1 acpi= [HW,ACPI,X86,ARM64] 2 Advanced Configuration and Power Interface 3 Format: { force | on | off | strict | noirq | rsdt | 4 copy_dsdt } 5 force -- enable ACPI if default was off 6 on -- enable ACPI but allow fallback to DT [arm64] 7 off -- disable ACPI if default was on 8 noirq -- do not use ACPI for IRQ routing 9 strict -- Be less tolerant of platforms that are not 10 strictly ACPI specification compliant. 11 rsdt -- prefer RSDT over (default) XSDT 12 copy_dsdt -- copy DSDT to memory 13 For ARM64, ONLY "acpi=off", "acpi=on" or "acpi=force" 14 are available 15 16 See also Documentation/power/runtime_pm.rst, pci=noacpi 17 18 acpi_apic_instance= [ACPI, IOAPIC] 19 Format: <int> 20 2: use 2nd APIC table, if available 21 1,0: use 1st APIC table 22 default: 0 23 24 acpi_backlight= [HW,ACPI] 25 { vendor | video | native | none } 26 If set to vendor, prefer vendor-specific driver 27 (e.g. thinkpad_acpi, sony_acpi, etc.) instead 28 of the ACPI video.ko driver. 29 If set to video, use the ACPI video.ko driver. 30 If set to native, use the device's native backlight mode. 31 If set to none, disable the ACPI backlight interface. 32 33 acpi_force_32bit_fadt_addr 34 force FADT to use 32 bit addresses rather than the 35 64 bit X_* addresses. Some firmware have broken 64 36 bit addresses for force ACPI ignore these and use 37 the older legacy 32 bit addresses. 38 39 acpica_no_return_repair [HW, ACPI] 40 Disable AML predefined validation mechanism 41 This mechanism can repair the evaluation result to make 42 the return objects more ACPI specification compliant. 43 This option is useful for developers to identify the 44 root cause of an AML interpreter issue when the issue 45 has something to do with the repair mechanism. 46 47 acpi.debug_layer= [HW,ACPI,ACPI_DEBUG] 48 acpi.debug_level= [HW,ACPI,ACPI_DEBUG] 49 Format: <int> 50 CONFIG_ACPI_DEBUG must be enabled to produce any ACPI 51 debug output. Bits in debug_layer correspond to a 52 _COMPONENT in an ACPI source file, e.g., 53 #define _COMPONENT ACPI_PCI_COMPONENT 54 Bits in debug_level correspond to a level in 55 ACPI_DEBUG_PRINT statements, e.g., 56 ACPI_DEBUG_PRINT((ACPI_DB_INFO, ... 57 The debug_level mask defaults to "info". See 58 Documentation/firmware-guide/acpi/debug.rst for more information about 59 debug layers and levels. 60 61 Enable processor driver info messages: 62 acpi.debug_layer=0x20000000 63 Enable PCI/PCI interrupt routing info messages: 64 acpi.debug_layer=0x400000 65 Enable AML "Debug" output, i.e., stores to the Debug 66 object while interpreting AML: 67 acpi.debug_layer=0xffffffff acpi.debug_level=0x2 68 Enable all messages related to ACPI hardware: 69 acpi.debug_layer=0x2 acpi.debug_level=0xffffffff 70 71 Some values produce so much output that the system is 72 unusable. The "log_buf_len" parameter may be useful 73 if you need to capture more output. 74 75 acpi_enforce_resources= [ACPI] 76 { strict | lax | no } 77 Check for resource conflicts between native drivers 78 and ACPI OperationRegions (SystemIO and SystemMemory 79 only). IO ports and memory declared in ACPI might be 80 used by the ACPI subsystem in arbitrary AML code and 81 can interfere with legacy drivers. 82 strict (default): access to resources claimed by ACPI 83 is denied; legacy drivers trying to access reserved 84 resources will fail to bind to device using them. 85 lax: access to resources claimed by ACPI is allowed; 86 legacy drivers trying to access reserved resources 87 will bind successfully but a warning message is logged. 88 no: ACPI OperationRegions are not marked as reserved, 89 no further checks are performed. 90 91 acpi_force_table_verification [HW,ACPI] 92 Enable table checksum verification during early stage. 93 By default, this is disabled due to x86 early mapping 94 size limitation. 95 96 acpi_irq_balance [HW,ACPI] 97 ACPI will balance active IRQs 98 default in APIC mode 99 100 acpi_irq_nobalance [HW,ACPI] 101 ACPI will not move active IRQs (default) 102 default in PIC mode 103 104 acpi_irq_isa= [HW,ACPI] If irq_balance, mark listed IRQs used by ISA 105 Format: <irq>,<irq>... 106 107 acpi_irq_pci= [HW,ACPI] If irq_balance, clear listed IRQs for 108 use by PCI 109 Format: <irq>,<irq>... 110 111 acpi_mask_gpe= [HW,ACPI] 112 Due to the existence of _Lxx/_Exx, some GPEs triggered 113 by unsupported hardware/firmware features can result in 114 GPE floodings that cannot be automatically disabled by 115 the GPE dispatcher. 116 This facility can be used to prevent such uncontrolled 117 GPE floodings. 118 Format: <byte> 119 120 acpi_no_auto_serialize [HW,ACPI] 121 Disable auto-serialization of AML methods 122 AML control methods that contain the opcodes to create 123 named objects will be marked as "Serialized" by the 124 auto-serialization feature. 125 This feature is enabled by default. 126 This option allows to turn off the feature. 127 128 acpi_no_memhotplug [ACPI] Disable memory hotplug. Useful for kdump 129 kernels. 130 131 acpi_no_static_ssdt [HW,ACPI] 132 Disable installation of static SSDTs at early boot time 133 By default, SSDTs contained in the RSDT/XSDT will be 134 installed automatically and they will appear under 135 /sys/firmware/acpi/tables. 136 This option turns off this feature. 137 Note that specifying this option does not affect 138 dynamic table installation which will install SSDT 139 tables to /sys/firmware/acpi/tables/dynamic. 140 141 acpi_no_watchdog [HW,ACPI,WDT] 142 Ignore the ACPI-based watchdog interface (WDAT) and let 143 a native driver control the watchdog device instead. 144 145 acpi_rsdp= [ACPI,EFI,KEXEC] 146 Pass the RSDP address to the kernel, mostly used 147 on machines running EFI runtime service to boot the 148 second kernel for kdump. 149 150 acpi_os_name= [HW,ACPI] Tell ACPI BIOS the name of the OS 151 Format: To spoof as Windows 98: ="Microsoft Windows" 152 153 acpi_rev_override [ACPI] Override the _REV object to return 5 (instead 154 of 2 which is mandated by ACPI 6) as the supported ACPI 155 specification revision (when using this switch, it may 156 be necessary to carry out a cold reboot _twice_ in a 157 row to make it take effect on the platform firmware). 158 159 acpi_osi= [HW,ACPI] Modify list of supported OS interface strings 160 acpi_osi="string1" # add string1 161 acpi_osi="!string2" # remove string2 162 acpi_osi=!* # remove all strings 163 acpi_osi=! # disable all built-in OS vendor 164 strings 165 acpi_osi=!! # enable all built-in OS vendor 166 strings 167 acpi_osi= # disable all strings 168 169 'acpi_osi=!' can be used in combination with single or 170 multiple 'acpi_osi="string1"' to support specific OS 171 vendor string(s). Note that such command can only 172 affect the default state of the OS vendor strings, thus 173 it cannot affect the default state of the feature group 174 strings and the current state of the OS vendor strings, 175 specifying it multiple times through kernel command line 176 is meaningless. This command is useful when one do not 177 care about the state of the feature group strings which 178 should be controlled by the OSPM. 179 Examples: 180 1. 'acpi_osi=! acpi_osi="Windows 2000"' is equivalent 181 to 'acpi_osi="Windows 2000" acpi_osi=!', they all 182 can make '_OSI("Windows 2000")' TRUE. 183 184 'acpi_osi=' cannot be used in combination with other 185 'acpi_osi=' command lines, the _OSI method will not 186 exist in the ACPI namespace. NOTE that such command can 187 only affect the _OSI support state, thus specifying it 188 multiple times through kernel command line is also 189 meaningless. 190 Examples: 191 1. 'acpi_osi=' can make 'CondRefOf(_OSI, Local1)' 192 FALSE. 193 194 'acpi_osi=!*' can be used in combination with single or 195 multiple 'acpi_osi="string1"' to support specific 196 string(s). Note that such command can affect the 197 current state of both the OS vendor strings and the 198 feature group strings, thus specifying it multiple times 199 through kernel command line is meaningful. But it may 200 still not able to affect the final state of a string if 201 there are quirks related to this string. This command 202 is useful when one want to control the state of the 203 feature group strings to debug BIOS issues related to 204 the OSPM features. 205 Examples: 206 1. 'acpi_osi="Module Device" acpi_osi=!*' can make 207 '_OSI("Module Device")' FALSE. 208 2. 'acpi_osi=!* acpi_osi="Module Device"' can make 209 '_OSI("Module Device")' TRUE. 210 3. 'acpi_osi=! acpi_osi=!* acpi_osi="Windows 2000"' is 211 equivalent to 212 'acpi_osi=!* acpi_osi=! acpi_osi="Windows 2000"' 213 and 214 'acpi_osi=!* acpi_osi="Windows 2000" acpi_osi=!', 215 they all will make '_OSI("Windows 2000")' TRUE. 216 217 acpi_pm_good [X86] 218 Override the pmtimer bug detection: force the kernel 219 to assume that this machine's pmtimer latches its value 220 and always returns good values. 221 222 acpi_sci= [HW,ACPI] ACPI System Control Interrupt trigger mode 223 Format: { level | edge | high | low } 224 225 acpi_skip_timer_override [HW,ACPI] 226 Recognize and ignore IRQ0/pin2 Interrupt Override. 227 For broken nForce2 BIOS resulting in XT-PIC timer. 228 229 acpi_sleep= [HW,ACPI] Sleep options 230 Format: { s3_bios, s3_mode, s3_beep, s4_nohwsig, 231 old_ordering, nonvs, sci_force_enable, nobl } 232 See Documentation/power/video.rst for information on 233 s3_bios and s3_mode. 234 s3_beep is for debugging; it makes the PC's speaker beep 235 as soon as the kernel's real-mode entry point is called. 236 s4_nohwsig prevents ACPI hardware signature from being 237 used during resume from hibernation. 238 old_ordering causes the ACPI 1.0 ordering of the _PTS 239 control method, with respect to putting devices into 240 low power states, to be enforced (the ACPI 2.0 ordering 241 of _PTS is used by default). 242 nonvs prevents the kernel from saving/restoring the 243 ACPI NVS memory during suspend/hibernation and resume. 244 sci_force_enable causes the kernel to set SCI_EN directly 245 on resume from S1/S3 (which is against the ACPI spec, 246 but some broken systems don't work without it). 247 nobl causes the internal blacklist of systems known to 248 behave incorrectly in some ways with respect to system 249 suspend and resume to be ignored (use wisely). 250 251 acpi_use_timer_override [HW,ACPI] 252 Use timer override. For some broken Nvidia NF5 boards 253 that require a timer override, but don't have HPET 254 255 add_efi_memmap [EFI; X86] Include EFI memory map in 256 kernel's map of available physical RAM. 257 258 agp= [AGP] 259 { off | try_unsupported } 260 off: disable AGP support 261 try_unsupported: try to drive unsupported chipsets 262 (may crash computer or cause data corruption) 263 264 ALSA [HW,ALSA] 265 See Documentation/sound/alsa-configuration.rst 266 267 alignment= [KNL,ARM] 268 Allow the default userspace alignment fault handler 269 behaviour to be specified. Bit 0 enables warnings, 270 bit 1 enables fixups, and bit 2 sends a segfault. 271 272 align_va_addr= [X86-64] 273 Align virtual addresses by clearing slice [14:12] when 274 allocating a VMA at process creation time. This option 275 gives you up to 3% performance improvement on AMD F15h 276 machines (where it is enabled by default) for a 277 CPU-intensive style benchmark, and it can vary highly in 278 a microbenchmark depending on workload and compiler. 279 280 32: only for 32-bit processes 281 64: only for 64-bit processes 282 on: enable for both 32- and 64-bit processes 283 off: disable for both 32- and 64-bit processes 284 285 alloc_snapshot [FTRACE] 286 Allocate the ftrace snapshot buffer on boot up when the 287 main buffer is allocated. This is handy if debugging 288 and you need to use tracing_snapshot() on boot up, and 289 do not want to use tracing_snapshot_alloc() as it needs 290 to be done where GFP_KERNEL allocations are allowed. 291 292 amd_iommu= [HW,X86-64] 293 Pass parameters to the AMD IOMMU driver in the system. 294 Possible values are: 295 fullflush - enable flushing of IO/TLB entries when 296 they are unmapped. Otherwise they are 297 flushed before they will be reused, which 298 is a lot of faster 299 off - do not initialize any AMD IOMMU found in 300 the system 301 force_isolation - Force device isolation for all 302 devices. The IOMMU driver is not 303 allowed anymore to lift isolation 304 requirements as needed. This option 305 does not override iommu=pt 306 307 amd_iommu_dump= [HW,X86-64] 308 Enable AMD IOMMU driver option to dump the ACPI table 309 for AMD IOMMU. With this option enabled, AMD IOMMU 310 driver will print ACPI tables for AMD IOMMU during 311 IOMMU initialization. 312 313 amd_iommu_intr= [HW,X86-64] 314 Specifies one of the following AMD IOMMU interrupt 315 remapping modes: 316 legacy - Use legacy interrupt remapping mode. 317 vapic - Use virtual APIC mode, which allows IOMMU 318 to inject interrupts directly into guest. 319 This mode requires kvm-amd.avic=1. 320 (Default when IOMMU HW support is present.) 321 322 amijoy.map= [HW,JOY] Amiga joystick support 323 Map of devices attached to JOY0DAT and JOY1DAT 324 Format: <a>,<b> 325 See also Documentation/input/joydev/joystick.rst 326 327 analog.map= [HW,JOY] Analog joystick and gamepad support 328 Specifies type or capabilities of an analog joystick 329 connected to one of 16 gameports 330 Format: <type1>,<type2>,..<type16> 331 332 apc= [HW,SPARC] 333 Power management functions (SPARCstation-4/5 + deriv.) 334 Format: noidle 335 Disable APC CPU standby support. SPARCstation-Fox does 336 not play well with APC CPU idle - disable it if you have 337 APC and your system crashes randomly. 338 339 apic= [APIC,X86] Advanced Programmable Interrupt Controller 340 Change the output verbosity while booting 341 Format: { quiet (default) | verbose | debug } 342 Change the amount of debugging information output 343 when initialising the APIC and IO-APIC components. 344 For X86-32, this can also be used to specify an APIC 345 driver name. 346 Format: apic=driver_name 347 Examples: apic=bigsmp 348 349 apic_extnmi= [APIC,X86] External NMI delivery setting 350 Format: { bsp (default) | all | none } 351 bsp: External NMI is delivered only to CPU 0 352 all: External NMIs are broadcast to all CPUs as a 353 backup of CPU 0 354 none: External NMI is masked for all CPUs. This is 355 useful so that a dump capture kernel won't be 356 shot down by NMI 357 358 autoconf= [IPV6] 359 See Documentation/networking/ipv6.rst. 360 361 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 362 Limit apic dumping. The parameter defines the maximal 363 number of local apics being dumped. Also it is possible 364 to set it to "all" by meaning -- no limit here. 365 Format: { 1 (default) | 2 | ... | all }. 366 The parameter valid if only apic=debug or 367 apic=verbose is specified. 368 Example: apic=debug show_lapic=all 369 370 apm= [APM] Advanced Power Management 371 See header of arch/x86/kernel/apm_32.c. 372 373 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 374 Format: <io>,<irq>,<nodeID> 375 376 ataflop= [HW,M68k] 377 378 atarimouse= [HW,MOUSE] Atari Mouse 379 380 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 381 EzKey and similar keyboards 382 383 atkbd.reset= [HW] Reset keyboard during initialization 384 385 atkbd.set= [HW] Select keyboard code set 386 Format: <int> (2 = AT (default), 3 = PS/2) 387 388 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 389 keyboards 390 391 atkbd.softraw= [HW] Choose between synthetic and real raw mode 392 Format: <bool> (0 = real, 1 = synthetic (default)) 393 394 atkbd.softrepeat= [HW] 395 Use software keyboard repeat 396 397 audit= [KNL] Enable the audit sub-system 398 Format: { "0" | "1" | "off" | "on" } 399 0 | off - kernel audit is disabled and can not be 400 enabled until the next reboot 401 unset - kernel audit is initialized but disabled and 402 will be fully enabled by the userspace auditd. 