4430 rcuscale.kfree_rcu_test_double= [KNL] 4431 Test the double-argument variant of kfree_rcu(). 4432 If this parameter has the same value as 4433 rcuscale.kfree_rcu_test_single, both the single- 4434 and double-argument variants are tested. 4435 4436 rcuscale.kfree_rcu_test_single= [KNL] 4437 Test the single-argument variant of kfree_rcu(). 4438 If this parameter has the same value as 4439 rcuscale.kfree_rcu_test_double, both the single- 4440 and double-argument variants are tested. 4441 4442 rcuscale.kfree_nthreads= [KNL] 4443 The number of threads running loops of kfree_rcu(). 4444 4445 rcuscale.kfree_alloc_num= [KNL] 4446 Number of allocations and frees done in an iteration. 4447 4448 rcuscale.kfree_loops= [KNL] 4449 Number of loops doing rcuscale.kfree_alloc_num number 4450 of allocations and frees. 4451 4452 rcuscale.nreaders= [KNL] 4453 Set number of RCU readers. The value -1 selects 4454 N, where N is the number of CPUs. A value 4455 "n" less than -1 selects N-n+1, where N is again 4456 the number of CPUs. For example, -2 selects N 4457 (the number of CPUs), -3 selects N+1, and so on. 4458 A value of "n" less than or equal to -N selects 4459 a single reader. 4460 4461 rcuscale.nwriters= [KNL] 4462 Set number of RCU writers. The values operate 4463 the same as for rcuscale.nreaders. 4464 N, where N is the number of CPUs 4465 4466 rcuscale.perf_type= [KNL] 4467 Specify the RCU implementation to test. 4468 4469 rcuscale.shutdown= [KNL] 4470 Shut the system down after performance tests 4471 complete. This is useful for hands-off automated 4472 testing. 4473 4474 rcuscale.verbose= [KNL] 4475 Enable additional printk() statements. 4476 4477 rcuscale.writer_holdoff= [KNL] 4478 Write-side holdoff between grace periods, 4479 in microseconds. The default of zero says 4480 no holdoff. 4481 4482 rcutorture.fqs_duration= [KNL] 4483 Set duration of force_quiescent_state bursts 4484 in microseconds. 4485 4486 rcutorture.fqs_holdoff= [KNL] 4487 Set holdoff time within force_quiescent_state bursts 4488 in microseconds. 4489 4490 rcutorture.fqs_stutter= [KNL] 4491 Set wait time between force_quiescent_state bursts 4492 in seconds. 4493 4494 rcutorture.fwd_progress= [KNL] 4495 Enable RCU grace-period forward-progress testing 4496 for the types of RCU supporting this notion. 4497 4498 rcutorture.fwd_progress_div= [KNL] 4499 Specify the fraction of a CPU-stall-warning 4500 period to do tight-loop forward-progress testing. 4501 4502 rcutorture.fwd_progress_holdoff= [KNL] 4503 Number of seconds to wait between successive 4504 forward-progress tests. 4505 4506 rcutorture.fwd_progress_need_resched= [KNL] 4507 Enclose cond_resched() calls within checks for 4508 need_resched() during tight-loop forward-progress 4509 testing. 4510 4511 rcutorture.gp_cond= [KNL] 4512 Use conditional/asynchronous update-side 4513 primitives, if available. 4514 4515 rcutorture.gp_exp= [KNL] 4516 Use expedited update-side primitives, if available. 4517 4518 rcutorture.gp_normal= [KNL] 4519 Use normal (non-expedited) asynchronous 4520 update-side primitives, if available. 4521 4522 rcutorture.gp_sync= [KNL] 4523 Use normal (non-expedited) synchronous 4524 update-side primitives, if available. If all 4525 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4526 rcutorture.gp_normal=, and rcutorture.gp_sync= 4527 are zero, rcutorture acts as if is interpreted 4528 they are all non-zero. 4529 4530 rcutorture.irqreader= [KNL] 4531 Run RCU readers from irq handlers, or, more 4532 accurately, from a timer handler. Not all RCU 4533 flavors take kindly to this sort of thing. 4534 4535 rcutorture.leakpointer= [KNL] 4536 Leak an RCU-protected pointer out of the reader. 4537 This can of course result in splats, and is 4538 intended to test the ability of things like 4539 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 4540 such leaks. 4541 4542 rcutorture.n_barrier_cbs= [KNL] 4543 Set callbacks/threads for rcu_barrier() testing. 4544 4545 rcutorture.nfakewriters= [KNL] 4546 Set number of concurrent RCU writers. These just 4547 stress RCU, they don't participate in the actual 4548 test, hence the "fake". 4549 4550 rcutorture.nocbs_nthreads= [KNL] 4551 Set number of RCU callback-offload togglers. 4552 Zero (the default) disables toggling. 4553 4554 rcutorture.nocbs_toggle= [KNL] 4555 Set the delay in milliseconds between successive 4556 callback-offload toggling attempts. 4557 4558 rcutorture.nreaders= [KNL] 4559 Set number of RCU readers. The value -1 selects 4560 N-1, where N is the number of CPUs. A value 4561 "n" less than -1 selects N-n-2, where N is again 4562 the number of CPUs. For example, -2 selects N 4563 (the number of CPUs), -3 selects N+1, and so on. 4564 4565 rcutorture.object_debug= [KNL] 4566 Enable debug-object double-call_rcu() testing. 4567 4568 rcutorture.onoff_holdoff= [KNL] 4569 Set time (s) after boot for CPU-hotplug testing. 4570 4571 rcutorture.onoff_interval= [KNL] 4572 Set time (jiffies) between CPU-hotplug operations, 4573 or zero to disable CPU-hotplug testing. 4574 4575 rcutorture.read_exit= [KNL] 4576 Set the number of read-then-exit kthreads used 4577 to test the interaction of RCU updaters and 4578 task-exit processing. 4579 4580 rcutorture.read_exit_burst= [KNL] 4581 The number of times in a given read-then-exit 4582 episode that a set of read-then-exit kthreads 4583 is spawned. 4584 4585 rcutorture.read_exit_delay= [KNL] 4586 The delay, in seconds, between successive 4587 read-then-exit testing episodes. 4588 4589 rcutorture.shuffle_interval= [KNL] 4590 Set task-shuffle interval (s). Shuffling tasks 4591 allows some CPUs to go into dyntick-idle mode 4592 during the rcutorture test. 4593 4594 rcutorture.shutdown_secs= [KNL] 4595 Set time (s) after boot system shutdown. This 4596 is useful for hands-off automated testing. 4597 4598 rcutorture.stall_cpu= [KNL] 4599 Duration of CPU stall (s) to test RCU CPU stall 4600 warnings, zero to disable. 4601 4602 rcutorture.stall_cpu_block= [KNL] 4603 Sleep while stalling if set. This will result 4604 in warnings from preemptible RCU in addition 4605 to any other stall-related activity. 4606 4607 rcutorture.stall_cpu_holdoff= [KNL] 4608 Time to wait (s) after boot before inducing stall. 4609 4610 rcutorture.stall_cpu_irqsoff= [KNL] 4611 Disable interrupts while stalling if set. 4612 4613 rcutorture.stall_gp_kthread= [KNL] 4614 Duration (s) of forced sleep within RCU 4615 grace-period kthread to test RCU CPU stall 4616 warnings, zero to disable. If both stall_cpu 4617 and stall_gp_kthread are specified, the 4618 kthread is starved first, then the CPU. 4619 4620 rcutorture.stat_interval= [KNL] 4621 Time (s) between statistics printk()s. 4622 4623 rcutorture.stutter= [KNL] 4624 Time (s) to stutter testing, for example, specifying 4625 five seconds causes the test to run for five seconds, 4626 wait for five seconds, and so on. This tests RCU's 4627 ability to transition abruptly to and from idle. 4628 4629 rcutorture.test_boost= [KNL] 4630 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4631 "Maybe" means test if the RCU implementation 4632 under test support RCU priority boosting. 4633 4634 rcutorture.test_boost_duration= [KNL] 4635 Duration (s) of each individual boost test. 4636 4637 rcutorture.test_boost_interval= [KNL] 4638 Interval (s) between each boost test. 4639 4640 rcutorture.test_no_idle_hz= [KNL] 4641 Test RCU's dyntick-idle handling. See also the 4642 rcutorture.shuffle_interval parameter. 4643 4644 rcutorture.torture_type= [KNL] 4645 Specify the RCU implementation to test. 4646 4647 rcutorture.verbose= [KNL] 4648 Enable additional printk() statements. 4649 4650 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4651 Dump ftrace buffer after reporting RCU CPU 4652 stall warning. 4653 4654 rcupdate.rcu_cpu_stall_suppress= [KNL] 4655 Suppress RCU CPU stall warning messages. 4656 4657 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 4658 Suppress RCU CPU stall warning messages and 4659 rcutorture writer stall warnings that occur 4660 during early boot, that is, during the time 4661 before the init task is spawned. 4662 4663 rcupdate.rcu_cpu_stall_timeout= [KNL] 4664 Set timeout for RCU CPU stall warning messages. 4665 4666 rcupdate.rcu_expedited= [KNL] 4667 Use expedited grace-period primitives, for 4668 example, synchronize_rcu_expedited() instead 4669 of synchronize_rcu(). This reduces latency, 4670 but can increase CPU utilization, degrade 4671 real-time latency, and degrade energy efficiency. 4672 No effect on CONFIG_TINY_RCU kernels. 