403 1 | on - kernel audit is initialized and partially 404 enabled, storing at most audit_backlog_limit 405 messages in RAM until it is fully enabled by the 406 userspace auditd. 407 Default: unset 408 409 audit_backlog_limit= [KNL] Set the audit queue size limit. 410 Format: <int> (must be >=0) 411 Default: 64 412 413 bau= [X86_UV] Enable the BAU on SGI UV. The default 414 behavior is to disable the BAU (i.e. bau=0). 415 Format: { "0" | "1" } 416 0 - Disable the BAU. 417 1 - Enable the BAU. 418 unset - Disable the BAU. 419 420 baycom_epp= [HW,AX25] 421 Format: <io>,<mode> 422 423 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 424 Format: <io>,<mode> 425 See header of drivers/net/hamradio/baycom_par.c. 426 427 baycom_ser_fdx= [HW,AX25] 428 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 429 Format: <io>,<irq>,<mode>[,<baud>] 430 See header of drivers/net/hamradio/baycom_ser_fdx.c. 431 432 baycom_ser_hdx= [HW,AX25] 433 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 434 Format: <io>,<irq>,<mode> 435 See header of drivers/net/hamradio/baycom_ser_hdx.c. 436 437 blkdevparts= Manual partition parsing of block device(s) for 438 embedded devices based on command line input. 439 See Documentation/block/cmdline-partition.rst 440 441 boot_delay= Milliseconds to delay each printk during boot. 442 Values larger than 10 seconds (10000) are changed to 443 no delay (0). 444 Format: integer 445 446 bootconfig [KNL] 447 Extended command line options can be added to an initrd 448 and this will cause the kernel to look for it. 449 450 See Documentation/admin-guide/bootconfig.rst 451 452 bert_disable [ACPI] 453 Disable BERT OS support on buggy BIOSes. 454 455 bgrt_disable [ACPI][X86] 456 Disable BGRT to avoid flickering OEM logo. 457 458 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 459 bttv.radio= Most important insmod options are available as 460 kernel args too. 461 bttv.pll= See Documentation/admin-guide/media/bttv.rst 462 bttv.tuner= 463 464 bulk_remove=off [PPC] This parameter disables the use of the pSeries 465 firmware feature for flushing multiple hpte entries 466 at a time. 467 468 c101= [NET] Moxa C101 synchronous serial card 469 470 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 471 Sometimes CPU hardware bugs make them report the cache 472 size incorrectly. The kernel will attempt work arounds 473 to fix known problems, but for some CPUs it is not 474 possible to determine what the correct size should be. 475 This option provides an override for these situations. 476 477 carrier_timeout= 478 [NET] Specifies amount of time (in seconds) that 479 the kernel should wait for a network carrier. By default 480 it waits 120 seconds. 481 482 ca_keys= [KEYS] This parameter identifies a specific key(s) on 483 the system trusted keyring to be used for certificate 484 trust validation. 485 format: { id:<keyid> | builtin } 486 487 cca= [MIPS] Override the kernel pages' cache coherency 488 algorithm. Accepted values range from 0 to 7 489 inclusive. See arch/mips/include/asm/pgtable-bits.h 490 for platform specific values (SB1, Loongson3 and 491 others). 492 493 ccw_timeout_log [S390] 494 See Documentation/s390/common_io.rst for details. 495 496 cgroup_disable= [KNL] Disable a particular controller 497 Format: {name of the controller(s) to disable} 498 The effects of cgroup_disable=foo are: 499 - foo isn't auto-mounted if you mount all cgroups in 500 a single hierarchy 501 - foo isn't visible as an individually mountable 502 subsystem 503 {Currently only "memory" controller deal with this and 504 cut the overhead, others just disable the usage. So 505 only cgroup_disable=memory is actually worthy} 506 507 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 508 Format: { { controller | "all" | "named" } 509 [,{ controller | "all" | "named" }...] } 510 Like cgroup_disable, but only applies to cgroup v1; 511 the blacklisted controllers remain available in cgroup2. 512 "all" blacklists all controllers and "named" disables 513 named mounts. Specifying both "all" and "named" disables 514 all v1 hierarchies. 515 516 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 517 Format: <string> 518 nosocket -- Disable socket memory accounting. 519 nokmem -- Disable kernel memory accounting. 520 521 checkreqprot [SELINUX] Set initial checkreqprot flag value. 522 Format: { "0" | "1" } 523 See security/selinux/Kconfig help text. 524 0 -- check protection applied by kernel (includes 525 any implied execute protection). 526 1 -- check protection requested by application. 527 Default value is set via a kernel config option. 528 Value can be changed at runtime via 529 /sys/fs/selinux/checkreqprot. 530 Setting checkreqprot to 1 is deprecated. 531 532 cio_ignore= [S390] 533 See Documentation/s390/common_io.rst for details. 534 clk_ignore_unused 535 [CLK] 536 Prevents the clock framework from automatically gating 537 clocks that have not been explicitly enabled by a Linux 538 device driver but are enabled in hardware at reset or 539 by the bootloader/firmware. Note that this does not 540 force such clocks to be always-on nor does it reserve 541 those clocks in any way. This parameter is useful for 542 debug and development, but should not be needed on a 543 platform with proper driver support. For more 544 information, see Documentation/driver-api/clk.rst. 545 546 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 547 [Deprecated] 548 Forces specified clocksource (if available) to be used 549 when calculating gettimeofday(). If specified 550 clocksource is not available, it defaults to PIT. 551 Format: { pit | tsc | cyclone | pmtmr } 552 553 clocksource= Override the default clocksource 554 Format: <string> 555 Override the default clocksource and use the clocksource 556 with the name specified. 557 Some clocksource names to choose from, depending on 558 the platform: 559 [all] jiffies (this is the base, fallback clocksource) 560 [ACPI] acpi_pm 561 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 562 pxa_timer,timer3,32k_counter,timer0_1 563 [X86-32] pit,hpet,tsc; 564 scx200_hrt on Geode; cyclone on IBM x440 565 [MIPS] MIPS 566 [PARISC] cr16 567 [S390] tod 568 [SH] SuperH 569 [SPARC64] tick 570 [X86-64] hpet,tsc 571 572 clocksource.arm_arch_timer.evtstrm= 573 [ARM,ARM64] 574 Format: <bool> 575 Enable/disable the eventstream feature of the ARM 576 architected timer so that code using WFE-based polling 577 loops can be debugged more effectively on production 578 systems. 579 580 clearcpuid=BITNUM [X86] 581 Disable CPUID feature X for the kernel. See 582 arch/x86/include/asm/cpufeatures.h for the valid bit 583 numbers. Note the Linux specific bits are not necessarily 584 stable over kernel options, but the vendor specific 585 ones should be. 586 Also note that user programs calling CPUID directly 587 or using the feature without checking anything 588 will still see it. This just prevents it from 589 being used by the kernel or shown in /proc/cpuinfo. 590 Also note the kernel might malfunction if you disable 591 some critical bits. 592 593 cma=nn[MG]@[start[MG][-end[MG]]] 594 [ARM,X86,KNL] 595 Sets the size of kernel global memory area for 596 contiguous memory allocations and optionally the 597 placement constraint by the physical address range of 598 memory allocations. A value of 0 disables CMA 599 altogether. For more information, see 600 include/linux/dma-contiguous.h 601 602 cmo_free_hint= [PPC] Format: { yes | no } 603 Specify whether pages are marked as being inactive 604 when they are freed. This is used in CMO environments 605 to determine OS memory pressure for page stealing by 606 a hypervisor. 607 Default: yes 608 609 coherent_pool=nn[KMG] [ARM,KNL] 610 Sets the size of memory pool for coherent, atomic dma 611 allocations, by default set to 256K. 612 613 com20020= [HW,NET] ARCnet - COM20020 chipset 614 Format: 615 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 616 617 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 618 Format: <io>[,<irq>] 619 620 com90xx= [HW,NET] 621 ARCnet - COM90xx chipset (memory-mapped buffers) 622 Format: <io>[,<irq>[,<memstart>]] 623 624 condev= [HW,S390] console device 625 conmode= 626 627 console= [KNL] Output console device and options. 628 629 tty<n> Use the virtual console device <n>. 630 631 ttyS<n>[,options] 632 ttyUSB0[,options] 633 Use the specified serial port. The options are of 634 the form "bbbbpnf", where "bbbb" is the baud rate, 635 "p" is parity ("n", "o", or "e"), "n" is number of 636 bits, and "f" is flow control ("r" for RTS or 637 omit it). Default is "9600n8". 638 639 See Documentation/admin-guide/serial-console.rst for more 640 information. See 641 Documentation/networking/netconsole.rst for an 642 alternative. 643 644 uart[8250],io,<addr>[,options] 645 uart[8250],mmio,<addr>[,options] 646 uart[8250],mmio16,<addr>[,options] 647 uart[8250],mmio32,<addr>[,options] 648 uart[8250],0x<addr>[,options] 649 Start an early, polled-mode console on the 8250/16550 650 UART at the specified I/O port or MMIO address, 651 switching to the matching ttyS device later. 652 MMIO inter-register address stride is either 8-bit 653 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 654 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 655 to be equivalent to 'mmio'. 'options' are specified in 656 the same format described for ttyS above; if unspecified, 657 the h/w is not re-initialized. 658 659 hvc<n> Use the hypervisor console device <n>. This is for 660 both Xen and PowerPC hypervisors. 661 662 If the device connected to the port is not a TTY but a braille 663 device, prepend "brl," before the device type, for instance 664 console=brl,ttyS0 665 For now, only VisioBraille is supported. 666 667 console_msg_format= 668 [KNL] Change console messages format 669 default 670 By default we print messages on consoles in 671 "[time stamp] text\n" format (time stamp may not be 672 printed, depending on CONFIG_PRINTK_TIME or 673 `printk_time' param). 674 syslog 675 Switch to syslog format: "<%u>[time stamp] text\n" 676 IOW, each message will have a facility and loglevel 677 prefix. The format is similar to one used by syslog() 678 syscall, or to executing "dmesg -S --raw" or to reading 679 from /proc/kmsg. 680 681 consoleblank= [KNL] The console blank (screen saver) timeout in 682 seconds. A value of 0 disables the blank timer. 683 Defaults to 0. 684 685 coredump_filter= 686 [KNL] Change the default value for 687 /proc/<pid>/coredump_filter. 688 See also Documentation/filesystems/proc.rst. 689 690 coresight_cpu_debug.enable 691 [ARM,ARM64] 692 Format: <bool> 693 Enable/disable the CPU sampling based debugging. 694 0: default value, disable debugging 695 1: enable debugging at boot time 696 697 cpuidle.off=1 [CPU_IDLE] 698 disable the cpuidle sub-system 699 700 cpuidle.governor= 701 [CPU_IDLE] Name of the cpuidle governor to use. 702 703 cpufreq.off=1 [CPU_FREQ] 704 disable the cpufreq sub-system 705 706 cpufreq.default_governor= 707 [CPU_FREQ] Name of the default cpufreq governor or 708 policy to use. This governor must be registered in the 709 kernel before the cpufreq driver probes. 710 711 cpu_init_udelay=N 712 [X86] Delay for N microsec between assert and de-assert 713 of APIC INIT to start processors. This delay occurs 714 on every CPU online, such as boot, and resume from suspend. 715 Default: 10000 716 717 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 718 Format: 719 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 720 721 crashkernel=size[KMG][@offset[KMG]] 722 [KNL] Using kexec, Linux can switch to a 'crash kernel' 723 upon panic. This parameter reserves the physical 724 memory region [offset, offset + size] for that kernel 725 image. If '@offset' is omitted, then a suitable offset 726 is selected automatically. 727 [KNL, x86_64] select a region under 4G first, and 728 fall back to reserve region above 4G when '@offset' 729 hasn't been specified. 730 See Documentation/admin-guide/kdump/kdump.rst for further details. 731 732 crashkernel=range1:size1[,range2:size2,...][@offset] 733 [KNL] Same as above, but depends on the memory 734 in the running system. The syntax of range is 735 start-[end] where start and end are both 736 a memory unit (amount[KMG]). See also 737 Documentation/admin-guide/kdump/kdump.rst for an example. 738 739 crashkernel=size[KMG],high 740 [KNL, x86_64] range could be above 4G. Allow kernel 741 to allocate physical memory region from top, so could 742 be above 4G if system have more than 4G ram installed. 743 Otherwise memory region will be allocated below 4G, if 744 available. 745 It will be ignored if crashkernel=X is specified. 746 crashkernel=size[KMG],low 747 [KNL, x86_64] range under 4G. When crashkernel=X,high 748 is passed, kernel could allocate physical memory region 749 above 4G, that cause second kernel crash on system 750 that require some amount of low memory, e.g. swiotlb 751 requires at least 64M+32K low memory, also enough extra 752 low memory is needed to make sure DMA buffers for 32-bit 753 devices won't run out. Kernel would try to allocate at 754 at least 256M below 4G automatically. 755 This one let user to specify own low range under 4G 756 for second kernel instead. 757 0: to disable low allocation. 758 It will be ignored when crashkernel=X,high is not used 759 or memory reserved is below 4G. 760 761 cryptomgr.notests 762 [KNL] Disable crypto self-tests 763 764 cs89x0_dma= [HW,NET] 765 Format: <dma> 766 767 cs89x0_media= [HW,NET] 768 Format: { rj45 | aui | bnc } 769 770 dasd= [HW,NET] 771 See header of drivers/s390/block/dasd_devmap.c. 772 773 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 774 (one device per port) 775 Format: <port#>,<type> 776 See also Documentation/input/devices/joystick-parport.rst 777 778 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 779 time. See 780 Documentation/admin-guide/dynamic-debug-howto.rst for 781 details. Deprecated, see dyndbg. 782 783 debug [KNL] Enable kernel debugging (events log level). 784 785 debug_boot_weak_hash 786 [KNL] Enable printing [hashed] pointers early in the 787 boot sequence. If enabled, we use a weak hash instead 788 of siphash to hash pointers. Use this option if you are 789 seeing instances of '(___ptrval___)') and need to see a 790 value (hashed pointer) instead. Cryptographically 791 insecure, please do not use on production kernels. 792 793 debug_locks_verbose= 794 [KNL] verbose self-tests 795 Format=<0|1> 796 Print debugging info while doing the locking API 797 self-tests. 798 We default to 0 (no extra messages), setting it to 799 1 will print _a lot_ more information - normally 800 only useful to kernel developers. 801 802 debug_objects [KNL] Enable object debugging 803 804 no_debug_objects 805 [KNL] Disable object debugging 806 807 debug_guardpage_minorder= 808 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 809 parameter allows control of the order of pages that will 810 be intentionally kept free (and hence protected) by the 811 buddy allocator. Bigger value increase the probability 812 of catching random memory corruption, but reduce the 813 amount of memory for normal system use. The maximum 814 possible value is MAX_ORDER/2. Setting this parameter 815 to 1 or 2 should be enough to identify most random 816 memory corruption problems caused by bugs in kernel or 817 driver code when a CPU writes to (or reads from) a 818 random memory location. Note that there exists a class 819 of memory corruptions problems caused by buggy H/W or 820 F/W or by drivers badly programing DMA (basically when 821 memory is written at bus level and the CPU MMU is 822 bypassed) which are not detectable by 823 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 824 tracking down these problems. 825 826 debug_pagealloc= 827 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 828 enables the feature at boot time. By default, it is 829 disabled and the system will work mostly the same as a 830 kernel built without CONFIG_DEBUG_PAGEALLOC. 831 Note: to get most of debug_pagealloc error reports, it's 832 useful to also enable the page_owner functionality. 833 on: enable the feature 834 835 debugpat [X86] Enable PAT debugging 836 837 decnet.addr= [HW,NET] 838 Format: <area>[,<node>] 839 See also Documentation/networking/decnet.rst. 840 841 default_hugepagesz= 842 [HW] The size of the default HugeTLB page. This is 843 the size represented by the legacy /proc/ hugepages 844 APIs. In addition, this is the default hugetlb size 845 used for shmget(), mmap() and mounting hugetlbfs 846 filesystems. If not specified, defaults to the 847 architecture's default huge page size. Huge page 848 sizes are architecture dependent. See also 849 Documentation/admin-guide/mm/hugetlbpage.rst. 850 Format: size[KMG] 851 852 deferred_probe_timeout= 853 [KNL] Debugging option to set a timeout in seconds for 854 deferred probe to give up waiting on dependencies to 855 probe. Only specific dependencies (subsystems or 856 drivers) that have opted in will be ignored. A timeout of 0 857 will timeout at the end of initcalls. This option will also 858 dump out devices still on the deferred probe list after 859 retrying. 860 861 dfltcc= [HW,S390] 862 Format: { on | off | def_only | inf_only | always } 863 on: s390 zlib hardware support for compression on 864 level 1 and decompression (default) 865 off: No s390 zlib hardware support 866 def_only: s390 zlib hardware support for deflate 867 only (compression on level 1) 868 inf_only: s390 zlib hardware support for inflate 869 only (decompression) 870 always: Same as 'on' but ignores the selected compression 871 level always using hardware support (used for debugging) 872 873 dhash_entries= [KNL] 874 Set number of hash buckets for dentry cache. 875 876 disable_1tb_segments [PPC] 877 Disables the use of 1TB hash page table segments. This 878 causes the kernel to fall back to 256MB segments which 879 can be useful when debugging issues that require an SLB 880 miss to occur. 881 882 stress_slb [PPC] 883 Limits the number of kernel SLB entries, and flushes 884 them frequently to increase the rate of SLB faults 885 on kernel addresses. 886 887 disable= [IPV6] 888 See Documentation/networking/ipv6.rst. 889 890 hardened_usercopy= 891 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 892 hardening is enabled for this boot. Hardened 893 usercopy checking is used to protect the kernel 894 from reading or writing beyond known memory 895 allocation boundaries as a proactive defense 896 against bounds-checking flaws in the kernel's 897 copy_to_user()/copy_from_user() interface. 898 on Perform hardened usercopy checks (default). 899 off Disable hardened usercopy checks. 900 901 disable_radix [PPC] 902 Disable RADIX MMU mode on POWER9 903 904 disable_tlbie [PPC] 905 Disable TLBIE instruction. Currently does not work 906 with KVM, with HASH MMU, or with coherent accelerators. 907 908 disable_cpu_apicid= [X86,APIC,SMP] 909 Format: <int> 910 The number of initial APIC ID for the 911 corresponding CPU to be disabled at boot, 912 mostly used for the kdump 2nd kernel to 913 disable BSP to wake up multiple CPUs without 914 causing system reset or hang due to sending 915 INIT from AP to BSP. 916 917 perf_v4_pmi= [X86,INTEL] 918 Format: <bool> 919 Disable Intel PMU counter freezing feature. 920 The feature only exists starting from 921 Arch Perfmon v4 (Skylake and newer). 922 923 disable_ddw [PPC/PSERIES] 924 Disable Dynamic DMA Window support. Use this if 925 to workaround buggy firmware. 926 927 disable_ipv6= [IPV6] 928 See Documentation/networking/ipv6.rst. 929 930 disable_mtrr_cleanup [X86] 931 The kernel tries to adjust MTRR layout from continuous 932 to discrete, to make X server driver able to add WB 933 entry later. This parameter disables that. 934 935 disable_mtrr_trim [X86, Intel and AMD only] 936 By default the kernel will trim any uncacheable 937 memory out of your available memory pool based on 938 MTRR settings. This parameter disables that behavior, 939 possibly causing your machine to run very slowly. 940 941 disable_timer_pin_1 [X86] 942 Disable PIN 1 of APIC timer 943 Can be useful to work around chipset bugs. 944 945 dis_ucode_ldr [X86] Disable the microcode loader. 946 947 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 948 this option disables the debugging code at boot. 949 950 dma_debug_entries=<number> 951 This option allows to tune the number of preallocated 952 entries for DMA-API debugging code. One entry is 953 required per DMA-API allocation. Use this if the 954 DMA-API debugging code disables itself because the 955 architectural default is too low. 956 957 dma_debug_driver=<driver_name> 958 With this option the DMA-API debugging driver 959 filter feature can be enabled at boot time. Just 960 pass the driver to filter for as the parameter. 961 The filter can be disabled or changed to another 962 driver later using sysfs. 963 964 driver_async_probe= [KNL] 965 List of driver names to be probed asynchronously. 966 Format: <driver_name1>,<driver_name2>... 967 968 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 969 Broken monitors, graphic adapters, KVMs and EDIDless 970 panels may send no or incorrect EDID data sets. 971 This parameter allows to specify an EDID data sets 972 in the /lib/firmware directory that are used instead. 973 Generic built-in EDID data sets are used, if one of 974 edid/1024x768.bin, edid/1280x1024.bin, 975 edid/1680x1050.bin, or edid/1920x1080.bin is given 976 and no file with the same name exists. Details and 977 instructions how to build your own EDID data are 978 available in Documentation/admin-guide/edid.rst. An EDID 979 data set will only be used for a particular connector, 980 if its name and a colon are prepended to the EDID 981 name. Each connector may use a unique EDID data 982 set by separating the files with a comma. An EDID 983 data set with no connector name will be used for 984 any connectors not explicitly specified. 985 986 dscc4.setup= [NET] 987 988 dt_cpu_ftrs= [PPC] 989 Format: {"off" | "known"} 990 Control how the dt_cpu_ftrs device-tree binding is 991 used for CPU feature discovery and setup (if it 992 exists). 993 off: Do not use it, fall back to legacy cpu table. 994 known: Do not pass through unknown features to guests 995 or userspace, only those that the kernel is aware of. 996 997 dump_apple_properties [X86] 998 Dump name and content of EFI device properties on 999 x86 Macs. Useful for driver authors to determine 1000 what data is available or for reverse-engineering. 1001 1002 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1003 module.dyndbg[="val"] 1004 Enable debug messages at boot time. See 1005 Documentation/admin-guide/dynamic-debug-howto.rst 1006 for details. 1007 1008 nopku [X86] Disable Memory Protection Keys CPU feature found 1009 in some Intel CPUs. 1010 1011 module.async_probe [KNL] 1012 Enable asynchronous probe on this module. 1013 1014 early_ioremap_debug [KNL] 1015 Enable debug messages in early_ioremap support. This 1016 is useful for tracking down temporary early mappings 1017 which are not unmapped. 1018 1019 earlycon= [KNL] Output early console device and options. 1020 1021 When used with no options, the early console is 1022 determined by stdout-path property in device tree's 1023 chosen node or the ACPI SPCR table if supported by 1024 the platform. 1025 1026 cdns,<addr>[,options] 1027 Start an early, polled-mode console on a Cadence 1028 (xuartps) serial port at the specified address. Only 1029 supported option is baud rate. If baud rate is not 1030 specified, the serial port must already be setup and 1031 configured. 1032 1033 uart[8250],io,<addr>[,options] 1034 uart[8250],mmio,<addr>[,options] 1035 uart[8250],mmio32,<addr>[,options] 1036 uart[8250],mmio32be,<addr>[,options] 1037 uart[8250],0x<addr>[,options] 1038 Start an early, polled-mode console on the 8250/16550 1039 UART at the specified I/O port or MMIO address. 1040 MMIO inter-register address stride is either 8-bit 1041 (mmio) or 32-bit (mmio32 or mmio32be). 1042 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1043 to be equivalent to 'mmio'. 'options' are specified 1044 in the same format described for "console=ttyS<n>"; if 1045 unspecified, the h/w is not initialized. 1046 1047 pl011,<addr> 1048 pl011,mmio32,<addr> 1049 Start an early, polled-mode console on a pl011 serial 1050 port at the specified address. The pl011 serial port 1051 must already be setup and configured. Options are not 1052 yet supported. If 'mmio32' is specified, then only 1053 the driver will use only 32-bit accessors to read/write 1054 the device registers. 1055 1056 meson,<addr> 1057 Start an early, polled-mode console on a meson serial 1058 port at the specified address. The serial port must 1059 already be setup and configured. Options are not yet 1060 supported. 1061 1062 msm_serial,<addr> 1063 Start an early, polled-mode console on an msm serial 1064 port at the specified address. The serial port 1065 must already be setup and configured. Options are not 1066 yet supported. 1067 1068 msm_serial_dm,<addr> 1069 Start an early, polled-mode console on an msm serial 1070 dm port at the specified address. The serial port 1071 must already be setup and configured. Options are not 1072 yet supported. 1073 1074 owl,<addr> 1075 Start an early, polled-mode console on a serial port 1076 of an Actions Semi SoC, such as S500 or S900, at the 1077 specified address. The serial port must already be 1078 setup and configured. Options are not yet supported. 1079 1080 rda,<addr> 1081 Start an early, polled-mode console on a serial port 1082 of an RDA Micro SoC, such as RDA8810PL, at the 1083 specified address. The serial port must already be 1084 setup and configured. Options are not yet supported. 1085 1086 sbi 1087 Use RISC-V SBI (Supervisor Binary Interface) for early 1088 console. 1089 1090 smh Use ARM semihosting calls for early console. 1091 1092 s3c2410,<addr> 1093 s3c2412,<addr> 1094 s3c2440,<addr> 1095 s3c6400,<addr> 1096 s5pv210,<addr> 1097 exynos4210,<addr> 1098 Use early console provided by serial driver available 1099 on Samsung SoCs, requires selecting proper type and 1100 a correct base address of the selected UART port. The 1101 serial port must already be setup and configured. 1102 Options are not yet supported. 1103 1104 lantiq,<addr> 1105 Start an early, polled-mode console on a lantiq serial 1106 (lqasc) port at the specified address. The serial port 1107 must already be setup and configured. Options are not 1108 yet supported. 1109 1110 lpuart,<addr> 1111 lpuart32,<addr> 1112 Use early console provided by Freescale LP UART driver 1113 found on Freescale Vybrid and QorIQ LS1021A processors. 1114 A valid base address must be provided, and the serial 1115 port must already be setup and configured. 1116 1117 ec_imx21,<addr> 1118 ec_imx6q,<addr> 1119 Start an early, polled-mode, output-only console on the 1120 Freescale i.MX UART at the specified address. The UART 1121 must already be setup and configured. 1122 1123 ar3700_uart,<addr> 1124 Start an early, polled-mode console on the 1125 Armada 3700 serial port at the specified 1126 address. The serial port must already be setup 1127 and configured. Options are not yet supported. 1128 1129 qcom_geni,<addr> 1130 Start an early, polled-mode console on a Qualcomm 1131 Generic Interface (GENI) based serial port at the 1132 specified address. The serial port must already be 1133 setup and configured. Options are not yet supported. 1134 1135 efifb,[options] 1136 Start an early, unaccelerated console on the EFI 1137 memory mapped framebuffer (if available). On cache 1138 coherent non-x86 systems that use system memory for 1139 the framebuffer, pass the 'ram' option so that it is 1140 mapped with the correct attributes. 1141 1142 linflex,<addr> 1143 Use early console provided by Freescale LINFlexD UART 1144 serial driver for NXP S32V234 SoCs. A valid base 1145 address must be provided, and the serial port must 1146 already be setup and configured. 1147 1148 earlyprintk= [X86,SH,ARM,M68k,S390] 1149 earlyprintk=vga 1150 earlyprintk=sclp 1151 earlyprintk=xen 1152 earlyprintk=serial[,ttySn[,baudrate]] 1153 earlyprintk=serial[,0x...[,baudrate]] 1154 earlyprintk=ttySn[,baudrate] 1155 earlyprintk=dbgp[debugController#] 1156 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1157 earlyprintk=xdbc[xhciController#] 1158 1159 earlyprintk is useful when the kernel crashes before 1160 the normal console is initialized. It is not enabled by 1161 default because it has some cosmetic problems. 1162 1163 Append ",keep" to not disable it when the real console 1164 takes over. 1165 1166 Only one of vga, efi, serial, or usb debug port can 1167 be used at a time. 1168 1169 Currently only ttyS0 and ttyS1 may be specified by 1170 name. Other I/O ports may be explicitly specified 1171 on some architectures (x86 and arm at least) by 1172 replacing ttySn with an I/O port address, like this: 1173 earlyprintk=serial,0x1008,115200 1174 You can find the port for a given device in 1175 /proc/tty/driver/serial: 1176 2: uart:ST16650V2 port:00001008 irq:18 ... 1177 1178 Interaction with the standard serial driver is not 1179 very good. 1180 1181 The VGA and EFI output is eventually overwritten by 1182 the real console. 1183 1184 The xen output can only be used by Xen PV guests. 1185 1186 The sclp output can only be used on s390. 1187 1188 The optional "force" to "pciserial" enables use of a 1189 PCI device even when its classcode is not of the 1190 UART class. 1191 1192 edac_report= [HW,EDAC] Control how to report EDAC event 1193 Format: {"on" | "off" | "force"} 1194 on: enable EDAC to report H/W event. May be overridden 1195 by other higher priority error reporting module. 1196 off: disable H/W event reporting through EDAC. 1197 force: enforce the use of EDAC to report H/W event. 1198 default: on. 1199 1200 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1201 ekgdboc=kbd 1202 1203 This is designed to be used in conjunction with 1204 the boot argument: earlyprintk=vga 1205 1206 This parameter works in place of the kgdboc parameter 1207 but can only be used if the backing tty is available 1208 very early in the boot process. For early debugging 1209 via a serial port see kgdboc_earlycon instead. 1210 1211 edd= [EDD] 1212 Format: {"off" | "on" | "skip[mbr]"} 1213 1214 efi= [EFI] 1215 Format: { "debug", "disable_early_pci_dma", 1216 "nochunk", "noruntime", "nosoftreserve", 1217 "novamap", "no_disable_early_pci_dma", 1218 "old_map" } 1219 debug: enable misc debug output. 1220 disable_early_pci_dma: disable the busmaster bit on all 1221 PCI bridges while in the EFI boot stub. 1222 nochunk: disable reading files in "chunks" in the EFI 1223 boot stub, as chunking can cause problems with some 1224 firmware implementations. 1225 noruntime : disable EFI runtime services support 1226 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1227 attribute may cause the kernel to reserve the 1228 memory range for a memory mapping driver to 1229 claim. Specify efi=nosoftreserve to disable this 1230 reservation and treat the memory by its base type 1231 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1232 novamap: do not call SetVirtualAddressMap(). 