4673 4674 rcupdate.rcu_normal= [KNL] 4675 Use only normal grace-period primitives, 4676 for example, synchronize_rcu() instead of 4677 synchronize_rcu_expedited(). This improves 4678 real-time latency, CPU utilization, and 4679 energy efficiency, but can expose users to 4680 increased grace-period latency. This parameter 4681 overrides rcupdate.rcu_expedited. No effect on 4682 CONFIG_TINY_RCU kernels. 4683 4684 rcupdate.rcu_normal_after_boot= [KNL] 4685 Once boot has completed (that is, after 4686 rcu_end_inkernel_boot() has been invoked), use 4687 only normal grace-period primitives. No effect 4688 on CONFIG_TINY_RCU kernels. 4689 4690 But note that CONFIG_PREEMPT_RT=y kernels enables 4691 this kernel boot parameter, forcibly setting 4692 it to the value one, that is, converting any 4693 post-boot attempt at an expedited RCU grace 4694 period to instead use normal non-expedited 4695 grace-period processing. 4696 4697 rcupdate.rcu_task_ipi_delay= [KNL] 4698 Set time in jiffies during which RCU tasks will 4699 avoid sending IPIs, starting with the beginning 4700 of a given grace period. Setting a large 4701 number avoids disturbing real-time workloads, 4702 but lengthens grace periods. 4703 4704 rcupdate.rcu_task_stall_timeout= [KNL] 4705 Set timeout in jiffies for RCU task stall warning 4706 messages. Disable with a value less than or equal 4707 to zero. 4708 4709 rcupdate.rcu_self_test= [KNL] 4710 Run the RCU early boot self tests 4711 4712 rdinit= [KNL] 4713 Format: <full_path> 4714 Run specified binary instead of /init from the ramdisk, 4715 used for early userspace startup. See initrd. 4716 4717 rdrand= [X86] 4718 force - Override the decision by the kernel to hide the 4719 advertisement of RDRAND support (this affects 4720 certain AMD processors because of buggy BIOS 4721 support, specifically around the suspend/resume 4722 path). 4723 4724 rdt= [HW,X86,RDT] 4725 Turn on/off individual RDT features. List is: 4726 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4727 mba. 4728 E.g. to turn on cmt and turn off mba use: 4729 rdt=cmt,!mba 4730 4731 reboot= [KNL] 4732 Format (x86 or x86_64): 4733 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4734 [[,]s[mp]#### \ 4735 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4736 [[,]f[orce] 4737 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4738 (prefix with 'panic_' to set mode for panic 4739 reboot only), 4740 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4741 reboot_force is either force or not specified, 4742 reboot_cpu is s[mp]#### with #### being the processor 4743 to be used for rebooting. 4744 4745 refscale.holdoff= [KNL] 4746 Set test-start holdoff period. The purpose of 4747 this parameter is to delay the start of the 4748 test until boot completes in order to avoid 4749 interference. 4750 4751 refscale.loops= [KNL] 4752 Set the number of loops over the synchronization 4753 primitive under test. Increasing this number 4754 reduces noise due to loop start/end overhead, 4755 but the default has already reduced the per-pass 4756 noise to a handful of picoseconds on ca. 2020 4757 x86 laptops. 4758 4759 refscale.nreaders= [KNL] 4760 Set number of readers. The default value of -1 4761 selects N, where N is roughly 75% of the number 4762 of CPUs. A value of zero is an interesting choice. 4763 4764 refscale.nruns= [KNL] 4765 Set number of runs, each of which is dumped onto 4766 the console log. 4767 4768 refscale.readdelay= [KNL] 4769 Set the read-side critical-section duration, 4770 measured in microseconds. 4771 4772 refscale.scale_type= [KNL] 4773 Specify the read-protection implementation to test. 4774 4775 refscale.shutdown= [KNL] 4776 Shut down the system at the end of the performance 4777 test. This defaults to 1 (shut it down) when 4778 refscale is built into the kernel and to 0 (leave 4779 it running) when refscale is built as a module. 4780 4781 refscale.verbose= [KNL] 4782 Enable additional printk() statements. 4783 4784 refscale.verbose_batched= [KNL] 4785 Batch the additional printk() statements. If zero 4786 (the default) or negative, print everything. Otherwise, 4787 print every Nth verbose statement, where N is the value 4788 specified. 4789 4790 relax_domain_level= 4791 [KNL, SMP] Set scheduler's default relax_domain_level. 4792 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4793 4794 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4795 Format: <base1>,<size1>[,<base2>,<size2>,...] 4796 Reserve I/O ports or memory so the kernel won't use 4797 them. If <base> is less than 0x10000, the region 4798 is assumed to be I/O ports; otherwise it is memory. 4799 4800 reservetop= [X86-32] 4801 Format: nn[KMG] 4802 Reserves a hole at the top of the kernel virtual 4803 address space. 4804 4805 reservelow= [X86] 4806 Format: nn[K] 4807 Set the amount of memory to reserve for BIOS at 4808 the bottom of the address space. 4809 4810 reset_devices [KNL] Force drivers to reset the underlying device 4811 during initialization. 4812 4813 resume= [SWSUSP] 4814 Specify the partition device for software suspend 4815 Format: 4816 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4817 4818 resume_offset= [SWSUSP] 4819 Specify the offset from the beginning of the partition 4820 given by "resume=" at which the swap header is located, 4821 in <PAGE_SIZE> units (needed only for swap files). 4822 See Documentation/power/swsusp-and-swap-files.rst 4823 4824 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4825 read the resume files 4826 4827 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4828 Useful for devices that are detected asynchronously 4829 (e.g. USB and MMC devices). 4830 4831 hibernate= [HIBERNATION] 4832 noresume Don't check if there's a hibernation image 4833 present during boot. 4834 nocompress Don't compress/decompress hibernation images. 4835 no Disable hibernation and resume. 4836 protect_image Turn on image protection during restoration 4837 (that will set all pages holding image data 4838 during restoration read-only). 4839 4840 retain_initrd [RAM] Keep initrd memory after extraction 4841 4842 rfkill.default_state= 4843 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4844 etc. communication is blocked by default. 4845 1 Unblocked. 4846 4847 rfkill.master_switch_mode= 4848 0 The "airplane mode" button does nothing. 4849 1 The "airplane mode" button toggles between everything 4850 blocked and the previous configuration. 4851 2 The "airplane mode" button toggles between everything 4852 blocked and everything unblocked. 4853 4854 rhash_entries= [KNL,NET] 4855 Set number of hash buckets for route cache 4856 4857 ring3mwait=disable 4858 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4859 CPUs. 4860 4861 ro [KNL] Mount root device read-only on boot 4862 4863 rodata= [KNL] 4864 on Mark read-only kernel memory as read-only (default). 4865 off Leave read-only kernel memory writable for debugging. 4866 4867 rockchip.usb_uart 4868 Enable the uart passthrough on the designated usb port 4869 on Rockchip SoCs. When active, the signals of the 4870 debug-uart get routed to the D+ and D- pins of the usb 4871 port and the regular usb controller gets disabled. 4872 4873 root= [KNL] Root filesystem 4874 See name_to_dev_t comment in init/do_mounts.c. 4875 4876 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4877 mount the root filesystem 4878 4879 rootflags= [KNL] Set root filesystem mount option string 4880 4881 rootfstype= [KNL] Set root filesystem type 4882 4883 rootwait [KNL] Wait (indefinitely) for root device to show up. 4884 Useful for devices that are detected asynchronously 4885 (e.g. USB and MMC devices). 4886 4887 rproc_mem=nn[KMG][@address] 4888 [KNL,ARM,CMA] Remoteproc physical memory block. 4889 Memory area to be used by remote processor image, 4890 managed by CMA. 4891 4892 rw [KNL] Mount root device read-write on boot 4893 4894 S [KNL] Run init in single mode 4895 4896 s390_iommu= [HW,S390] 4897 Set s390 IOTLB flushing mode 4898 strict 4899 With strict flushing every unmap operation will result in 4900 an IOTLB flush. Default is lazy flushing before reuse, 4901 which is faster. 4902 4903 sa1100ir [NET] 4904 See drivers/net/irda/sa1100_ir.c. 4905 4906 sbni= [NET] Granch SBNI12 leased line adapter 4907 4908 sched_verbose [KNL] Enables verbose scheduler debug messages. 4909 4910 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4911 Allowed values are enable and disable. This feature 4912 incurs a small amount of overhead in the scheduler 4913 but is useful for debugging and performance tuning. 4914 4915 sched_thermal_decay_shift= 4916 [KNL, SMP] Set a decay shift for scheduler thermal 4917 pressure signal. Thermal pressure signal follows the 4918 default decay period of other scheduler pelt 4919 signals(usually 32 ms but configurable). Setting 4920 sched_thermal_decay_shift will left shift the decay 4921 period for the thermal pressure signal by the shift 4922 value. 