1233 no_disable_early_pci_dma: Leave the busmaster bit set 1234 on all PCI bridges while in the EFI boot stub 1235 old_map [X86-64]: switch to the old ioremap-based EFI 1236 runtime services mapping. [Needs CONFIG_X86_UV=y] 1237 1238 efi_no_storage_paranoia [EFI; X86] 1239 Using this parameter you can use more than 50% of 1240 your efi variable storage. Use this parameter only if 1241 you are really sure that your UEFI does sane gc and 1242 fulfills the spec otherwise your board may brick. 1243 1244 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1245 Add arbitrary attribute to specific memory range by 1246 updating original EFI memory map. 1247 Region of memory which aa attribute is added to is 1248 from ss to ss+nn. 1249 1250 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1251 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1252 attribute is added to range 0x100000000-0x180000000 and 1253 0x10a0000000-0x1120000000. 1254 1255 If efi_fake_mem=8G@9G:0x40000 is specified, the 1256 EFI_MEMORY_SP(0x40000) attribute is added to 1257 range 0x240000000-0x43fffffff. 1258 1259 Using this parameter you can do debugging of EFI memmap 1260 related features. For example, you can do debugging of 1261 Address Range Mirroring feature even if your box 1262 doesn't support it, or mark specific memory as 1263 "soft reserved". 1264 1265 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1266 that is to be dynamically loaded by Linux. If there are 1267 multiple variables with the same name but with different 1268 vendor GUIDs, all of them will be loaded. See 1269 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1270 1271 1272 eisa_irq_edge= [PARISC,HW] 1273 See header of drivers/parisc/eisa.c. 1274 1275 elanfreq= [X86-32] 1276 See comment before function elanfreq_setup() in 1277 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1278 1279 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1280 Specifies physical address of start of kernel core 1281 image elf header and optionally the size. Generally 1282 kexec loader will pass this option to capture kernel. 1283 See Documentation/admin-guide/kdump/kdump.rst for details. 1284 1285 enable_mtrr_cleanup [X86] 1286 The kernel tries to adjust MTRR layout from continuous 1287 to discrete, to make X server driver able to add WB 1288 entry later. This parameter enables that. 1289 1290 enable_timer_pin_1 [X86] 1291 Enable PIN 1 of APIC timer 1292 Can be useful to work around chipset bugs 1293 (in particular on some ATI chipsets). 1294 The kernel tries to set a reasonable default. 1295 1296 enforcing [SELINUX] Set initial enforcing status. 1297 Format: {"0" | "1"} 1298 See security/selinux/Kconfig help text. 1299 0 -- permissive (log only, no denials). 1300 1 -- enforcing (deny and log). 1301 Default value is 0. 1302 Value can be changed at runtime via 1303 /sys/fs/selinux/enforce. 1304 1305 erst_disable [ACPI] 1306 Disable Error Record Serialization Table (ERST) 1307 support. 1308 1309 ether= [HW,NET] Ethernet cards parameters 1310 This option is obsoleted by the "netdev=" option, which 1311 has equivalent usage. See its documentation for details. 1312 1313 evm= [EVM] 1314 Format: { "fix" } 1315 Permit 'security.evm' to be updated regardless of 1316 current integrity status. 1317 1318 failslab= 1319 fail_page_alloc= 1320 fail_make_request=[KNL] 1321 General fault injection mechanism. 1322 Format: <interval>,<probability>,<space>,<times> 1323 See also Documentation/fault-injection/. 1324 1325 floppy= [HW] 1326 See Documentation/admin-guide/blockdev/floppy.rst. 1327 1328 force_pal_cache_flush 1329 [IA-64] Avoid check_sal_cache_flush which may hang on 1330 buggy SAL_CACHE_FLUSH implementations. Using this 1331 parameter will force ia64_sal_cache_flush to call 1332 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1333 1334 forcepae [X86-32] 1335 Forcefully enable Physical Address Extension (PAE). 1336 Many Pentium M systems disable PAE but may have a 1337 functionally usable PAE implementation. 1338 Warning: use of this parameter will taint the kernel 1339 and may cause unknown problems. 1340 1341 ftrace=[tracer] 1342 [FTRACE] will set and start the specified tracer 1343 as early as possible in order to facilitate early 1344 boot debugging. 1345 1346 ftrace_dump_on_oops[=orig_cpu] 1347 [FTRACE] will dump the trace buffers on oops. 1348 If no parameter is passed, ftrace will dump 1349 buffers of all CPUs, but if you pass orig_cpu, it will 1350 dump only the buffer of the CPU that triggered the 1351 oops. 1352 1353 ftrace_filter=[function-list] 1354 [FTRACE] Limit the functions traced by the function 1355 tracer at boot up. function-list is a comma separated 1356 list of functions. This list can be changed at run 1357 time by the set_ftrace_filter file in the debugfs 1358 tracing directory. 1359 1360 ftrace_notrace=[function-list] 1361 [FTRACE] Do not trace the functions specified in 1362 function-list. This list can be changed at run time 1363 by the set_ftrace_notrace file in the debugfs 1364 tracing directory. 1365 1366 ftrace_graph_filter=[function-list] 1367 [FTRACE] Limit the top level callers functions traced 1368 by the function graph tracer at boot up. 1369 function-list is a comma separated list of functions 1370 that can be changed at run time by the 1371 set_graph_function file in the debugfs tracing directory. 1372 1373 ftrace_graph_notrace=[function-list] 1374 [FTRACE] Do not trace from the functions specified in 1375 function-list. This list is a comma separated list of 1376 functions that can be changed at run time by the 1377 set_graph_notrace file in the debugfs tracing directory. 1378 1379 ftrace_graph_max_depth=<uint> 1380 [FTRACE] Used with the function graph tracer. This is 1381 the max depth it will trace into a function. This value 1382 can be changed at run time by the max_graph_depth file 1383 in the tracefs tracing directory. default: 0 (no limit) 1384 1385 fw_devlink= [KNL] Create device links between consumer and supplier 1386 devices by scanning the firmware to infer the 1387 consumer/supplier relationships. This feature is 1388 especially useful when drivers are loaded as modules as 1389 it ensures proper ordering of tasks like device probing 1390 (suppliers first, then consumers), supplier boot state 1391 clean up (only after all consumers have probed), 1392 suspend/resume & runtime PM (consumers first, then 1393 suppliers). 1394 Format: { off | permissive | on | rpm } 1395 off -- Don't create device links from firmware info. 1396 permissive -- Create device links from firmware info 1397 but use it only for ordering boot state clean 1398 up (sync_state() calls). 1399 on -- Create device links from firmware info and use it 1400 to enforce probe and suspend/resume ordering. 1401 rpm -- Like "on", but also use to order runtime PM. 1402 1403 gamecon.map[2|3]= 1404 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1405 support via parallel port (up to 5 devices per port) 1406 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1407 See also Documentation/input/devices/joystick-parport.rst 1408 1409 gamma= [HW,DRM] 1410 1411 gart_fix_e820= [X86_64] disable the fix e820 for K8 GART 1412 Format: off | on 1413 default: on 1414 1415 gcov_persist= [GCOV] When non-zero (default), profiling data for 1416 kernel modules is saved and remains accessible via 1417 debugfs, even when the module is unloaded/reloaded. 1418 When zero, profiling data is discarded and associated 1419 debugfs files are removed at module unload time. 1420 1421 goldfish [X86] Enable the goldfish android emulator platform. 1422 Don't use this when you are not running on the 1423 android emulator 1424 1425 gpt [EFI] Forces disk with valid GPT signature but 1426 invalid Protective MBR to be treated as GPT. If the 1427 primary GPT is corrupted, it enables the backup/alternate 1428 GPT to be used instead. 1429 1430 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1431 the "Enable 0" bit of the configuration register. 1432 Format: 0 | 1 1433 Default: 0 1434 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1435 the "Enable 0" bit of the configuration register. 1436 Format: 0 | 1 1437 Default: 0 1438 grcan.select= [HW] Select which physical interface to use. 1439 Format: 0 | 1 1440 Default: 0 1441 grcan.txsize= [HW] Sets the size of the tx buffer. 1442 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1443 Default: 1024 1444 grcan.rxsize= [HW] Sets the size of the rx buffer. 1445 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1446 Default: 1024 1447 1448 gpio-mockup.gpio_mockup_ranges 1449 [HW] Sets the ranges of gpiochip of for this device. 1450 Format: <start1>,<end1>,<start2>,<end2>... 1451 1452 hardlockup_all_cpu_backtrace= 1453 [KNL] Should the hard-lockup detector generate 1454 backtraces on all cpus. 1455 Format: 0 | 1 1456 1457 hashdist= [KNL,NUMA] Large hashes allocated during boot 1458 are distributed across NUMA nodes. Defaults on 1459 for 64-bit NUMA, off otherwise. 1460 Format: 0 | 1 (for off | on) 1461 1462 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1463 1464 hd= [EIDE] (E)IDE hard drive subsystem geometry 1465 Format: <cyl>,<head>,<sect> 1466 1467 hest_disable [ACPI] 1468 Disable Hardware Error Source Table (HEST) support; 1469 corresponding firmware-first mode error processing 1470 logic will be disabled. 1471 1472 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1473 size of <nn>. This works even on boxes that have no 1474 highmem otherwise. This also works to reduce highmem 1475 size on bigger boxes. 1476 1477 highres= [KNL] Enable/disable high resolution timer mode. 1478 Valid parameters: "on", "off" 1479 Default: "on" 1480 1481 hlt [BUGS=ARM,SH] 1482 1483 hpet= [X86-32,HPET] option to control HPET usage 1484 Format: { enable (default) | disable | force | 1485 verbose } 1486 disable: disable HPET and use PIT instead 1487 force: allow force enabled of undocumented chips (ICH4, 1488 VIA, nVidia) 1489 verbose: show contents of HPET registers during setup 1490 1491 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1492 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1493 1494 hugetlb_cma= [HW] The size of a cma area used for allocation 1495 of gigantic hugepages. 1496 Format: nn[KMGTPE] 1497 1498 Reserve a cma area of given size and allocate gigantic 1499 hugepages using the cma allocator. If enabled, the 1500 boot-time allocation of gigantic hugepages is skipped. 1501 1502 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1503 If this follows hugepagesz (below), it specifies 1504 the number of pages of hugepagesz to be allocated. 1505 If this is the first HugeTLB parameter on the command 1506 line, it specifies the number of pages to allocate for 1507 the default huge page size. See also 1508 Documentation/admin-guide/mm/hugetlbpage.rst. 1509 Format: <integer> 1510 1511 hugepagesz= 1512 [HW] The size of the HugeTLB pages. This is used in 1513 conjunction with hugepages (above) to allocate huge 1514 pages of a specific size at boot. The pair 1515 hugepagesz=X hugepages=Y can be specified once for 1516 each supported huge page size. Huge page sizes are 1517 architecture dependent. See also 1518 Documentation/admin-guide/mm/hugetlbpage.rst. 1519 Format: size[KMG] 1520 1521 hung_task_panic= 1522 [KNL] Should the hung task detector generate panics. 1523 Format: 0 | 1 1524 1525 A value of 1 instructs the kernel to panic when a 1526 hung task is detected. The default value is controlled 1527 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1528 option. The value selected by this boot parameter can 1529 be changed later by the kernel.hung_task_panic sysctl. 1530 1531 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1532 terminal devices. Valid values: 0..8 1533 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1534 If specified, z/VM IUCV HVC accepts connections 1535 from listed z/VM user IDs only. 1536 1537 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1538 which allow the hypervisor to 'idle' the 1539 guest on lock contention. 1540 1541 keep_bootcon [KNL] 1542 Do not unregister boot console at start. This is only 1543 useful for debugging when something happens in the window 1544 between unregistering the boot console and initializing 1545 the real console. 1546 1547 i2c_bus= [HW] Override the default board specific I2C bus speed 1548 or register an additional I2C bus that is not 1549 registered from board initialization code. 1550 Format: 1551 <bus_id>,<clkrate> 1552 1553 i8042.debug [HW] Toggle i8042 debug mode 1554 i8042.unmask_kbd_data 1555 [HW] Enable printing of interrupt data from the KBD port 1556 (disabled by default, and as a pre-condition 1557 requires that i8042.debug=1 be enabled) 1558 i8042.direct [HW] Put keyboard port into non-translated mode 1559 i8042.dumbkbd [HW] Pretend that controller can only read data from 1560 keyboard and cannot control its state 1561 (Don't attempt to blink the leds) 1562 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1563 i8042.nokbd [HW] Don't check/create keyboard port 1564 i8042.noloop [HW] Disable the AUX Loopback command while probing 1565 for the AUX port 1566 i8042.nomux [HW] Don't check presence of an active multiplexing 1567 controller 1568 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1569 controllers 1570 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1571 i8042.reset [HW] Reset the controller during init, cleanup and 1572 suspend-to-ram transitions, only during s2r 1573 transitions, or never reset 1574 Format: { 1 | Y | y | 0 | N | n } 1575 1, Y, y: always reset controller 1576 0, N, n: don't ever reset controller 1577 Default: only on s2r transitions on x86; most other 1578 architectures force reset to be always executed 1579 i8042.unlock [HW] Unlock (ignore) the keylock 1580 i8042.kbdreset [HW] Reset device connected to KBD port 1581 1582 i810= [HW,DRM] 1583 1584 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1585 indicates that the driver is running on unsupported 1586 hardware. 1587 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1588 does not match list of supported models. 1589 i8k.power_status 1590 [HW] Report power status in /proc/i8k 1591 (disabled by default) 1592 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1593 capability is set. 1594 1595 i915.invert_brightness= 1596 [DRM] Invert the sense of the variable that is used to 1597 set the brightness of the panel backlight. Normally a 1598 brightness value of 0 indicates backlight switched off, 1599 and the maximum of the brightness value sets the backlight 1600 to maximum brightness. If this parameter is set to 0 1601 (default) and the machine requires it, or this parameter 1602 is set to 1, a brightness value of 0 sets the backlight 1603 to maximum brightness, and the maximum of the brightness 1604 value switches the backlight off. 1605 -1 -- never invert brightness 1606 0 -- machine default 1607 1 -- force brightness inversion 1608 1609 icn= [HW,ISDN] 1610 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1611 1612 ide-core.nodma= [HW] (E)IDE subsystem 1613 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1614 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1615 .cdrom .chs .ignore_cable are additional options 1616 See Documentation/ide/ide.rst. 1617 1618 ide-generic.probe-mask= [HW] (E)IDE subsystem 1619 Format: <int> 1620 Probe mask for legacy ISA IDE ports. Depending on 1621 platform up to 6 ports are supported, enabled by 1622 setting corresponding bits in the mask to 1. The 1623 default value is 0x0, which has a special meaning. 