4923 i.e. with the default pelt decay period of 32 ms 4924 sched_thermal_decay_shift thermal pressure decay pr 4925 1 64 ms 4926 2 128 ms 4927 and so on. 4928 Format: integer between 0 and 10 4929 Default is 0. 4930 4931 scftorture.holdoff= [KNL] 4932 Number of seconds to hold off before starting 4933 test. Defaults to zero for module insertion and 4934 to 10 seconds for built-in smp_call_function() 4935 tests. 4936 4937 scftorture.longwait= [KNL] 4938 Request ridiculously long waits randomly selected 4939 up to the chosen limit in seconds. Zero (the 4940 default) disables this feature. Please note 4941 that requesting even small non-zero numbers of 4942 seconds can result in RCU CPU stall warnings, 4943 softlockup complaints, and so on. 4944 4945 scftorture.nthreads= [KNL] 4946 Number of kthreads to spawn to invoke the 4947 smp_call_function() family of functions. 4948 The default of -1 specifies a number of kthreads 4949 equal to the number of CPUs. 4950 4951 scftorture.onoff_holdoff= [KNL] 4952 Number seconds to wait after the start of the 4953 test before initiating CPU-hotplug operations. 4954 4955 scftorture.onoff_interval= [KNL] 4956 Number seconds to wait between successive 4957 CPU-hotplug operations. Specifying zero (which 4958 is the default) disables CPU-hotplug operations. 4959 4960 scftorture.shutdown_secs= [KNL] 4961 The number of seconds following the start of the 4962 test after which to shut down the system. The 4963 default of zero avoids shutting down the system. 4964 Non-zero values are useful for automated tests. 4965 4966 scftorture.stat_interval= [KNL] 4967 The number of seconds between outputting the 4968 current test statistics to the console. A value 4969 of zero disables statistics output. 4970 4971 scftorture.stutter_cpus= [KNL] 4972 The number of jiffies to wait between each change 4973 to the set of CPUs under test. 4974 4975 scftorture.use_cpus_read_lock= [KNL] 4976 Use use_cpus_read_lock() instead of the default 4977 preempt_disable() to disable CPU hotplug 4978 while invoking one of the smp_call_function*() 4979 functions. 4980 4981 scftorture.verbose= [KNL] 4982 Enable additional printk() statements. 4983 4984 scftorture.weight_single= [KNL] 4985 The probability weighting to use for the 4986 smp_call_function_single() function with a zero 4987 "wait" parameter. A value of -1 selects the 4988 default if all other weights are -1. However, 4989 if at least one weight has some other value, a 4990 value of -1 will instead select a weight of zero. 4991 4992 scftorture.weight_single_wait= [KNL] 4993 The probability weighting to use for the 4994 smp_call_function_single() function with a 4995 non-zero "wait" parameter. See weight_single. 4996 4997 scftorture.weight_many= [KNL] 4998 The probability weighting to use for the 4999 smp_call_function_many() function with a zero 5000 "wait" parameter. See weight_single. 5001 Note well that setting a high probability for 5002 this weighting can place serious IPI load 5003 on the system. 5004 5005 scftorture.weight_many_wait= [KNL] 5006 The probability weighting to use for the 5007 smp_call_function_many() function with a 5008 non-zero "wait" parameter. See weight_single 5009 and weight_many. 5010 5011 scftorture.weight_all= [KNL] 5012 The probability weighting to use for the 5013 smp_call_function_all() function with a zero 5014 "wait" parameter. See weight_single and 5015 weight_many. 5016 5017 scftorture.weight_all_wait= [KNL] 5018 The probability weighting to use for the 5019 smp_call_function_all() function with a 5020 non-zero "wait" parameter. See weight_single 5021 and weight_many. 5022 5023 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 5024 xtime_lock contention on larger systems, and/or RCU lock 5025 contention on all systems with CONFIG_MAXSMP set. 5026 Format: { "0" | "1" } 5027 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 5028 1 -- enable. 5029 Note: increases power consumption, thus should only be 5030 enabled if running jitter sensitive (HPC/RT) workloads. 5031 5032 security= [SECURITY] Choose a legacy "major" security module to 5033 enable at boot. This has been deprecated by the 5034 "lsm=" parameter. 5035 5036 selinux= [SELINUX] Disable or enable SELinux at boot time. 5037 Format: { "0" | "1" } 5038 See security/selinux/Kconfig help text. 5039 0 -- disable. 5040 1 -- enable. 5041 Default value is 1. 5042 5043 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 5044 Format: { "0" | "1" } 5045 See security/apparmor/Kconfig help text 5046 0 -- disable. 5047 1 -- enable. 5048 Default value is set via kernel config option. 5049 5050 serialnumber [BUGS=X86-32] 5051 5052 shapers= [NET] 5053 Maximal number of shapers. 5054 5055 simeth= [IA-64] 5056 simscsi= 5057 5058 slram= [HW,MTD] 5059 5060 slab_merge [MM] 5061 Enable merging of slabs with similar size when the 5062 kernel is built without CONFIG_SLAB_MERGE_DEFAULT. 5063 5064 slab_nomerge [MM] 5065 Disable merging of slabs with similar size. May be 5066 necessary if there is some reason to distinguish 5067 allocs to different slabs, especially in hardened 5068 environments where the risk of heap overflows and 5069 layout control by attackers can usually be 5070 frustrated by disabling merging. This will reduce 5071 most of the exposure of a heap attack to a single 5072 cache (risks via metadata attacks are mostly 5073 unchanged). Debug options disable merging on their 5074 own. 5075 For more information see Documentation/vm/slub.rst. 5076 5077 slab_max_order= [MM, SLAB] 5078 Determines the maximum allowed order for slabs. 5079 A high setting may cause OOMs due to memory 5080 fragmentation. Defaults to 1 for systems with 5081 more than 32MB of RAM, 0 otherwise. 5082 5083 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB] 5084 Enabling slub_debug allows one to determine the 5085 culprit if slab objects become corrupted. Enabling 5086 slub_debug can create guard zones around objects and 5087 may poison objects when not in use. Also tracks the 5088 last alloc / free. For more information see 5089 Documentation/vm/slub.rst. 5090 5091 slub_max_order= [MM, SLUB] 5092 Determines the maximum allowed order for slabs. 5093 A high setting may cause OOMs due to memory 5094 fragmentation. For more information see 5095 Documentation/vm/slub.rst. 5096 5097 slub_min_objects= [MM, SLUB] 5098 The minimum number of objects per slab. SLUB will 5099 increase the slab order up to slub_max_order to 5100 generate a sufficiently large slab able to contain 5101 the number of objects indicated. The higher the number 5102 of objects the smaller the overhead of tracking slabs 5103 and the less frequently locks need to be acquired. 5104 For more information see Documentation/vm/slub.rst. 5105 5106 slub_min_order= [MM, SLUB] 5107 Determines the minimum page order for slabs. Must be 5108 lower than slub_max_order. 5109 For more information see Documentation/vm/slub.rst. 5110 5111 slub_merge [MM, SLUB] 5112 Same with slab_merge. 5113 5114 slub_nomerge [MM, SLUB] 5115 Same with slab_nomerge. This is supported for legacy. 5116 See slab_nomerge for more information. 5117 5118 smart2= [HW] 5119 Format: <io1>[,<io2>[,...,<io8>]] 5120 5121 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 5122 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 5123 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 5124 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 5125 smsc-ircc2.ircc_irq= [HW] IRQ line 5126 smsc-ircc2.ircc_dma= [HW] DMA channel 5127 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 5128 0: Toshiba Satellite 1800 (GP data pin select) 5129 1: Fast pin select (default) 5130 2: ATC IRMode 5131 5132 smt [KNL,S390] Set the maximum number of threads (logical 5133 CPUs) to use per physical CPU on systems capable of 5134 symmetric multithreading (SMT). Will be capped to the 5135 actual hardware limit. 5136 Format: <integer> 5137 Default: -1 (no limit) 5138 5139 softlockup_panic= 5140 [KNL] Should the soft-lockup detector generate panics. 5141 Format: 0 | 1 5142 5143 A value of 1 instructs the soft-lockup detector 5144 to panic the machine when a soft-lockup occurs. It is 5145 also controlled by the kernel.softlockup_panic sysctl 5146 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 5147 respective build-time switch to that functionality. 5148 5149 softlockup_all_cpu_backtrace= 5150 [KNL] Should the soft-lockup detector generate 5151 backtraces on all cpus. 5152 Format: 0 | 1 5153 5154 sonypi.*= [HW] Sony Programmable I/O Control Device driver 5155 See Documentation/admin-guide/laptops/sonypi.rst 5156 5157 spectre_v2= [X86] Control mitigation of Spectre variant 2 5158 (indirect branch speculation) vulnerability. 5159 The default operation protects the kernel from 5160 user space attacks. 5161 5162 on - unconditionally enable, implies 5163 spectre_v2_user=on 5164 off - unconditionally disable, implies 5165 spectre_v2_user=off 5166 auto - kernel detects whether your CPU model is 5167 vulnerable 5168 5169 Selecting 'on' will, and 'auto' may, choose a 5170 mitigation method at run time according to the 5171 CPU, the available microcode, the setting of the 5172 CONFIG_RETPOLINE configuration option, and the 5173 compiler with which the kernel was built. 5174 5175 Selecting 'on' will also enable the mitigation 5176 against user space to user space task attacks. 5177 5178 Selecting 'off' will disable both the kernel and 5179 the user space protections. 5180 5181 Specific mitigations can also be selected manually: 5182 5183 retpoline - replace indirect branches 5184 retpoline,generic - google's original retpoline 5185 retpoline,amd - AMD-specific minimal thunk 5186 5187 Not specifying this option is equivalent to 5188 spectre_v2=auto. 5189 5190 spectre_v2_user= 5191 [X86] Control mitigation of Spectre variant 2 5192 (indirect branch speculation) vulnerability between 5193 user space tasks 5194 5195 on - Unconditionally enable mitigations. Is 5196 enforced by spectre_v2=on 5197 5198 off - Unconditionally disable mitigations. Is 5199 enforced by spectre_v2=off 5200 5201 prctl - Indirect branch speculation is enabled, 5202 but mitigation can be enabled via prctl 5203 per thread. The mitigation control state 5204 is inherited on fork. 5205 5206 prctl,ibpb 5207 - Like "prctl" above, but only STIBP is 5208 controlled per thread. IBPB is issued 5209 always when switching between different user 5210 space processes. 5211 5212 seccomp 5213 - Same as "prctl" above, but all seccomp 5214 threads will enable the mitigation unless 5215 they explicitly opt out. 5216 5217 seccomp,ibpb 5218 - Like "seccomp" above, but only STIBP is 5219 controlled per thread. IBPB is issued 5220 always when switching between different 5221 user space processes. 5222 5223 auto - Kernel selects the mitigation depending on 5224 the available CPU features and vulnerability. 5225 5226 Default mitigation: 5227 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 5228 5229 Not specifying this option is equivalent to 5230 spectre_v2_user=auto. 5231 5232 spec_store_bypass_disable= 5233 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 5234 (Speculative Store Bypass vulnerability) 5235 5236 Certain CPUs are vulnerable to an exploit against a 5237 a common industry wide performance optimization known 5238 as "Speculative Store Bypass" in which recent stores 5239 to the same memory location may not be observed by 5240 later loads during speculative execution. The idea 5241 is that such stores are unlikely and that they can 5242 be detected prior to instruction retirement at the 5243 end of a particular speculation execution window. 5244 5245 In vulnerable processors, the speculatively forwarded 5246 store can be used in a cache side channel attack, for 5247 example to read memory to which the attacker does not 5248 directly have access (e.g. inside sandboxed code). 5249 5250 This parameter controls whether the Speculative Store 5251 Bypass optimization is used. 5252 5253 On x86 the options are: 5254 5255 on - Unconditionally disable Speculative Store Bypass 5256 off - Unconditionally enable Speculative Store Bypass 5257 auto - Kernel detects whether the CPU model contains an 5258 implementation of Speculative Store Bypass and 5259 picks the most appropriate mitigation. If the 5260 CPU is not vulnerable, "off" is selected. If the 5261 CPU is vulnerable the default mitigation is 5262 architecture and Kconfig dependent. See below. 5263 prctl - Control Speculative Store Bypass per thread 5264 via prctl. Speculative Store Bypass is enabled 5265 for a process by default. The state of the control 5266 is inherited on fork. 5267 seccomp - Same as "prctl" above, but all seccomp threads 5268 will disable SSB unless they explicitly opt out. 5269 5270 Default mitigations: 5271 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 5272 5273 On powerpc the options are: 5274 5275 on,auto - On Power8 and Power9 insert a store-forwarding 5276 barrier on kernel entry and exit. On Power7 5277 perform a software flush on kernel entry and 5278 exit. 5279 off - No action. 5280 5281 Not specifying this option is equivalent to 5282 spec_store_bypass_disable=auto. 5283 5284 spia_io_base= [HW,MTD] 5285 spia_fio_base= 5286 spia_pedr= 5287 spia_peddr= 5288 5289 split_lock_detect= 5290 [X86] Enable split lock detection or bus lock detection 5291 5292 When enabled (and if hardware support is present), atomic 5293 instructions that access data across cache line 5294 boundaries will result in an alignment check exception 5295 for split lock detection or a debug exception for 5296 bus lock detection. 5297 5298 off - not enabled 5299 5300 warn - the kernel will emit rate-limited warnings 5301 about applications triggering the #AC 5302 exception or the #DB exception. This mode is 5303 the default on CPUs that support split lock 5304 detection or bus lock detection. Default 5305 behavior is by #AC if both features are 5306 enabled in hardware. 5307 5308 fatal - the kernel will send SIGBUS to applications 5309 that trigger the #AC exception or the #DB 5310 exception. Default behavior is by #AC if 5311 both features are enabled in hardware. 5312 5313 If an #AC exception is hit in the kernel or in 5314 firmware (i.e. not while executing in user mode) 5315 the kernel will oops in either "warn" or "fatal" 5316 mode. 5317 5318 #DB exception for bus lock is triggered only when 5319 CPL > 0. 5320 5321 srbds= [X86,INTEL] 5322 Control the Special Register Buffer Data Sampling 5323 (SRBDS) mitigation. 5324 5325 Certain CPUs are vulnerable to an MDS-like 5326 exploit which can leak bits from the random 5327 number generator. 5328 5329 By default, this issue is mitigated by 5330 microcode. However, the microcode fix can cause 5331 the RDRAND and RDSEED instructions to become 5332 much slower. Among other effects, this will 5333 result in reduced throughput from /dev/urandom. 5334 5335 The microcode mitigation can be disabled with 5336 the following option: 5337 5338 off: Disable mitigation and remove 5339 performance impact to RDRAND and RDSEED 5340 5341 srcutree.counter_wrap_check [KNL] 5342 Specifies how frequently to check for 5343 grace-period sequence counter wrap for the 5344 srcu_data structure's ->srcu_gp_seq_needed field. 5345 The greater the number of bits set in this kernel 5346 parameter, the less frequently counter wrap will 5347 be checked for. Note that the bottom two bits 5348 are ignored. 5349 5350 srcutree.exp_holdoff [KNL] 5351 Specifies how many nanoseconds must elapse 5352 since the end of the last SRCU grace period for 5353 a given srcu_struct until the next normal SRCU 5354 grace period will be considered for automatic 5355 expediting. Set to zero to disable automatic 5356 expediting. 5357 5358 ssbd= [ARM64,HW] 5359 Speculative Store Bypass Disable control 5360 5361 On CPUs that are vulnerable to the Speculative 5362 Store Bypass vulnerability and offer a 5363 firmware based mitigation, this parameter 5364 indicates how the mitigation should be used: 5365 5366 force-on: Unconditionally enable mitigation for 5367 for both kernel and userspace 5368 force-off: Unconditionally disable mitigation for 5369 for both kernel and userspace 5370 kernel: Always enable mitigation in the 5371 kernel, and offer a prctl interface 5372 to allow userspace to register its 5373 interest in being mitigated too. 5374 5375 stack_guard_gap= [MM] 5376 override the default stack gap protection. The value 5377 is in page units and it defines how many pages prior 5378 to (for stacks growing down) resp. after (for stacks 5379 growing up) the main stack are reserved for no other 5380 mapping. Default value is 256 pages. 5381 5382 stack_depot_disable= [KNL] 5383 Setting this to true through kernel command line will 5384 disable the stack depot thereby saving the static memory 5385 consumed by the stack hash table. By default this is set 5386 to false. 5387 5388 stacktrace [FTRACE] 5389 Enabled the stack tracer on boot up. 5390 5391 stacktrace_filter=[function-list] 5392 [FTRACE] Limit the functions that the stack tracer 5393 will trace at boot up. function-list is a comma-separated 5394 list of functions. This list can be changed at run 5395 time by the stack_trace_filter file in the debugfs 5396 tracing directory. Note, this enables stack tracing 5397 and the stacktrace above is not needed. 