1624 On systems that have PCI, it triggers scanning the 1625 PCI bus for the first and the second port, which 1626 are then probed. On systems without PCI the value 1627 of 0x0 enables probing the two first ports as if it 1628 was 0x3. 1629 1630 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1631 Claim all unknown PCI IDE storage controllers. 1632 1633 idle= [X86] 1634 Format: idle=poll, idle=halt, idle=nomwait 1635 Poll forces a polling idle loop that can slightly 1636 improve the performance of waking up a idle CPU, but 1637 will use a lot of power and make the system run hot. 1638 Not recommended. 1639 idle=halt: Halt is forced to be used for CPU idle. 1640 In such case C2/C3 won't be used again. 1641 idle=nomwait: Disable mwait for CPU C-states 1642 1643 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1644 Format: { strict | legacy | 2008 | relaxed } 1645 Default: strict 1646 1647 Choose which programs will be accepted for execution 1648 based on the IEEE 754 NaN encoding(s) supported by 1649 the FPU and the NaN encoding requested with the value 1650 of an ELF file header flag individually set by each 1651 binary. Hardware implementations are permitted to 1652 support either or both of the legacy and the 2008 NaN 1653 encoding mode. 1654 1655 Available settings are as follows: 1656 strict accept binaries that request a NaN encoding 1657 supported by the FPU 1658 legacy only accept legacy-NaN binaries, if supported 1659 by the FPU 1660 2008 only accept 2008-NaN binaries, if supported 1661 by the FPU 1662 relaxed accept any binaries regardless of whether 1663 supported by the FPU 1664 1665 The FPU emulator is always able to support both NaN 1666 encodings, so if no FPU hardware is present or it has 1667 been disabled with 'nofpu', then the settings of 1668 'legacy' and '2008' strap the emulator accordingly, 1669 'relaxed' straps the emulator for both legacy-NaN and 1670 2008-NaN, whereas 'strict' enables legacy-NaN only on 1671 legacy processors and both NaN encodings on MIPS32 or 1672 MIPS64 CPUs. 1673 1674 The setting for ABS.fmt/NEG.fmt instruction execution 1675 mode generally follows that for the NaN encoding, 1676 except where unsupported by hardware. 1677 1678 ignore_loglevel [KNL] 1679 Ignore loglevel setting - this will print /all/ 1680 kernel messages to the console. Useful for debugging. 1681 We also add it as printk module parameter, so users 1682 could change it dynamically, usually by 1683 /sys/module/printk/parameters/ignore_loglevel. 1684 1685 ignore_rlimit_data 1686 Ignore RLIMIT_DATA setting for data mappings, 1687 print warning at first misuse. Can be changed via 1688 /sys/module/kernel/parameters/ignore_rlimit_data. 1689 1690 ihash_entries= [KNL] 1691 Set number of hash buckets for inode cache. 1692 1693 ima_appraise= [IMA] appraise integrity measurements 1694 Format: { "off" | "enforce" | "fix" | "log" } 1695 default: "enforce" 1696 1697 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1698 The builtin appraise policy appraises all files 1699 owned by uid=0. 1700 1701 ima_canonical_fmt [IMA] 1702 Use the canonical format for the binary runtime 1703 measurements, instead of host native format. 1704 1705 ima_hash= [IMA] 1706 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1707 | sha512 | ... } 1708 default: "sha1" 1709 1710 The list of supported hash algorithms is defined 1711 in crypto/hash_info.h. 1712 1713 ima_policy= [IMA] 1714 The builtin policies to load during IMA setup. 1715 Format: "tcb | appraise_tcb | secure_boot | 1716 fail_securely" 1717 1718 The "tcb" policy measures all programs exec'd, files 1719 mmap'd for exec, and all files opened with the read 1720 mode bit set by either the effective uid (euid=0) or 1721 uid=0. 1722 1723 The "appraise_tcb" policy appraises the integrity of 1724 all files owned by root. 1725 1726 The "secure_boot" policy appraises the integrity 1727 of files (eg. kexec kernel image, kernel modules, 1728 firmware, policy, etc) based on file signatures. 1729 1730 The "fail_securely" policy forces file signature 1731 verification failure also on privileged mounted 1732 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1733 flag. 1734 1735 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1736 Load a policy which meets the needs of the Trusted 1737 Computing Base. This means IMA will measure all 1738 programs exec'd, files mmap'd for exec, and all files 1739 opened for read by uid=0. 1740 1741 ima_template= [IMA] 1742 Select one of defined IMA measurements template formats. 1743 Formats: { "ima" | "ima-ng" | "ima-sig" } 1744 Default: "ima-ng" 1745 1746 ima_template_fmt= 1747 [IMA] Define a custom template format. 1748 Format: { "field1|...|fieldN" } 1749 1750 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1751 Format: <min_file_size> 1752 Set the minimal file size for using asynchronous hash. 1753 If left unspecified, ahash usage is disabled. 1754 1755 ahash performance varies for different data sizes on 1756 different crypto accelerators. This option can be used 1757 to achieve the best performance for a particular HW. 1758 1759 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1760 Format: <bufsize> 1761 Set hashing buffer size. Default: 4k. 1762 1763 ahash performance varies for different chunk sizes on 1764 different crypto accelerators. This option can be used 1765 to achieve best performance for particular HW. 1766 1767 init= [KNL] 1768 Format: <full_path> 1769 Run specified binary instead of /sbin/init as init 1770 process. 1771 1772 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1773 for working out where the kernel is dying during 1774 startup. 1775 1776 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1777 initcall functions. Useful for debugging built-in 1778 modules and initcalls. 1779 1780 initrd= [BOOT] Specify the location of the initial ramdisk 1781 1782 initrdmem= [KNL] Specify a physical address and size from which to 1783 load the initrd. If an initrd is compiled in or 1784 specified in the bootparams, it takes priority over this 1785 setting. 1786 Format: ss[KMG],nn[KMG] 1787 Default is 0, 0 1788 1789 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1790 zeroes. 1791 Format: 0 | 1 1792 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1793 1794 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1795 Format: 0 | 1 1796 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1797 1798 init_pkru= [x86] Specify the default memory protection keys rights 1799 register contents for all processes. 0x55555554 by 1800 default (disallow access to all but pkey 0). Can 1801 override in debugfs after boot. 1802 1803 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1804 Format: <irq> 1805 1806 int_pln_enable [x86] Enable power limit notification interrupt 1807 1808 integrity_audit=[IMA] 1809 Format: { "0" | "1" } 1810 0 -- basic integrity auditing messages. (Default) 1811 1 -- additional integrity auditing messages. 1812 1813 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1814 on 1815 Enable intel iommu driver. 1816 off 1817 Disable intel iommu driver. 1818 igfx_off [Default Off] 1819 By default, gfx is mapped as normal device. If a gfx 1820 device has a dedicated DMAR unit, the DMAR unit is 1821 bypassed by not enabling DMAR with this option. In 1822 this case, gfx device will use physical address for 1823 DMA. 1824 forcedac [x86_64] 1825 With this option iommu will not optimize to look 1826 for io virtual address below 32-bit forcing dual 1827 address cycle on pci bus for cards supporting greater 1828 than 32-bit addressing. The default is to look 1829 for translation below 32-bit and if not available 1830 then look in the higher range. 1831 strict [Default Off] 1832 With this option on every unmap_single operation will 1833 result in a hardware IOTLB flush operation as opposed 1834 to batching them for performance. 1835 sp_off [Default Off] 1836 By default, super page will be supported if Intel IOMMU 1837 has the capability. With this option, super page will 1838 not be supported. 1839 sm_on [Default Off] 1840 By default, scalable mode will be disabled even if the 1841 hardware advertises that it has support for the scalable 1842 mode translation. With this option set, scalable mode 1843 will be used on hardware which claims to support it. 1844 tboot_noforce [Default Off] 1845 Do not force the Intel IOMMU enabled under tboot. 1846 By default, tboot will force Intel IOMMU on, which 1847 could harm performance of some high-throughput 1848 devices like 40GBit network cards, even if identity 1849 mapping is enabled. 1850 Note that using this option lowers the security 1851 provided by tboot because it makes the system 1852 vulnerable to DMA attacks. 1853 nobounce [Default off] 1854 Disable bounce buffer for untrusted devices such as 1855 the Thunderbolt devices. This will treat the untrusted 1856 devices as the trusted ones, hence might expose security 1857 risks of DMA attacks. 1858 1859 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1860 0 disables intel_idle and fall back on acpi_idle. 1861 1 to 9 specify maximum depth of C-state. 1862 1863 intel_pstate= [X86] 1864 disable 1865 Do not enable intel_pstate as the default 1866 scaling driver for the supported processors 1867 passive 1868 Use intel_pstate as a scaling driver, but configure it 1869 to work with generic cpufreq governors (instead of 1870 enabling its internal governor). This mode cannot be 1871 used along with the hardware-managed P-states (HWP) 1872 feature. 1873 force 1874 Enable intel_pstate on systems that prohibit it by default 1875 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1876 instead of acpi-cpufreq may disable platform features, such 1877 as thermal controls and power capping, that rely on ACPI 1878 P-States information being indicated to OSPM and therefore 1879 should be used with caution. This option does not work with 1880 processors that aren't supported by the intel_pstate driver 1881 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1882 no_hwp 1883 Do not enable hardware P state control (HWP) 1884 if available. 1885 hwp_only 1886 Only load intel_pstate on systems which support 1887 hardware P state control (HWP) if available. 1888 support_acpi_ppc 1889 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1890 Description Table, specifies preferred power management 1891 profile as "Enterprise Server" or "Performance Server", 1892 then this feature is turned on by default. 1893 per_cpu_perf_limits 1894 Allow per-logical-CPU P-State performance control limits using 1895 cpufreq sysfs interface 1896 1897 intremap= [X86-64, Intel-IOMMU] 1898 on enable Interrupt Remapping (default) 1899 off disable Interrupt Remapping 1900 nosid disable Source ID checking 1901 no_x2apic_optout 1902 BIOS x2APIC opt-out request will be ignored 1903 nopost disable Interrupt Posting 1904 1905 iomem= Disable strict checking of access to MMIO memory 1906 strict regions from userspace. 1907 relaxed 1908 1909 iommu= [x86] 1910 off 1911 force 1912 noforce 1913 biomerge 1914 panic 1915 nopanic 1916 merge 1917 nomerge 1918 soft 1919 pt [x86] 1920 nopt [x86] 1921 nobypass [PPC/POWERNV] 1922 Disable IOMMU bypass, using IOMMU for PCI devices. 1923 1924 iommu.strict= [ARM64] Configure TLB invalidation behaviour 1925 Format: { "0" | "1" } 1926 0 - Lazy mode. 1927 Request that DMA unmap operations use deferred 1928 invalidation of hardware TLBs, for increased 1929 throughput at the cost of reduced device isolation. 1930 Will fall back to strict mode if not supported by 1931 the relevant IOMMU driver. 1932 1 - Strict mode (default). 1933 DMA unmap operations invalidate IOMMU hardware TLBs 1934 synchronously. 1935 1936 iommu.passthrough= 1937 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 1938 Format: { "0" | "1" } 1939 0 - Use IOMMU translation for DMA. 1940 1 - Bypass the IOMMU for DMA. 1941 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 1942 1943 io7= [HW] IO7 for Marvel based alpha systems 1944 See comment before marvel_specify_io7 in 1945 arch/alpha/kernel/core_marvel.c. 1946 1947 io_delay= [X86] I/O delay method 1948 0x80 1949 Standard port 0x80 based delay 1950 0xed 1951 Alternate port 0xed based delay (needed on some systems) 1952 udelay 1953 Simple two microseconds delay 1954 none 1955 No delay 1956 1957 ip= [IP_PNP] 1958 See Documentation/admin-guide/nfs/nfsroot.rst. 1959 1960 ipcmni_extend [KNL] Extend the maximum number of unique System V 1961 IPC identifiers from 32,768 to 16,777,216. 1962 1963 irqaffinity= [SMP] Set the default irq affinity mask 1964 The argument is a cpu list, as described above. 1965 1966 irqchip.gicv2_force_probe= 1967 [ARM, ARM64] 1968 Format: <bool> 1969 Force the kernel to look for the second 4kB page 1970 of a GICv2 controller even if the memory range 1971 exposed by the device tree is too small. 1972 1973 irqchip.gicv3_nolpi= 1974 [ARM, ARM64] 1975 Force the kernel to ignore the availability of 1976 LPIs (and by consequence ITSs). Intended for system 1977 that use the kernel as a bootloader, and thus want 1978 to let secondary kernels in charge of setting up 1979 LPIs. 1980 1981 irqchip.gicv3_pseudo_nmi= [ARM64] 1982 Enables support for pseudo-NMIs in the kernel. This 1983 requires the kernel to be built with 1984 CONFIG_ARM64_PSEUDO_NMI. 1985 1986 irqfixup [HW] 1987 When an interrupt is not handled search all handlers 1988 for it. Intended to get systems with badly broken 1989 firmware running. 1990 1991 irqpoll [HW] 1992 When an interrupt is not handled search all handlers 1993 for it. Also check all handlers each timer 1994 interrupt. Intended to get systems with badly broken 1995 firmware running. 1996 1997 isapnp= [ISAPNP] 1998 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 1999 2000 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2001 [Deprecated - use cpusets instead] 2002 Format: [flag-list,]<cpu-list> 2003 2004 Specify one or more CPUs to isolate from disturbances 2005 specified in the flag list (default: domain): 2006 2007 nohz 2008 Disable the tick when a single task runs. 2009 2010 A residual 1Hz tick is offloaded to workqueues, which you 2011 need to affine to housekeeping through the global 2012 workqueue's affinity configured via the 2013 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2014 by using the 'domain' flag described below. 2015 2016 NOTE: by default the global workqueue runs on all CPUs, 2017 so to protect individual CPUs the 'cpumask' file has to 2018 be configured manually after bootup. 2019 2020 domain 2021 Isolate from the general SMP balancing and scheduling 2022 algorithms. Note that performing domain isolation this way 2023 is irreversible: it's not possible to bring back a CPU to 2024 the domains once isolated through isolcpus. It's strongly 2025 advised to use cpusets instead to disable scheduler load 2026 balancing through the "cpuset.sched_load_balance" file. 2027 It offers a much more flexible interface where CPUs can 2028 move in and out of an isolated set anytime. 2029 2030 You can move a process onto or off an "isolated" CPU via 2031 the CPU affinity syscalls or cpuset. 2032 <cpu number> begins at 0 and the maximum value is 2033 "number of CPUs in system - 1". 2034 2035 managed_irq 2036 2037 Isolate from being targeted by managed interrupts 2038 which have an interrupt mask containing isolated 2039 CPUs. The affinity of managed interrupts is 2040 handled by the kernel and cannot be changed via 2041 the /proc/irq/* interfaces. 2042 2043 This isolation is best effort and only effective 2044 if the automatically assigned interrupt mask of a 2045 device queue contains isolated and housekeeping 2046 CPUs. If housekeeping CPUs are online then such 2047 interrupts are directed to the housekeeping CPU 2048 so that IO submitted on the housekeeping CPU 2049 cannot disturb the isolated CPU. 2050 2051 If a queue's affinity mask contains only isolated 2052 CPUs then this parameter has no effect on the 2053 interrupt routing decision, though interrupts are 2054 only delivered when tasks running on those 2055 isolated CPUs submit IO. IO submitted on 2056 housekeeping CPUs has no influence on those 2057 queues. 2058 2059 The format of <cpu-list> is described above. 2060 2061 iucv= [HW,NET] 2062 2063 ivrs_ioapic [HW,X86_64] 2064 Provide an override to the IOAPIC-ID<->DEVICE-ID 2065 mapping provided in the IVRS ACPI table. For 2066 example, to map IOAPIC-ID decimal 10 to 2067 PCI device 00:14.0 write the parameter as: 2068 ivrs_ioapic[10]=00:14.0 2069 2070 ivrs_hpet [HW,X86_64] 2071 Provide an override to the HPET-ID<->DEVICE-ID 2072 mapping provided in the IVRS ACPI table. For 2073 example, to map HPET-ID decimal 0 to 2074 PCI device 00:14.0 write the parameter as: 2075 ivrs_hpet[0]=00:14.0 2076 2077 ivrs_acpihid [HW,X86_64] 2078 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2079 mapping provided in the IVRS ACPI table. For 2080 example, to map UART-HID:UID AMD0020:0 to 2081 PCI device 00:14.5 write the parameter as: 2082 ivrs_acpihid[00:14.5]=AMD0020:0 2083 2084 js= [HW,JOY] Analog joystick 2085 See Documentation/input/joydev/joystick.rst. 2086 2087 nokaslr [KNL] 2088 When CONFIG_RANDOMIZE_BASE is set, this disables 2089 kernel and module base offset ASLR (Address Space 2090 Layout Randomization). 2091 2092 kasan_multi_shot 2093 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2094 report on every invalid memory access. Without this 2095 parameter KASAN will print report only for the first 2096 invalid access. 2097 2098 keepinitrd [HW,ARM] 2099 2100 kernelcore= [KNL,X86,IA-64,PPC] 2101 Format: nn[KMGTPE] | nn% | "mirror" 2102 This parameter specifies the amount of memory usable by 2103 the kernel for non-movable allocations. The requested 2104 amount is spread evenly throughout all nodes in the 2105 system as ZONE_NORMAL. The remaining memory is used for 2106 movable memory in its own zone, ZONE_MOVABLE. In the 2107 event, a node is too small to have both ZONE_NORMAL and 2108 ZONE_MOVABLE, kernelcore memory will take priority and 2109 other nodes will have a larger ZONE_MOVABLE. 2110 2111 ZONE_MOVABLE is used for the allocation of pages that 2112 may be reclaimed or moved by the page migration 2113 subsystem. Note that allocations like PTEs-from-HighMem 2114 still use the HighMem zone if it exists, and the Normal 2115 zone if it does not. 2116 2117 It is possible to specify the exact amount of memory in 2118 the form of "nn[KMGTPE]", a percentage of total system 2119 memory in the form of "nn%", or "mirror". If "mirror" 2120 option is specified, mirrored (reliable) memory is used 2121 for non-movable allocations and remaining memory is used 2122 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2123 are exclusive, so you cannot specify multiple forms. 2124 2125 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2126 Format: <Controller#>[,poll interval] 2127 The controller # is the number of the ehci usb debug 2128 port as it is probed via PCI. The poll interval is 2129 optional and is the number seconds in between 2130 each poll cycle to the debug port in case you need 2131 the functionality for interrupting the kernel with 2132 gdb or control-c on the dbgp connection. When 2133 not using this parameter you use sysrq-g to break into 2134 the kernel debugger. 2135 2136 kgdboc= [KGDB,HW] kgdb over consoles. 2137 Requires a tty driver that supports console polling, 2138 or a supported polling keyboard driver (non-usb). 2139 Serial only format: <serial_device>[,baud] 2140 keyboard only format: kbd 2141 keyboard and serial format: kbd,<serial_device>[,baud] 2142 Optional Kernel mode setting: 2143 kms, kbd format: kms,kbd 2144 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2145 2146 kgdboc_earlycon= [KGDB,HW] 2147 If the boot console provides the ability to read 2148 characters and can work in polling mode, you can use 2149 this parameter to tell kgdb to use it as a backend 2150 until the normal console is registered. Intended to 2151 be used together with the kgdboc parameter which 2152 specifies the normal console to transition to. 2153 2154 The name of the early console should be specified 2155 as the value of this parameter. Note that the name of 2156 the early console might be different than the tty 2157 name passed to kgdboc. It's OK to leave the value 2158 blank and the first boot console that implements 2159 read() will be picked. 2160 2161 kgdbwait [KGDB] Stop kernel execution and enter the 2162 kernel debugger at the earliest opportunity. 2163 2164 kmac= [MIPS] korina ethernet MAC address. 2165 Configure the RouterBoard 532 series on-chip 2166 Ethernet adapter MAC address. 2167 2168 kmemleak= [KNL] Boot-time kmemleak enable/disable 2169 Valid arguments: on, off 2170 Default: on 2171 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2172 the default is off. 2173 2174 kprobe_event=[probe-list] 2175 [FTRACE] Add kprobe events and enable at boot time. 2176 The probe-list is a semicolon delimited list of probe 2177 definitions. Each definition is same as kprobe_events 2178 interface, but the parameters are comma delimited. 2179 For example, to add a kprobe event on vfs_read with 2180 arg1 and arg2, add to the command line; 2181 2182 kprobe_event=p,vfs_read,$arg1,$arg2 2183 2184 See also Documentation/trace/kprobetrace.rst "Kernel 2185 Boot Parameter" section. 2186 2187 kpti= [ARM64] Control page table isolation of user 2188 and kernel address spaces. 2189 Default: enabled on cores which need mitigation. 2190 0: force disabled 2191 1: force enabled 2192 2193 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2194 Default is 0 (don't ignore, but inject #GP) 2195 2196 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2197 Default is false (don't support). 2198 2199 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2200 KVM MMU at runtime. 2201 Default is 0 (off) 2202 2203 kvm.nx_huge_pages= 2204 [KVM] Controls the software workaround for the 2205 X86_BUG_ITLB_MULTIHIT bug. 2206 force : Always deploy workaround. 2207 off : Never deploy workaround. 2208 auto : Deploy workaround based on the presence of 2209 X86_BUG_ITLB_MULTIHIT. 2210 2211 Default is 'auto'. 2212 2213 If the software workaround is enabled for the host, 2214 guests do need not to enable it for nested guests. 2215 2216 kvm.nx_huge_pages_recovery_ratio= 2217 [KVM] Controls how many 4KiB pages are periodically zapped 2218 back to huge pages. 0 disables the recovery, otherwise if 2219 the value is N KVM will zap 1/Nth of the 4KiB pages every 2220 minute. The default is 60. 2221 2222 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2223 Default is 1 (enabled) 2224 2225 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2226 for all guests. 2227 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2228 2229 kvm-arm.vgic_v3_group0_trap= 2230 [KVM,ARM] Trap guest accesses to GICv3 group-0 2231 system registers 2232 2233 kvm-arm.vgic_v3_group1_trap= 2234 [KVM,ARM] Trap guest accesses to GICv3 group-1 2235 system registers 2236 2237 kvm-arm.vgic_v3_common_trap= 2238 [KVM,ARM] Trap guest accesses to GICv3 common 2239 system registers 2240 2241 kvm-arm.vgic_v4_enable= 2242 [KVM,ARM] Allow use of GICv4 for direct injection of 2243 LPIs. 2244 2245 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2246 (virtualized MMU) support on capable Intel chips. 2247 Default is 1 (enabled) 2248 2249 kvm-intel.emulate_invalid_guest_state= 2250 [KVM,Intel] Enable emulation of invalid guest states 2251 Default is 0 (disabled) 2252 2253 kvm-intel.flexpriority= 2254 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2255 Default is 1 (enabled) 2256 2257 kvm-intel.nested= 2258 [KVM,Intel] Enable VMX nesting (nVMX). 2259 Default is 0 (disabled) 2260 2261 kvm-intel.unrestricted_guest= 2262 [KVM,Intel] Disable unrestricted guest feature 2263 (virtualized real and unpaged mode) on capable 2264 Intel chips. Default is 1 (enabled) 2265 2266 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2267 CVE-2018-3620. 2268 2269 Valid arguments: never, cond, always 2270 2271 always: L1D cache flush on every VMENTER. 2272 cond: Flush L1D on VMENTER only when the code between 2273 VMEXIT and VMENTER can leak host memory. 2274 never: Disables the mitigation 2275 2276 Default is cond (do L1 cache flush in specific instances) 2277 2278 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2279 feature (tagged TLBs) on capable Intel chips. 2280 Default is 1 (enabled) 2281 2282 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2283 affected CPUs 2284 2285 The kernel PTE inversion protection is unconditionally 2286 enabled and cannot be disabled. 2287 2288 full 2289 Provides all available mitigations for the 2290 L1TF vulnerability. Disables SMT and 2291 enables all mitigations in the 2292 hypervisors, i.e. unconditional L1D flush. 2293 2294 SMT control and L1D flush control via the 2295 sysfs interface is still possible after 2296 boot. Hypervisors will issue a warning 2297 when the first VM is started in a 2298 potentially insecure configuration, 2299 i.e. SMT enabled or L1D flush disabled. 2300 2301 full,force 2302 Same as 'full', but disables SMT and L1D 2303 flush runtime control. Implies the 2304 'nosmt=force' command line option. 2305 (i.e. sysfs control of SMT is disabled.) 2306 2307 flush 2308 Leaves SMT enabled and enables the default 2309 hypervisor mitigation, i.e. conditional 2310 L1D flush. 2311 2312 SMT control and L1D flush control via the 2313 sysfs interface is still possible after 2314 boot. Hypervisors will issue a warning 2315 when the first VM is started in a 2316 potentially insecure configuration, 2317 i.e. SMT enabled or L1D flush disabled. 2318 2319 flush,nosmt 2320 2321 Disables SMT and enables the default 2322 hypervisor mitigation. 2323 2324 SMT control and L1D flush control via the 2325 sysfs interface is still possible after 2326 boot. Hypervisors will issue a warning 2327 when the first VM is started in a 2328 potentially insecure configuration, 2329 i.e. SMT enabled or L1D flush disabled. 2330 2331 flush,nowarn 2332 Same as 'flush', but hypervisors will not 2333 warn when a VM is started in a potentially 2334 insecure configuration. 2335 2336 off 2337 Disables hypervisor mitigations and doesn't 2338 emit any warnings. 2339 It also drops the swap size and available 2340 RAM limit restriction on both hypervisor and 2341 bare metal. 2342 2343 Default is 'flush'. 2344 2345 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2346 2347 l2cr= [PPC] 2348 2349 l3cr= [PPC] 2350 2351 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2352 disabled it. 2353 2354 lapic= [x86,APIC] "notscdeadline" Do not use TSC deadline 2355 value for LAPIC timer one-shot implementation. Default 2356 back to the programmable timer unit in the LAPIC. 2357 2358 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2359 in C2 power state. 2360 2361 libata.dma= [LIBATA] DMA control 2362 libata.dma=0 Disable all PATA and SATA DMA 2363 libata.dma=1 PATA and SATA Disk DMA only 2364 libata.dma=2 ATAPI (CDROM) DMA only 2365 libata.dma=4 Compact Flash DMA only 2366 Combinations also work, so libata.dma=3 enables DMA 2367 for disks and CDROMs, but not CFs. 2368 2369 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2370 libata.ignore_hpa=0 keep BIOS limits (default) 2371 libata.ignore_hpa=1 ignore limits, using full disk 2372 2373 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2374 when set. 2375 Format: <int> 2376 2377 libata.force= [LIBATA] Force configurations. The format is comma 2378 separated list of "[ID:]VAL" where ID is 2379 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2380 matching port, link or device. Basically, it matches 2381 the ATA ID string printed on console by libata. If 2382 the whole ID part is omitted, the last PORT and DEVICE 2383 values are used. If ID hasn't been specified yet, the 2384 configuration applies to all ports, links and devices. 2385 2386 If only DEVICE is omitted, the parameter applies to 2387 the port and all links and devices behind it. DEVICE 2388 number of 0 either selects the first device or the 2389 first fan-out link behind PMP device. It does not 2390 select the host link. DEVICE number of 15 selects the 2391 host link and device attached to it. 2392 2393 The VAL specifies the configuration to force. As long 2394 as there's no ambiguity shortcut notation is allowed. 2395 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2396 The following configurations can be forced. 2397 2398 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2399 Any ID with matching PORT is used. 2400 2401 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2402 2403 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2404 udma[/][16,25,33,44,66,100,133] notation is also 2405 allowed. 2406 2407 * [no]ncq: Turn on or off NCQ. 2408 2409 * [no]ncqtrim: Turn off queued DSM TRIM. 2410 2411 * nohrst, nosrst, norst: suppress hard, soft 2412 and both resets. 2413 2414 * rstonce: only attempt one reset during 2415 hot-unplug link recovery 2416 2417 * dump_id: dump IDENTIFY data. 2418 2419 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2420 2421 * disable: Disable this device. 2422 2423 If there are multiple matching configurations changing 2424 the same attribute, the last one is used. 2425 2426 memblock=debug [KNL] Enable memblock debug messages. 2427 2428 load_ramdisk= [RAM] List of ramdisks to load from floppy 2429 See Documentation/admin-guide/blockdev/ramdisk.rst. 2430 2431 lockd.nlm_grace_period=P [NFS] Assign grace period. 2432 Format: <integer> 2433 2434 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2435 Format: <integer> 2436 2437 lockd.nlm_timeout=T [NFS] Assign timeout value. 2438 Format: <integer> 2439 2440 lockd.nlm_udpport=M [NFS] Assign UDP port. 2441 Format: <integer> 2442 2443 lockdown= [SECURITY] 2444 { integrity | confidentiality } 2445 Enable the kernel lockdown feature. If set to 2446 integrity, kernel features that allow userland to 2447 modify the running kernel are disabled. If set to 2448 confidentiality, kernel features that allow userland 2449 to extract confidential information from the kernel 2450 are also disabled. 2451 2452 locktorture.nreaders_stress= [KNL] 2453 Set the number of locking read-acquisition kthreads. 2454 Defaults to being automatically set based on the 2455 number of online CPUs. 2456 2457 locktorture.nwriters_stress= [KNL] 2458 Set the number of locking write-acquisition kthreads. 2459 2460 locktorture.onoff_holdoff= [KNL] 2461 Set time (s) after boot for CPU-hotplug testing. 2462 2463 locktorture.onoff_interval= [KNL] 2464 Set time (s) between CPU-hotplug operations, or 2465 zero to disable CPU-hotplug testing. 2466 2467 locktorture.shuffle_interval= [KNL] 2468 Set task-shuffle interval (jiffies). Shuffling 2469 tasks allows some CPUs to go into dyntick-idle 2470 mode during the locktorture test. 2471 2472 locktorture.shutdown_secs= [KNL] 2473 Set time (s) after boot system shutdown. This 2474 is useful for hands-off automated testing. 2475 2476 locktorture.stat_interval= [KNL] 2477 Time (s) between statistics printk()s. 2478 2479 locktorture.stutter= [KNL] 2480 Time (s) to stutter testing, for example, 2481 specifying five seconds causes the test to run for 2482 five seconds, wait for five seconds, and so on. 2483 This tests the locking primitive's ability to 2484 transition abruptly to and from idle. 2485 2486 locktorture.torture_type= [KNL] 2487 Specify the locking implementation to test. 2488 2489 locktorture.verbose= [KNL] 2490 Enable additional printk() statements. 