5398 5399 sti= [PARISC,HW] 5400 Format: <num> 5401 Set the STI (builtin display/keyboard on the HP-PARISC 5402 machines) console (graphic card) which should be used 5403 as the initial boot-console. 5404 See also comment in drivers/video/console/sticore.c. 5405 5406 sti_font= [HW] 5407 See comment in drivers/video/console/sticore.c. 5408 5409 stifb= [HW] 5410 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 5411 5412 sunrpc.min_resvport= 5413 sunrpc.max_resvport= 5414 [NFS,SUNRPC] 5415 SunRPC servers often require that client requests 5416 originate from a privileged port (i.e. a port in the 5417 range 0 < portnr < 1024). 5418 An administrator who wishes to reserve some of these 5419 ports for other uses may adjust the range that the 5420 kernel's sunrpc client considers to be privileged 5421 using these two parameters to set the minimum and 5422 maximum port values. 5423 5424 sunrpc.svc_rpc_per_connection_limit= 5425 [NFS,SUNRPC] 5426 Limit the number of requests that the server will 5427 process in parallel from a single connection. 5428 The default value is 0 (no limit). 5429 5430 sunrpc.pool_mode= 5431 [NFS] 5432 Control how the NFS server code allocates CPUs to 5433 service thread pools. Depending on how many NICs 5434 you have and where their interrupts are bound, this 5435 option will affect which CPUs will do NFS serving. 5436 Note: this parameter cannot be changed while the 5437 NFS server is running. 5438 5439 auto the server chooses an appropriate mode 5440 automatically using heuristics 5441 global a single global pool contains all CPUs 5442 percpu one pool for each CPU 5443 pernode one pool for each NUMA node (equivalent 5444 to global on non-NUMA machines) 5445 5446 sunrpc.tcp_slot_table_entries= 5447 sunrpc.udp_slot_table_entries= 5448 [NFS,SUNRPC] 5449 Sets the upper limit on the number of simultaneous 5450 RPC calls that can be sent from the client to a 5451 server. Increasing these values may allow you to 5452 improve throughput, but will also increase the 5453 amount of memory reserved for use by the client. 5454 5455 suspend.pm_test_delay= 5456 [SUSPEND] 5457 Sets the number of seconds to remain in a suspend test 5458 mode before resuming the system (see 5459 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 5460 is set. Default value is 5. 5461 5462 svm= [PPC] 5463 Format: { on | off | y | n | 1 | 0 } 5464 This parameter controls use of the Protected 5465 Execution Facility on pSeries. 5466 5467 swapaccount=[0|1] 5468 [KNL] Enable accounting of swap in memory resource 5469 controller if no parameter or 1 is given or disable 5470 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 5471 5472 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 5473 Format: { <int> | force | noforce } 5474 <int> -- Number of I/O TLB slabs 5475 force -- force using of bounce buffers even if they 5476 wouldn't be automatically used by the kernel 5477 noforce -- Never use bounce buffers (for debugging) 5478 5479 switches= [HW,M68k] 5480 5481 sysctl.*= [KNL] 5482 Set a sysctl parameter, right before loading the init 5483 process, as if the value was written to the respective 5484 /proc/sys/... file. Both '.' and '/' are recognized as 5485 separators. Unrecognized parameters and invalid values 5486 are reported in the kernel log. Sysctls registered 5487 later by a loaded module cannot be set this way. 5488 Example: sysctl.vm.swappiness=40 5489 5490 sysfs.deprecated=0|1 [KNL] 5491 Enable/disable old style sysfs layout for old udev 5492 on older distributions. When this option is enabled 5493 very new udev will not work anymore. When this option 5494 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 5495 in older udev will not work anymore. 5496 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 5497 the kernel configuration. 5498 5499 sysrq_always_enabled 5500 [KNL] 5501 Ignore sysrq setting - this boot parameter will 5502 neutralize any effect of /proc/sys/kernel/sysrq. 5503 Useful for debugging. 5504 5505 tcpmhash_entries= [KNL,NET] 5506 Set the number of tcp_metrics_hash slots. 5507 Default value is 8192 or 16384 depending on total 5508 ram pages. This is used to specify the TCP metrics 5509 cache size. See Documentation/networking/ip-sysctl.rst 5510 "tcp_no_metrics_save" section for more details. 5511 5512 tdfx= [HW,DRM] 5513 5514 test_suspend= [SUSPEND][,N] 5515 Specify "mem" (for Suspend-to-RAM) or "standby" (for 5516 standby suspend) or "freeze" (for suspend type freeze) 5517 as the system sleep state during system startup with 5518 the optional capability to repeat N number of times. 5519 The system is woken from this state using a 5520 wakeup-capable RTC alarm. 5521 5522 thash_entries= [KNL,NET] 5523 Set number of hash buckets for TCP connection 5524 5525 thermal.act= [HW,ACPI] 5526 -1: disable all active trip points in all thermal zones 5527 <degrees C>: override all lowest active trip points 5528 5529 thermal.crt= [HW,ACPI] 5530 -1: disable all critical trip points in all thermal zones 5531 <degrees C>: override all critical trip points 5532 5533 thermal.nocrt= [HW,ACPI] 5534 Set to disable actions on ACPI thermal zone 5535 critical and hot trip points. 5536 5537 thermal.off= [HW,ACPI] 5538 1: disable ACPI thermal control 5539 5540 thermal.psv= [HW,ACPI] 5541 -1: disable all passive trip points 5542 <degrees C>: override all passive trip points to this 5543 value 5544 5545 thermal.tzp= [HW,ACPI] 5546 Specify global default ACPI thermal zone polling rate 5547 <deci-seconds>: poll all this frequency 5548 0: no polling (default) 5549 5550 threadirqs [KNL] 5551 Force threading of all interrupt handlers except those 5552 marked explicitly IRQF_NO_THREAD. 5553 5554 topology= [S390] 5555 Format: {off | on} 5556 Specify if the kernel should make use of the cpu 5557 topology information if the hardware supports this. 5558 The scheduler will make use of this information and 5559 e.g. base its process migration decisions on it. 5560 Default is on. 5561 5562 topology_updates= [KNL, PPC, NUMA] 5563 Format: {off} 5564 Specify if the kernel should ignore (off) 5565 topology updates sent by the hypervisor to this 5566 LPAR. 5567 5568 torture.disable_onoff_at_boot= [KNL] 5569 Prevent the CPU-hotplug component of torturing 5570 until after init has spawned. 5571 5572 torture.ftrace_dump_at_shutdown= [KNL] 5573 Dump the ftrace buffer at torture-test shutdown, 5574 even if there were no errors. This can be a 5575 very costly operation when many torture tests 5576 are running concurrently, especially on systems 5577 with rotating-rust storage. 5578 5579 torture.verbose_sleep_frequency= [KNL] 5580 Specifies how many verbose printk()s should be 5581 emitted between each sleep. The default of zero 5582 disables verbose-printk() sleeping. 5583 5584 torture.verbose_sleep_duration= [KNL] 5585 Duration of each verbose-printk() sleep in jiffies. 5586 5587 tp720= [HW,PS2] 5588 5589 tpm_suspend_pcr=[HW,TPM] 5590 Format: integer pcr id 5591 Specify that at suspend time, the tpm driver 5592 should extend the specified pcr with zeros, 5593 as a workaround for some chips which fail to 5594 flush the last written pcr on TPM_SaveState. 5595 This will guarantee that all the other pcrs 5596 are saved. 5597 5598 trace_buf_size=nn[KMG] 5599 [FTRACE] will set tracing buffer size on each cpu. 5600 5601 trace_event=[event-list] 5602 [FTRACE] Set and start specified trace events in order 5603 to facilitate early boot debugging. The event-list is a 5604 comma-separated list of trace events to enable. See 5605 also Documentation/trace/events.rst 5606 5607 trace_options=[option-list] 5608 [FTRACE] Enable or disable tracer options at boot. 5609 The option-list is a comma delimited list of options 5610 that can be enabled or disabled just as if you were 5611 to echo the option name into 5612 5613 /sys/kernel/debug/tracing/trace_options 5614 5615 For example, to enable stacktrace option (to dump the 5616 stack trace of each event), add to the command line: 5617 5618 trace_options=stacktrace 5619 5620 See also Documentation/trace/ftrace.rst "trace options" 5621 section. 5622 5623 tp_printk[FTRACE] 5624 Have the tracepoints sent to printk as well as the 5625 tracing ring buffer. This is useful for early boot up 5626 where the system hangs or reboots and does not give the 5627 option for reading the tracing buffer or performing a 5628 ftrace_dump_on_oops. 5629 5630 To turn off having tracepoints sent to printk, 5631 echo 0 > /proc/sys/kernel/tracepoint_printk 5632 Note, echoing 1 into this file without the 5633 tracepoint_printk kernel cmdline option has no effect. 