2491 2492 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2493 Format: <irq> 2494 2495 loglevel= All Kernel Messages with a loglevel smaller than the 2496 console loglevel will be printed to the console. It can 2497 also be changed with klogd or other programs. The 2498 loglevels are defined as follows: 2499 2500 0 (KERN_EMERG) system is unusable 2501 1 (KERN_ALERT) action must be taken immediately 2502 2 (KERN_CRIT) critical conditions 2503 3 (KERN_ERR) error conditions 2504 4 (KERN_WARNING) warning conditions 2505 5 (KERN_NOTICE) normal but significant condition 2506 6 (KERN_INFO) informational 2507 7 (KERN_DEBUG) debug-level messages 2508 2509 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2510 in bytes. n must be a power of two and greater 2511 than the minimal size. The minimal size is defined 2512 by LOG_BUF_SHIFT kernel config parameter. There is 2513 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2514 that allows to increase the default size depending on 2515 the number of CPUs. See init/Kconfig for more details. 2516 2517 logo.nologo [FB] Disables display of the built-in Linux logo. 2518 This may be used to provide more screen space for 2519 kernel log messages and is useful when debugging 2520 kernel boot problems. 2521 2522 lp=0 [LP] Specify parallel ports to use, e.g, 2523 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2524 lp=reset first parallel port). 'lp=0' disables the 2525 lp=auto printer driver. 'lp=reset' (which can be 2526 specified in addition to the ports) causes 2527 attached printers to be reset. Using 2528 lp=port1,port2,... specifies the parallel ports 2529 to associate lp devices with, starting with 2530 lp0. A port specification may be 'none' to skip 2531 that lp device, or a parport name such as 2532 'parport0'. Specifying 'lp=auto' instead of a 2533 port specification list means that device IDs 2534 from each port should be examined, to see if 2535 an IEEE 1284-compliant printer is attached; if 2536 so, the driver will manage that printer. 2537 See also header of drivers/char/lp.c. 2538 2539 lpj=n [KNL] 2540 Sets loops_per_jiffy to given constant, thus avoiding 2541 time-consuming boot-time autodetection (up to 250 ms per 2542 CPU). 0 enables autodetection (default). To determine 2543 the correct value for your kernel, boot with normal 2544 autodetection and see what value is printed. Note that 2545 on SMP systems the preset will be applied to all CPUs, 2546 which is likely to cause problems if your CPUs need 2547 significantly divergent settings. An incorrect value 2548 will cause delays in the kernel to be wrong, leading to 2549 unpredictable I/O errors and other breakage. Although 2550 unlikely, in the extreme case this might damage your 2551 hardware. 2552 2553 ltpc= [NET] 2554 Format: <io>,<irq>,<dma> 2555 2556 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2557 2558 lsm=lsm1,...,lsmN 2559 [SECURITY] Choose order of LSM initialization. This 2560 overrides CONFIG_LSM, and the "security=" parameter. 2561 2562 machvec= [IA-64] Force the use of a particular machine-vector 2563 (machvec) in a generic kernel. 2564 Example: machvec=hpzx1 2565 2566 machtype= [Loongson] Share the same kernel image file between different 2567 yeeloong laptop. 2568 Example: machtype=lemote-yeeloong-2f-7inch 2569 2570 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2571 than or equal to this physical address is ignored. 2572 2573 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2574 will bring up during bootup. maxcpus=n : n >= 0 limits 2575 the kernel to bring up 'n' processors. Surely after 2576 bootup you can bring up the other plugged cpu by executing 2577 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2578 only takes effect during system bootup. 2579 While n=0 is a special case, it is equivalent to "nosmp", 2580 which also disables the IO APIC. 2581 2582 max_loop= [LOOP] The number of loop block devices that get 2583 (loop.max_loop) unconditionally pre-created at init time. The default 2584 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2585 of statically allocating a predefined number, loop 2586 devices can be requested on-demand with the 2587 /dev/loop-control interface. 2588 2589 mce [X86-32] Machine Check Exception 2590 2591 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2592 2593 md= [HW] RAID subsystems devices and level 2594 See Documentation/admin-guide/md.rst. 2595 2596 mdacon= [MDA] 2597 Format: <first>,<last> 2598 Specifies range of consoles to be captured by the MDA. 2599 2600 mds= [X86,INTEL] 2601 Control mitigation for the Micro-architectural Data 2602 Sampling (MDS) vulnerability. 2603 2604 Certain CPUs are vulnerable to an exploit against CPU 2605 internal buffers which can forward information to a 2606 disclosure gadget under certain conditions. 2607 2608 In vulnerable processors, the speculatively 2609 forwarded data can be used in a cache side channel 2610 attack, to access data to which the attacker does 2611 not have direct access. 2612 2613 This parameter controls the MDS mitigation. The 2614 options are: 2615 2616 full - Enable MDS mitigation on vulnerable CPUs 2617 full,nosmt - Enable MDS mitigation and disable 2618 SMT on vulnerable CPUs 2619 off - Unconditionally disable MDS mitigation 2620 2621 On TAA-affected machines, mds=off can be prevented by 2622 an active TAA mitigation as both vulnerabilities are 2623 mitigated with the same mechanism so in order to disable 2624 this mitigation, you need to specify tsx_async_abort=off 2625 too. 2626 2627 Not specifying this option is equivalent to 2628 mds=full. 2629 2630 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2631 2632 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2633 Amount of memory to be used in cases as follows: 2634 2635 1 for test; 2636 2 when the kernel is not able to see the whole system memory; 2637 3 memory that lies after 'mem=' boundary is excluded from 2638 the hypervisor, then assigned to KVM guests. 2639 2640 [X86] Work as limiting max address. Use together 2641 with memmap= to avoid physical address space collisions. 2642 Without memmap= PCI devices could be placed at addresses 2643 belonging to unused RAM. 2644 2645 Note that this only takes effects during boot time since 2646 in above case 3, memory may need be hot added after boot 2647 if system memory of hypervisor is not sufficient. 2648 2649 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2650 memory. 2651 2652 memchunk=nn[KMG] 2653 [KNL,SH] Allow user to override the default size for 2654 per-device physically contiguous DMA buffers. 2655 2656 memhp_default_state=online/offline 2657 [KNL] Set the initial state for the memory hotplug 2658 onlining policy. If not specified, the default value is 2659 set according to the 2660 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2661 option. 2662 See Documentation/admin-guide/mm/memory-hotplug.rst. 2663 2664 memmap=exactmap [KNL,X86] Enable setting of an exact 2665 E820 memory map, as specified by the user. 2666 Such memmap=exactmap lines can be constructed based on 2667 BIOS output or other requirements. See the memmap=nn@ss 2668 option description. 2669 2670 memmap=nn[KMG]@ss[KMG] 2671 [KNL] Force usage of a specific region of memory. 2672 Region of memory to be used is from ss to ss+nn. 2673 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2674 which limits max address to nn[KMG]. 2675 Multiple different regions can be specified, 2676 comma delimited. 2677 Example: 2678 memmap=100M@2G,100M#3G,1G!1024G 2679 2680 memmap=nn[KMG]#ss[KMG] 2681 [KNL,ACPI] Mark specific memory as ACPI data. 2682 Region of memory to be marked is from ss to ss+nn. 2683 2684 memmap=nn[KMG]$ss[KMG] 2685 [KNL,ACPI] Mark specific memory as reserved. 2686 Region of memory to be reserved is from ss to ss+nn. 2687 Example: Exclude memory from 0x18690000-0x1869ffff 2688 memmap=64K$0x18690000 2689 or 2690 memmap=0x10000$0x18690000 2691 Some bootloaders may need an escape character before '$', 2692 like Grub2, otherwise '$' and the following number 2693 will be eaten. 2694 2695 memmap=nn[KMG]!ss[KMG] 2696 [KNL,X86] Mark specific memory as protected. 2697 Region of memory to be used, from ss to ss+nn. 2698 The memory region may be marked as e820 type 12 (0xc) 2699 and is NVDIMM or ADR memory. 2700 2701 memmap=<size>%<offset>-<oldtype>+<newtype> 2702 [KNL,ACPI] Convert memory within the specified region 2703 from <oldtype> to <newtype>. If "-<oldtype>" is left 2704 out, the whole region will be marked as <newtype>, 2705 even if previously unavailable. If "+<newtype>" is left 2706 out, matching memory will be removed. Types are 2707 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2708 3 = ACPI, 12 = PRAM. 2709 2710 memory_corruption_check=0/1 [X86] 2711 Some BIOSes seem to corrupt the first 64k of 2712 memory when doing things like suspend/resume. 2713 Setting this option will scan the memory 2714 looking for corruption. Enabling this will 2715 both detect corruption and prevent the kernel 2716 from using the memory being corrupted. 2717 However, its intended as a diagnostic tool; if 2718 repeatable BIOS-originated corruption always 2719 affects the same memory, you can use memmap= 2720 to prevent the kernel from using that memory. 2721 2722 memory_corruption_check_size=size [X86] 2723 By default it checks for corruption in the low 2724 64k, making this memory unavailable for normal 2725 use. Use this parameter to scan for 2726 corruption in more or less memory. 2727 2728 memory_corruption_check_period=seconds [X86] 2729 By default it checks for corruption every 60 2730 seconds. Use this parameter to check at some 2731 other rate. 0 disables periodic checking. 2732 2733 memtest= [KNL,X86,ARM,PPC] Enable memtest 2734 Format: <integer> 2735 default : 0 <disable> 2736 Specifies the number of memtest passes to be 2737 performed. Each pass selects another test 2738 pattern from a given set of patterns. Memtest 2739 fills the memory with this pattern, validates 2740 memory contents and reserves bad memory 2741 regions that are detected. 2742 2743 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2744 Valid arguments: on, off 2745 Default (depends on kernel configuration option): 2746 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2747 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2748 mem_encrypt=on: Activate SME 2749 mem_encrypt=off: Do not activate SME 2750 2751 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2752 for details on when memory encryption can be activated. 2753 2754 mem_sleep_default= [SUSPEND] Default system suspend mode: 2755 s2idle - Suspend-To-Idle 2756 shallow - Power-On Suspend or equivalent (if supported) 2757 deep - Suspend-To-RAM or equivalent (if supported) 2758 See Documentation/admin-guide/pm/sleep-states.rst. 2759 2760 meye.*= [HW] Set MotionEye Camera parameters 2761 See Documentation/admin-guide/media/meye.rst. 2762 2763 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2764 Multi-Function General Purpose Timers on AMD Geode 2765 platforms. 2766 2767 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2768 the BIOS has incorrectly applied a workaround. TinyBIOS 2769 version 0.98 is known to be affected, 0.99 fixes the 2770 problem by letting the user disable the workaround. 2771 2772 mga= [HW,DRM] 2773 2774 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2775 physical address is ignored. 2776 2777 mini2440= [ARM,HW,KNL] 2778 Format:[0..2][b][c][t] 2779 Default: "0tb" 2780 MINI2440 configuration specification: 2781 0 - The attached screen is the 3.5" TFT 2782 1 - The attached screen is the 7" TFT 2783 2 - The VGA Shield is attached (1024x768) 2784 Leaving out the screen size parameter will not load 2785 the TFT driver, and the framebuffer will be left 2786 unconfigured. 2787 b - Enable backlight. The TFT backlight pin will be 2788 linked to the kernel VESA blanking code and a GPIO 2789 LED. This parameter is not necessary when using the 2790 VGA shield. 2791 c - Enable the s3c camera interface. 2792 t - Reserved for enabling touchscreen support. The 2793 touchscreen support is not enabled in the mainstream 2794 kernel as of 2.6.30, a preliminary port can be found 2795 in the "bleeding edge" mini2440 support kernel at 2796 https://repo.or.cz/w/linux-2.6/mini2440.git 2797 2798 mitigations= 2799 [X86,PPC,S390,ARM64] Control optional mitigations for 2800 CPU vulnerabilities. This is a set of curated, 2801 arch-independent options, each of which is an 2802 aggregation of existing arch-specific options. 2803 2804 off 2805 Disable all optional CPU mitigations. This 2806 improves system performance, but it may also 2807 expose users to several CPU vulnerabilities. 2808 Equivalent to: nopti [X86,PPC] 2809 kpti=0 [ARM64] 2810 nospectre_v1 [X86,PPC] 2811 nobp=0 [S390] 2812 nospectre_v2 [X86,PPC,S390,ARM64] 2813 spectre_v2_user=off [X86] 2814 spec_store_bypass_disable=off [X86,PPC] 2815 ssbd=force-off [ARM64] 2816 l1tf=off [X86] 2817 mds=off [X86] 2818 tsx_async_abort=off [X86] 2819 kvm.nx_huge_pages=off [X86] 2820 2821 Exceptions: 2822 This does not have any effect on 2823 kvm.nx_huge_pages when 2824 kvm.nx_huge_pages=force. 2825 2826 auto (default) 2827 Mitigate all CPU vulnerabilities, but leave SMT 2828 enabled, even if it's vulnerable. This is for 2829 users who don't want to be surprised by SMT 2830 getting disabled across kernel upgrades, or who 2831 have other ways of avoiding SMT-based attacks. 2832 Equivalent to: (default behavior) 2833 2834 auto,nosmt 2835 Mitigate all CPU vulnerabilities, disabling SMT 2836 if needed. This is for users who always want to 2837 be fully mitigated, even if it means losing SMT. 2838 Equivalent to: l1tf=flush,nosmt [X86] 2839 mds=full,nosmt [X86] 2840 tsx_async_abort=full,nosmt [X86] 2841 2842 mminit_loglevel= 2843 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2844 parameter allows control of the logging verbosity for 2845 the additional memory initialisation checks. A value 2846 of 0 disables mminit logging and a level of 4 will 2847 log everything. Information is printed at KERN_DEBUG 2848 so loglevel=8 may also need to be specified. 2849 2850 module.sig_enforce 2851 [KNL] When CONFIG_MODULE_SIG is set, this means that 2852 modules without (valid) signatures will fail to load. 2853 Note that if CONFIG_MODULE_SIG_FORCE is set, that 2854 is always true, so this option does nothing. 2855 2856 module_blacklist= [KNL] Do not load a comma-separated list of 2857 modules. Useful for debugging problem modules. 2858 2859 mousedev.tap_time= 2860 [MOUSE] Maximum time between finger touching and 2861 leaving touchpad surface for touch to be considered 2862 a tap and be reported as a left button click (for 2863 touchpads working in absolute mode only). 2864 Format: <msecs> 2865 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 2866 reporting absolute coordinates, such as tablets 2867 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 2868 reporting absolute coordinates, such as tablets 2869 2870 movablecore= [KNL,X86,IA-64,PPC] 2871 Format: nn[KMGTPE] | nn% 2872 This parameter is the complement to kernelcore=, it 2873 specifies the amount of memory used for migratable 2874 allocations. If both kernelcore and movablecore is 2875 specified, then kernelcore will be at *least* the 2876 specified value but may be more. If movablecore on its 2877 own is specified, the administrator must be careful 2878 that the amount of memory usable for all allocations 2879 is not too small. 