5634 5635 ** CAUTION ** 5636 5637 Having tracepoints sent to printk() and activating high 5638 frequency tracepoints such as irq or sched, can cause 5639 the system to live lock. 5640 5641 traceoff_on_warning 5642 [FTRACE] enable this option to disable tracing when a 5643 warning is hit. This turns off "tracing_on". Tracing can 5644 be enabled again by echoing '1' into the "tracing_on" 5645 file located in /sys/kernel/debug/tracing/ 5646 5647 This option is useful, as it disables the trace before 5648 the WARNING dump is called, which prevents the trace to 5649 be filled with content caused by the warning output. 5650 5651 This option can also be set at run time via the sysctl 5652 option: kernel/traceoff_on_warning 5653 5654 transparent_hugepage= 5655 [KNL] 5656 Format: [always|madvise|never] 5657 Can be used to control the default behavior of the system 5658 with respect to transparent hugepages. 5659 See Documentation/admin-guide/mm/transhuge.rst 5660 for more details. 5661 5662 trusted.source= [KEYS] 5663 Format: <string> 5664 This parameter identifies the trust source as a backend 5665 for trusted keys implementation. Supported trust 5666 sources: 5667 - "tpm" 5668 - "tee" 5669 If not specified then it defaults to iterating through 5670 the trust source list starting with TPM and assigns the 5671 first trust source as a backend which is initialized 5672 successfully during iteration. 5673 5674 tsc= Disable clocksource stability checks for TSC. 5675 Format: <string> 5676 [x86] reliable: mark tsc clocksource as reliable, this 5677 disables clocksource verification at runtime, as well 5678 as the stability checks done at bootup. Used to enable 5679 high-resolution timer mode on older hardware, and in 5680 virtualized environment. 5681 [x86] noirqtime: Do not use TSC to do irq accounting. 5682 Used to run time disable IRQ_TIME_ACCOUNTING on any 5683 platforms where RDTSC is slow and this accounting 5684 can add overhead. 5685 [x86] unstable: mark the TSC clocksource as unstable, this 5686 marks the TSC unconditionally unstable at bootup and 5687 avoids any further wobbles once the TSC watchdog notices. 5688 [x86] nowatchdog: disable clocksource watchdog. Used 5689 in situations with strict latency requirements (where 5690 interruptions from clocksource watchdog are not 5691 acceptable). 5692 5693 tsc_early_khz= [X86] Skip early TSC calibration and use the given 5694 value instead. Useful when the early TSC frequency discovery 5695 procedure is not reliable, such as on overclocked systems 5696 with CPUID.16h support and partial CPUID.15h support. 5697 Format: <unsigned int> 5698 5699 tsx= [X86] Control Transactional Synchronization 5700 Extensions (TSX) feature in Intel processors that 5701 support TSX control. 5702 5703 This parameter controls the TSX feature. The options are: 5704 5705 on - Enable TSX on the system. Although there are 5706 mitigations for all known security vulnerabilities, 5707 TSX has been known to be an accelerator for 5708 several previous speculation-related CVEs, and 5709 so there may be unknown security risks associated 5710 with leaving it enabled. 5711 5712 off - Disable TSX on the system. (Note that this 5713 option takes effect only on newer CPUs which are 5714 not vulnerable to MDS, i.e., have 5715 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 5716 the new IA32_TSX_CTRL MSR through a microcode 5717 update. This new MSR allows for the reliable 5718 deactivation of the TSX functionality.) 5719 5720 auto - Disable TSX if X86_BUG_TAA is present, 5721 otherwise enable TSX on the system. 5722 5723 Not specifying this option is equivalent to tsx=off. 5724 5725 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5726 for more details. 5727 5728 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 5729 Abort (TAA) vulnerability. 5730 5731 Similar to Micro-architectural Data Sampling (MDS) 5732 certain CPUs that support Transactional 5733 Synchronization Extensions (TSX) are vulnerable to an 5734 exploit against CPU internal buffers which can forward 5735 information to a disclosure gadget under certain 5736 conditions. 5737 5738 In vulnerable processors, the speculatively forwarded 5739 data can be used in a cache side channel attack, to 5740 access data to which the attacker does not have direct 5741 access. 5742 5743 This parameter controls the TAA mitigation. The 5744 options are: 5745 5746 full - Enable TAA mitigation on vulnerable CPUs 5747 if TSX is enabled. 5748 5749 full,nosmt - Enable TAA mitigation and disable SMT on 5750 vulnerable CPUs. If TSX is disabled, SMT 5751 is not disabled because CPU is not 5752 vulnerable to cross-thread TAA attacks. 5753 off - Unconditionally disable TAA mitigation 5754 5755 On MDS-affected machines, tsx_async_abort=off can be 5756 prevented by an active MDS mitigation as both vulnerabilities 5757 are mitigated with the same mechanism so in order to disable 5758 this mitigation, you need to specify mds=off too. 5759 5760 Not specifying this option is equivalent to 5761 tsx_async_abort=full. On CPUs which are MDS affected 5762 and deploy MDS mitigation, TAA mitigation is not 5763 required and doesn't provide any additional 5764 mitigation. 5765 5766 For details see: 5767 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5768 5769 turbografx.map[2|3]= [HW,JOY] 5770 TurboGraFX parallel port interface 5771 Format: 5772 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 5773 See also Documentation/input/devices/joystick-parport.rst 5774 5775 udbg-immortal [PPC] When debugging early kernel crashes that 5776 happen after console_init() and before a proper 5777 console driver takes over, this boot options might 5778 help "seeing" what's going on. 5779 5780 uhash_entries= [KNL,NET] 5781 Set number of hash buckets for UDP/UDP-Lite connections 5782 5783 uhci-hcd.ignore_oc= 5784 [USB] Ignore overcurrent events (default N). 5785 Some badly-designed motherboards generate lots of 5786 bogus events, for ports that aren't wired to 5787 anything. Set this parameter to avoid log spamming. 5788 Note that genuine overcurrent events won't be 5789 reported either. 5790 5791 unknown_nmi_panic 5792 [X86] Cause panic on unknown NMI. 5793 5794 usbcore.authorized_default= 5795 [USB] Default USB device authorization: 5796 (default -1 = authorized except for wireless USB, 5797 0 = not authorized, 1 = authorized, 2 = authorized 5798 if device connected to internal port) 5799 5800 usbcore.autosuspend= 5801 [USB] The autosuspend time delay (in seconds) used 5802 for newly-detected USB devices (default 2). This 5803 is the time required before an idle device will be 5804 autosuspended. Devices for which the delay is set 5805 to a negative value won't be autosuspended at all. 5806 5807 usbcore.usbfs_snoop= 5808 [USB] Set to log all usbfs traffic (default 0 = off). 5809 5810 usbcore.usbfs_snoop_max= 5811 [USB] Maximum number of bytes to snoop in each URB 5812 (default = 65536). 5813 5814 usbcore.blinkenlights= 5815 [USB] Set to cycle leds on hubs (default 0 = off). 5816 5817 usbcore.old_scheme_first= 5818 [USB] Start with the old device initialization 5819 scheme (default 0 = off). 5820 5821 usbcore.usbfs_memory_mb= 5822 [USB] Memory limit (in MB) for buffers allocated by 5823 usbfs (default = 16, 0 = max = 2047). 5824 5825 usbcore.use_both_schemes= 5826 [USB] Try the other device initialization scheme 5827 if the first one fails (default 1 = enabled). 5828 5829 usbcore.initial_descriptor_timeout= 5830 [USB] Specifies timeout for the initial 64-byte 5831 USB_REQ_GET_DESCRIPTOR request in milliseconds 5832 (default 5000 = 5.0 seconds). 5833 5834 usbcore.nousb [USB] Disable the USB subsystem 5835 5836 usbcore.quirks= 5837 [USB] A list of quirk entries to augment the built-in 5838 usb core quirk list. List entries are separated by 5839 commas. Each entry has the form 5840 VendorID:ProductID:Flags. The IDs are 4-digit hex 5841 numbers and Flags is a set of letters. Each letter 5842 will change the built-in quirk; setting it if it is 5843 clear and clearing it if it is set. The letters have 5844 the following meanings: 5845 a = USB_QUIRK_STRING_FETCH_255 (string 5846 descriptors must not be fetched using 5847 a 255-byte read); 5848 b = USB_QUIRK_RESET_RESUME (device can't resume 5849 correctly so reset it instead); 5850 c = USB_QUIRK_NO_SET_INTF (device can't handle 5851 Set-Interface requests); 5852 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5853 handle its Configuration or Interface 5854 strings); 5855 e = USB_QUIRK_RESET (device can't be reset 5856 (e.g morph devices), don't use reset); 5857 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5858 more interface descriptions than the 5859 bNumInterfaces count, and can't handle 5860 talking to these interfaces); 5861 g = USB_QUIRK_DELAY_INIT (device needs a pause 5862 during initialization, after we read 5863 the device descriptor); 5864 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5865 high speed and super speed interrupt 5866 endpoints, the USB 2.0 and USB 3.0 spec 5867 require the interval in microframes (1 5868 microframe = 125 microseconds) to be 5869 calculated as interval = 2 ^ 5870 (bInterval-1). 5871 Devices with this quirk report their 5872 bInterval as the result of this 5873 calculation instead of the exponent 5874 variable used in the calculation); 5875 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5876 handle device_qualifier descriptor 5877 requests); 5878 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5879 generates spurious wakeup, ignore 5880 remote wakeup capability); 5881 k = USB_QUIRK_NO_LPM (device can't handle Link 5882 Power Management); 5883 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5884 (Device reports its bInterval as linear 5885 frames instead of the USB 2.0 5886 calculation); 5887 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5888 to be disconnected before suspend to 5889 prevent spurious wakeup); 5890 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5891 pause after every control message); 5892 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5893 delay after resetting its port); 5894 Example: quirks=0781:5580:bk,0a5c:5834:gij 5895 5896 usbhid.mousepoll= 5897 [USBHID] The interval which mice are to be polled at. 5898 5899 usbhid.jspoll= 5900 [USBHID] The interval which joysticks are to be polled at. 5901 5902 usbhid.kbpoll= 5903 [USBHID] The interval which keyboards are to be polled at. 5904 5905 usb-storage.delay_use= 5906 [UMS] The delay in seconds before a new device is 5907 scanned for Logical Units (default 1). 5908 5909 usb-storage.quirks= 5910 [UMS] A list of quirks entries to supplement or 5911 override the built-in unusual_devs list. List 5912 entries are separated by commas. Each entry has 5913 the form VID:PID:Flags where VID and PID are Vendor 5914 and Product ID values (4-digit hex numbers) and 5915 Flags is a set of characters, each corresponding 5916 to a common usb-storage quirk flag as follows: 5917 a = SANE_SENSE (collect more than 18 bytes 5918 of sense data, not on uas); 5919 b = BAD_SENSE (don't collect more than 18 5920 bytes of sense data, not on uas); 5921 c = FIX_CAPACITY (decrease the reported 5922 device capacity by one sector); 5923 d = NO_READ_DISC_INFO (don't use 5924 READ_DISC_INFO command, not on uas); 5925 e = NO_READ_CAPACITY_16 (don't use 5926 READ_CAPACITY_16 command); 5927 f = NO_REPORT_OPCODES (don't use report opcodes 5928 command, uas only); 5929 g = MAX_SECTORS_240 (don't transfer more than 5930 240 sectors at a time, uas only); 5931 h = CAPACITY_HEURISTICS (decrease the 5932 reported device capacity by one 5933 sector if the number is odd); 5934 i = IGNORE_DEVICE (don't bind to this 5935 device); 5936 j = NO_REPORT_LUNS (don't use report luns 5937 command, uas only); 5938 k = NO_SAME (do not use WRITE_SAME, uas only) 5939 l = NOT_LOCKABLE (don't try to lock and 5940 unlock ejectable media, not on uas); 5941 m = MAX_SECTORS_64 (don't transfer more 5942 than 64 sectors = 32 KB at a time, 5943 not on uas); 5944 n = INITIAL_READ10 (force a retry of the 5945 initial READ(10) command, not on uas); 5946 o = CAPACITY_OK (accept the capacity 5947 reported by the device, not on uas); 5948 p = WRITE_CACHE (the device cache is ON 5949 by default, not on uas); 5950 r = IGNORE_RESIDUE (the device reports 5951 bogus residue values, not on uas); 5952 s = SINGLE_LUN (the device has only one 5953 Logical Unit); 5954 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5955 commands, uas only); 5956 u = IGNORE_UAS (don't bind to the uas driver); 5957 w = NO_WP_DETECT (don't test whether the 5958 medium is write-protected). 5959 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5960 even if the device claims no cache, 5961 not on uas) 5962 Example: quirks=0419:aaf5:rl,0421:0433:rc 5963 5964 user_debug= [KNL,ARM] 5965 Format: <int> 5966 See arch/arm/Kconfig.debug help text. 5967 1 - undefined instruction events 5968 2 - system calls 5969 4 - invalid data aborts 5970 8 - SIGSEGV faults 5971 16 - SIGBUS faults 5972 Example: user_debug=31 5973 5974 userpte= 5975 [X86] Flags controlling user PTE allocations. 5976 5977 nohigh = do not allocate PTE pages in 5978 HIGHMEM regardless of setting 5979 of CONFIG_HIGHPTE. 5980 5981 vdso= [X86,SH] 5982 On X86_32, this is an alias for vdso32=. Otherwise: 5983 5984 vdso=1: enable VDSO (the default) 5985 vdso=0: disable VDSO mapping 5986 5987 vdso32= [X86] Control the 32-bit vDSO 5988 vdso32=1: enable 32-bit VDSO 5989 vdso32=0 or vdso32=2: disable 32-bit VDSO 5990 5991 See the help text for CONFIG_COMPAT_VDSO for more 5992 details. If CONFIG_COMPAT_VDSO is set, the default is 5993 vdso32=0; otherwise, the default is vdso32=1. 5994 5995 For compatibility with older kernels, vdso32=2 is an 5996 alias for vdso32=0. 5997 5998 Try vdso32=0 if you encounter an error that says: 5999 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 6000 6001 vector= [IA-64,SMP] 6002 vector=percpu: enable percpu vector domain 6003 6004 video= [FB] Frame buffer configuration 6005 See Documentation/fb/modedb.rst. 6006 6007 video.brightness_switch_enabled= [0,1] 6008 If set to 1, on receiving an ACPI notify event 6009 generated by hotkey, video driver will adjust brightness 6010 level and then send out the event to user space through 6011 the allocated input device; If set to 0, video driver 6012 will only send out the event without touching backlight 6013 brightness level. 6014 default: 1 6015 6016 virtio_mmio.device= 6017 [VMMIO] Memory mapped virtio (platform) device. 6018 6019 <size>@<baseaddr>:<irq>[:<id>] 6020 where: 6021 <size> := size (can use standard suffixes 6022 like K, M and G) 6023 <baseaddr> := physical base address 6024 <irq> := interrupt number (as passed to 6025 request_irq()) 6026 <id> := (optional) platform device id 6027 example: 6028 virtio_mmio.device=1K@0x100b0000:48:7 6029 6030 Can be used multiple times for multiple devices. 6031 6032 vga= [BOOT,X86-32] Select a particular video mode 6033 See Documentation/x86/boot.rst and 6034 Documentation/admin-guide/svga.rst. 6035 Use vga=ask for menu. 6036 This is actually a boot loader parameter; the value is 6037 passed to the kernel using a special protocol. 6038 6039 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 6040 May slow down system boot speed, especially when 6041 enabled on systems with a large amount of memory. 6042 All options are enabled by default, and this 6043 interface is meant to allow for selectively 6044 enabling or disabling specific virtual memory 6045 debugging features. 6046 6047 Available options are: 6048 P Enable page structure init time poisoning 6049 - Disable all of the above options 6050 6051 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 6052 size of <nn>. This can be used to increase the 6053 minimum size (128MB on x86). It can also be used to 6054 decrease the size and leave more room for directly 6055 mapped kernel RAM. 6056 6057 vmcp_cma=nn[MG] [KNL,S390] 6058 Sets the memory size reserved for contiguous memory 6059 allocations for the vmcp device driver. 6060 6061 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 6062 Format: <command> 6063 6064 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 6065 Format: <command> 6066 6067 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 6068 Format: <command> 6069 6070 vsyscall= [X86-64] 6071 Controls the behavior of vsyscalls (i.e. calls to 6072 fixed addresses of 0xffffffffff600x00 from legacy 6073 code). Most statically-linked binaries and older 6074 versions of glibc use these calls. Because these 6075 functions are at fixed addresses, they make nice 6076 targets for exploits that can control RIP. 6077 6078 emulate [default] Vsyscalls turn into traps and are 6079 emulated reasonably safely. The vsyscall 6080 page is readable. 6081 6082 xonly Vsyscalls turn into traps and are 6083 emulated reasonably safely. The vsyscall 6084 page is not readable. 6085 6086 none Vsyscalls don't work at all. This makes 6087 them quite hard to use for exploits but 6088 might break your system. 6089 6090 vt.color= [VT] Default text color. 6091 Format: 0xYX, X = foreground, Y = background. 6092 Default: 0x07 = light gray on black. 6093 6094 vt.cur_default= [VT] Default cursor shape. 6095 Format: 0xCCBBAA, where AA, BB, and CC are the same as 6096 the parameters of the <Esc>[?A;B;Cc escape sequence; 6097 see VGA-softcursor.txt. Default: 2 = underline. 6098 6099 vt.default_blu= [VT] 6100 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 6101 Change the default blue palette of the console. 6102 This is a 16-member array composed of values 6103 ranging from 0-255. 6104 6105 vt.default_grn= [VT] 6106 Format: <green0>,<green1>,<green2>,...,<green15> 6107 Change the default green palette of the console. 6108 This is a 16-member array composed of values 6109 ranging from 0-255. 6110 6111 vt.default_red= [VT] 6112 Format: <red0>,<red1>,<red2>,...,<red15> 6113 Change the default red palette of the console. 6114 This is a 16-member array composed of values 6115 ranging from 0-255. 6116 6117 vt.default_utf8= 6118 [VT] 6119 Format=<0|1> 6120 Set system-wide default UTF-8 mode for all tty's. 6121 Default is 1, i.e. UTF-8 mode is enabled for all 6122 newly opened terminals. 6123 6124 vt.global_cursor_default= 6125 [VT] 6126 Format=<-1|0|1> 6127 Set system-wide default for whether a cursor 6128 is shown on new VTs. Default is -1, 6129 i.e. cursors will be created by default unless 6130 overridden by individual drivers. 0 will hide 6131 cursors, 1 will display them. 6132 6133 vt.italic= [VT] Default color for italic text; 0-15. 6134 Default: 2 = green. 6135 6136 vt.underline= [VT] Default color for underlined text; 0-15. 6137 Default: 3 = cyan. 6138 6139 watchdog timers [HW,WDT] For information on watchdog timers, 6140 see Documentation/watchdog/watchdog-parameters.rst 6141 or other driver-specific files in the 6142 Documentation/watchdog/ directory. 6143 6144 watchdog_thresh= 6145 [KNL] 6146 Set the hard lockup detector stall duration 6147 threshold in seconds. The soft lockup detector 6148 threshold is set to twice the value. A value of 0 6149 disables both lockup detectors. Default is 10 6150 seconds. 6151 6152 workqueue.watchdog_thresh= 6153 If CONFIG_WQ_WATCHDOG is configured, workqueue can 6154 warn stall conditions and dump internal state to 6155 help debugging. 0 disables workqueue stall 6156 detection; otherwise, it's the stall threshold 6157 duration in seconds. The default value is 30 and 6158 it can be updated at runtime by writing to the 6159 corresponding sysfs file. 6160 6161 workqueue.disable_numa 6162 By default, all work items queued to unbound 6163 workqueues are affine to the NUMA nodes they're 6164 issued on, which results in better behavior in 6165 general. If NUMA affinity needs to be disabled for 6166 whatever reason, this option can be used. Note 6167 that this also can be controlled per-workqueue for 6168 workqueues visible under /sys/bus/workqueue/. 6169 6170 workqueue.power_efficient 6171 Per-cpu workqueues are generally preferred because 6172 they show better performance thanks to cache 6173 locality; unfortunately, per-cpu workqueues tend to 6174 be more power hungry than unbound workqueues. 6175 6176 Enabling this makes the per-cpu workqueues which 6177 were observed to contribute significantly to power 6178 consumption unbound, leading to measurably lower 6179 power usage at the cost of small performance 6180 overhead. 6181 6182 The default value of this parameter is determined by 6183 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 6184 6185 workqueue.debug_force_rr_cpu 6186 Workqueue used to implicitly guarantee that work 6187 items queued without explicit CPU specified are put 6188 on the local CPU. This guarantee is no longer true 6189 and while local CPU is still preferred work items 6190 may be put on foreign CPUs. This debug option 6191 forces round-robin CPU selection to flush out 6192 usages which depend on the now broken guarantee. 6193 When enabled, memory and cache locality will be 6194 impacted. 6195 6196 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 6197 default x2apic cluster mode on platforms 6198 supporting x2apic. 6199 6200 xen_512gb_limit [KNL,X86-64,XEN] 6201 Restricts the kernel running paravirtualized under Xen 6202 to use only up to 512 GB of RAM. The reason to do so is 6203 crash analysis tools and Xen tools for doing domain 6204 save/restore/migration must be enabled to handle larger 6205 domains. 6206 6207 xen_emul_unplug= [HW,X86,XEN] 6208 Unplug Xen emulated devices 6209 Format: [unplug0,][unplug1] 6210 ide-disks -- unplug primary master IDE devices 6211 aux-ide-disks -- unplug non-primary-master IDE devices 6212 nics -- unplug network devices 6213 all -- unplug all emulated devices (NICs and IDE disks) 6214 unnecessary -- unplugging emulated devices is 6215 unnecessary even if the host did not respond to 6216 the unplug protocol 6217 never -- do not unplug even if version check succeeds 6218 6219 xen_legacy_crash [X86,XEN] 6220 Crash from Xen panic notifier, without executing late 6221 panic() code such as dumping handler. 6222 6223 xen_nopvspin [X86,XEN] 6224 Disables the qspinlock slowpath using Xen PV optimizations. 6225 This parameter is obsoleted by "nopvspin" parameter, which 6226 has equivalent effect for XEN platform. 6227 6228 xen_nopv [X86] 6229 Disables the PV optimizations forcing the HVM guest to 6230 run as generic HVM guest with no PV drivers. 6231 This option is obsoleted by the "nopv" option, which 6232 has equivalent effect for XEN platform. 6233 6234 xen_no_vector_callback 6235 [KNL,X86,XEN] Disable the vector callback for Xen 6236 event channel interrupts. 6237 6238 xen_scrub_pages= [XEN] 6239 Boolean option to control scrubbing pages before giving them back 6240 to Xen, for use by other domains. Can be also changed at runtime 6241 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 6242 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 6243 6244 xen_timer_slop= [X86-64,XEN] 6245 Set the timer slop (in nanoseconds) for the virtual Xen 6246 timers (default is 100000). This adjusts the minimum 6247 delta of virtualized Xen timers, where lower values 6248 improve timer resolution at the expense of processing 6249 more timer interrupts. 6250 6251 xen.event_eoi_delay= [XEN] 6252 How long to delay EOI handling in case of event 6253 storms (jiffies). Default is 10. 6254 6255 xen.event_loop_timeout= [XEN] 6256 After which time (jiffies) the event handling loop 6257 should start to delay EOI handling. Default is 2. 6258 6259 xen.fifo_events= [XEN] 6260 Boolean parameter to disable using fifo event handling 6261 even if available. Normally fifo event handling is 6262 preferred over the 2-level event handling, as it is 6263 fairer and the number of possible event channels is 6264 much higher. Default is on (use fifo events). 6265 6266 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 6267 Disables the PV optimizations forcing the guest to run 6268 as generic guest with no PV drivers. Currently support 6269 XEN HVM, KVM, HYPER_V and VMWARE guest. 6270 6271 nopvspin [X86,XEN,KVM] 6272 Disables the qspinlock slow path using PV optimizations 6273 which allow the hypervisor to 'idle' the guest on lock 6274 contention. 6275 6276 xirc2ps_cs= [NET,PCMCIA] 6277 Format: 6278 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 6279 6280 xive= [PPC] 6281 By default on POWER9 and above, the kernel will 6282 natively use the XIVE interrupt controller. This option 6283 allows the fallback firmware mode to be used: 6284 6285 off Fallback to firmware control of XIVE interrupt 6286 controller on both pseries and powernv 6287 platforms. Only useful on POWER9 and above. 6288 6289 xhci-hcd.quirks [USB,KNL] 6290 A hex value specifying bitmask with supplemental xhci 6291 host controller quirks. Meaning of each bit can be 6292 consulted in header drivers/usb/host/xhci.h. 6293 6294 xmon [PPC] 6295 Format: { early | on | rw | ro | off } 6296 Controls if xmon debugger is enabled. Default is off. 6297 Passing only "xmon" is equivalent to "xmon=early". 6298 early Call xmon as early as possible on boot; xmon 6299 debugger is called from setup_arch(). 6300 on xmon debugger hooks will be installed so xmon 6301 is only called on a kernel crash. Default mode, 6302 i.e. either "ro" or "rw" mode, is controlled 6303 with CONFIG_XMON_DEFAULT_RO_MODE. 6304 rw xmon debugger hooks will be installed so xmon 6305 is called only on a kernel crash, mode is write, 6306 meaning SPR registers, memory and, other data 6307 can be written using xmon commands. 6308 ro same as "rw" option above but SPR registers, 6309 memory, and other data can't be written using 6310 xmon commands. 6311 off xmon is disabled. 6312