2880 2881 movable_node [KNL] Boot-time switch to make hotplugable memory 2882 NUMA nodes to be movable. This means that the memory 2883 of such nodes will be usable only for movable 2884 allocations which rules out almost all kernel 2885 allocations. Use with caution! 2886 2887 MTD_Partition= [MTD] 2888 Format: <name>,<region-number>,<size>,<offset> 2889 2890 MTD_Region= [MTD] Format: 2891 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 2892 2893 mtdparts= [MTD] 2894 See drivers/mtd/parsers/cmdlinepart.c 2895 2896 multitce=off [PPC] This parameter disables the use of the pSeries 2897 firmware feature for updating multiple TCE entries 2898 at a time. 2899 2900 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 2901 2902 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 2903 2904 boundary - index of last SLC block on Flex-OneNAND. 2905 The remaining blocks are configured as MLC blocks. 2906 lock - Configure if Flex-OneNAND boundary should be locked. 2907 Once locked, the boundary cannot be changed. 2908 1 indicates lock status, 0 indicates unlock status. 2909 2910 mtdset= [ARM] 2911 ARM/S3C2412 JIVE boot control 2912 2913 See arch/arm/mach-s3c2412/mach-jive.c 2914 2915 mtouchusb.raw_coordinates= 2916 [HW] Make the MicroTouch USB driver use raw coordinates 2917 ('y', default) or cooked coordinates ('n') 2918 2919 mtrr_chunk_size=nn[KMG] [X86] 2920 used for mtrr cleanup. It is largest continuous chunk 2921 that could hold holes aka. UC entries. 2922 2923 mtrr_gran_size=nn[KMG] [X86] 2924 Used for mtrr cleanup. It is granularity of mtrr block. 2925 Default is 1. 2926 Large value could prevent small alignment from 2927 using up MTRRs. 2928 2929 mtrr_spare_reg_nr=n [X86] 2930 Format: <integer> 2931 Range: 0,7 : spare reg number 2932 Default : 1 2933 Used for mtrr cleanup. It is spare mtrr entries number. 2934 Set to 2 or more if your graphical card needs more. 2935 2936 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 2937 2938 netdev= [NET] Network devices parameters 2939 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 2940 Note that mem_start is often overloaded to mean 2941 something different and driver-specific. 2942 This usage is only documented in each driver source 2943 file if at all. 2944 2945 nf_conntrack.acct= 2946 [NETFILTER] Enable connection tracking flow accounting 2947 0 to disable accounting 2948 1 to enable accounting 2949 Default value is 0. 2950 2951 nfsaddrs= [NFS] Deprecated. Use ip= instead. 2952 See Documentation/admin-guide/nfs/nfsroot.rst. 2953 2954 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 2955 See Documentation/admin-guide/nfs/nfsroot.rst. 2956 2957 nfsrootdebug [NFS] enable nfsroot debugging messages. 2958 See Documentation/admin-guide/nfs/nfsroot.rst. 2959 2960 nfs.callback_nr_threads= 2961 [NFSv4] set the total number of threads that the 2962 NFS client will assign to service NFSv4 callback 2963 requests. 2964 2965 nfs.callback_tcpport= 2966 [NFS] set the TCP port on which the NFSv4 callback 2967 channel should listen. 2968 2969 nfs.cache_getent= 2970 [NFS] sets the pathname to the program which is used 2971 to update the NFS client cache entries. 2972 2973 nfs.cache_getent_timeout= 2974 [NFS] sets the timeout after which an attempt to 2975 update a cache entry is deemed to have failed. 2976 2977 nfs.idmap_cache_timeout= 2978 [NFS] set the maximum lifetime for idmapper cache 2979 entries. 2980 2981 nfs.enable_ino64= 2982 [NFS] enable 64-bit inode numbers. 2983 If zero, the NFS client will fake up a 32-bit inode 2984 number for the readdir() and stat() syscalls instead 2985 of returning the full 64-bit number. 2986 The default is to return 64-bit inode numbers. 2987 2988 nfs.max_session_cb_slots= 2989 [NFSv4.1] Sets the maximum number of session 2990 slots the client will assign to the callback 2991 channel. This determines the maximum number of 2992 callbacks the client will process in parallel for 2993 a particular server. 2994 2995 nfs.max_session_slots= 2996 [NFSv4.1] Sets the maximum number of session slots 2997 the client will attempt to negotiate with the server. 2998 This limits the number of simultaneous RPC requests 2999 that the client can send to the NFSv4.1 server. 3000 Note that there is little point in setting this 3001 value higher than the max_tcp_slot_table_limit. 3002 3003 nfs.nfs4_disable_idmapping= 3004 [NFSv4] When set to the default of '1', this option 3005 ensures that both the RPC level authentication 3006 scheme and the NFS level operations agree to use 3007 numeric uids/gids if the mount is using the 3008 'sec=sys' security flavour. In effect it is 3009 disabling idmapping, which can make migration from 3010 legacy NFSv2/v3 systems to NFSv4 easier. 3011 Servers that do not support this mode of operation 3012 will be autodetected by the client, and it will fall 3013 back to using the idmapper. 3014 To turn off this behaviour, set the value to '0'. 3015 nfs.nfs4_unique_id= 3016 [NFS4] Specify an additional fixed unique ident- 3017 ification string that NFSv4 clients can insert into 3018 their nfs_client_id4 string. This is typically a 3019 UUID that is generated at system install time. 3020 3021 nfs.send_implementation_id = 3022 [NFSv4.1] Send client implementation identification 3023 information in exchange_id requests. 3024 If zero, no implementation identification information 3025 will be sent. 3026 The default is to send the implementation identification 3027 information. 3028 3029 nfs.recover_lost_locks = 3030 [NFSv4] Attempt to recover locks that were lost due 3031 to a lease timeout on the server. Please note that 3032 doing this risks data corruption, since there are 3033 no guarantees that the file will remain unchanged 3034 after the locks are lost. 3035 If you want to enable the kernel legacy behaviour of 3036 attempting to recover these locks, then set this 3037 parameter to '1'. 3038 The default parameter value of '0' causes the kernel 3039 not to attempt recovery of lost locks. 3040 3041 nfs4.layoutstats_timer = 3042 [NFSv4.2] Change the rate at which the kernel sends 3043 layoutstats to the pNFS metadata server. 3044 3045 Setting this to value to 0 causes the kernel to use 3046 whatever value is the default set by the layout 3047 driver. A non-zero value sets the minimum interval 3048 in seconds between layoutstats transmissions. 3049 3050 nfsd.nfs4_disable_idmapping= 3051 [NFSv4] When set to the default of '1', the NFSv4 3052 server will return only numeric uids and gids to 3053 clients using auth_sys, and will accept numeric uids 3054 and gids from such clients. This is intended to ease 3055 migration from NFSv2/v3. 3056 3057 nmi_debug= [KNL,SH] Specify one or more actions to take 3058 when a NMI is triggered. 3059 Format: [state][,regs][,debounce][,die] 3060 3061 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3062 Format: [panic,][nopanic,][num] 3063 Valid num: 0 or 1 3064 0 - turn hardlockup detector in nmi_watchdog off 3065 1 - turn hardlockup detector in nmi_watchdog on 3066 When panic is specified, panic when an NMI watchdog 3067 timeout occurs (or 'nopanic' to not panic on an NMI 3068 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3069 To disable both hard and soft lockup detectors, 3070 please see 'nowatchdog'. 3071 This is useful when you use a panic=... timeout and 3072 need the box quickly up again. 3073 3074 These settings can be accessed at runtime via 3075 the nmi_watchdog and hardlockup_panic sysctls. 3076 3077 netpoll.carrier_timeout= 3078 [NET] Specifies amount of time (in seconds) that 3079 netpoll should wait for a carrier. By default netpoll 3080 waits 4 seconds. 3081 3082 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3083 emulation library even if a 387 maths coprocessor 3084 is present. 3085 3086 no5lvl [X86-64] Disable 5-level paging mode. Forces 3087 kernel to use 4-level paging instead. 3088 3089 nofsgsbase [X86] Disables FSGSBASE instructions. 3090 3091 no_console_suspend 3092 [HW] Never suspend the console 3093 Disable suspending of consoles during suspend and 3094 hibernate operations. Once disabled, debugging 3095 messages can reach various consoles while the rest 3096 of the system is being put to sleep (ie, while 3097 debugging driver suspend/resume hooks). This may 3098 not work reliably with all consoles, but is known 3099 to work with serial and VGA consoles. 3100 To facilitate more flexible debugging, we also add 3101 console_suspend, a printk module parameter to control 3102 it. Users could use console_suspend (usually 3103 /sys/module/printk/parameters/console_suspend) to 3104 turn on/off it dynamically. 3105 3106 novmcoredd [KNL,KDUMP] 3107 Disable device dump. Device dump allows drivers to 3108 append dump data to vmcore so you can collect driver 3109 specified debug info. Drivers can append the data 3110 without any limit and this data is stored in memory, 3111 so this may cause significant memory stress. Disabling 3112 device dump can help save memory but the driver debug 3113 data will be no longer available. This parameter 3114 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3115 is set. 3116 3117 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3118 caches in the slab allocator. Saves per-node memory, 3119 but will impact performance. 3120 3121 noalign [KNL,ARM] 3122 3123 noaltinstr [S390] Disables alternative instructions patching 3124 (CPU alternatives feature). 3125 3126 noapic [SMP,APIC] Tells the kernel to not make use of any 3127 IOAPICs that may be present in the system. 3128 3129 noautogroup Disable scheduler automatic task group creation. 3130 3131 nobats [PPC] Do not use BATs for mapping kernel lowmem 3132 on "Classic" PPC cores. 3133 3134 nocache [ARM] 3135 3136 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 3137 3138 nodelayacct [KNL] Disable per-task delay accounting 3139 3140 nodsp [SH] Disable hardware DSP at boot time. 3141 3142 noefi Disable EFI runtime services support. 3143 3144 noexec [IA-64] 3145 3146 noexec [X86] 3147 On X86-32 available only on PAE configured kernels. 3148 noexec=on: enable non-executable mappings (default) 3149 noexec=off: disable non-executable mappings 3150 3151 nosmap [X86,PPC] 3152 Disable SMAP (Supervisor Mode Access Prevention) 3153 even if it is supported by processor. 3154 3155 nosmep [X86,PPC] 3156 Disable SMEP (Supervisor Mode Execution Prevention) 3157 even if it is supported by processor. 3158 3159 noexec32 [X86-64] 3160 This affects only 32-bit executables. 3161 noexec32=on: enable non-executable mappings (default) 3162 read doesn't imply executable mappings 3163 noexec32=off: disable non-executable mappings 3164 read implies executable mappings 3165 3166 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3167 3168 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3169 register save and restore. The kernel will only save 3170 legacy floating-point registers on task switch. 3171 3172 nohugeiomap [KNL,x86,PPC] Disable kernel huge I/O mappings. 3173 3174 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3175 Equivalent to smt=1. 3176 3177 [KNL,x86] Disable symmetric multithreading (SMT). 3178 nosmt=force: Force disable SMT, cannot be undone 3179 via the sysfs control file. 3180 3181 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3182 (bounds check bypass). With this option data leaks are 3183 possible in the system. 3184 3185 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3186 the Spectre variant 2 (indirect branch prediction) 3187 vulnerability. System may allow data leaks with this 3188 option. 3189 3190 nospec_store_bypass_disable 3191 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3192 3193 noxsave [BUGS=X86] Disables x86 extended register state save 3194 and restore using xsave. The kernel will fallback to 3195 enabling legacy floating-point and sse state. 3196 3197 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3198 register states. The kernel will fall back to use 3199 xsave to save the states. By using this parameter, 3200 performance of saving the states is degraded because 3201 xsave doesn't support modified optimization while 3202 xsaveopt supports it on xsaveopt enabled systems. 3203 3204 noxsaves [X86] Disables xsaves and xrstors used in saving and 3205 restoring x86 extended register state in compacted 3206 form of xsave area. The kernel will fall back to use 3207 xsaveopt and xrstor to save and restore the states 3208 in standard form of xsave area. By using this 3209 parameter, xsave area per process might occupy more 3210 memory on xsaves enabled systems. 3211 3212 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or 3213 wfi(ARM) instruction doesn't work correctly and not to 3214 use it. This is also useful when using JTAG debugger. 3215 3216 no_file_caps Tells the kernel not to honor file capabilities. The 3217 only way then for a file to be executed with privilege 3218 is to be setuid root or executed by root. 3219 3220 nohalt [IA-64] Tells the kernel not to use the power saving 3221 function PAL_HALT_LIGHT when idle. This increases 3222 power-consumption. On the positive side, it reduces 3223 interrupt wake-up latency, which may improve performance 3224 in certain environments such as networked servers or 3225 real-time systems. 3226 3227 nohibernate [HIBERNATION] Disable hibernation and resume. 3228 3229 nohz= [KNL] Boottime enable/disable dynamic ticks 3230 Valid arguments: on, off 3231 Default: on 3232 3233 nohz_full= [KNL,BOOT,SMP,ISOL] 3234 The argument is a cpu list, as described above. 3235 In kernels built with CONFIG_NO_HZ_FULL=y, set 3236 the specified list of CPUs whose tick will be stopped 3237 whenever possible. The boot CPU will be forced outside 3238 the range to maintain the timekeeping. Any CPUs 3239 in this list will have their RCU callbacks offloaded, 3240 just as if they had also been called out in the 3241 rcu_nocbs= boot parameter. 3242 3243 noiotrap [SH] Disables trapped I/O port accesses. 3244 3245 noirqdebug [X86-32] Disables the code which attempts to detect and 3246 disable unhandled interrupt sources. 3247 3248 no_timer_check [X86,APIC] Disables the code which tests for 3249 broken timer IRQ sources. 3250 3251 noisapnp [ISAPNP] Disables ISA PnP code. 3252 3253 noinitrd [RAM] Tells the kernel not to load any configured 3254 initial RAM disk. 3255 3256 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3257 remapping. 3258 [Deprecated - use intremap=off] 3259 3260 nointroute [IA-64] 3261 3262 noinvpcid [X86] Disable the INVPCID cpu feature. 3263 3264 nojitter [IA-64] Disables jitter checking for ITC timers. 3265 3266 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3267 3268 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3269 fault handling. 3270 3271 no-vmw-sched-clock 3272 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3273 clock and use the default one. 3274 3275 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time 3276 accounting. steal time is computed, but won't 3277 influence scheduler behaviour 3278 3279 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3280 3281 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3282 3283 noltlbs [PPC] Do not use large page/tlb entries for kernel 3284 lowmem mapping on PPC40x and PPC8xx 3285 3286 nomca [IA-64] Disable machine check abort handling 3287 3288 nomce [X86-32] Disable Machine Check Exception 3289 3290 nomfgpt [X86-32] Disable Multi-Function General Purpose 3291 Timer usage (for AMD Geode machines). 3292 3293 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3294 shutdown the other cpus. Instead use the REBOOT_VECTOR 3295 irq. 3296 3297 nomodule Disable module load 3298 3299 nopat [X86] Disable PAT (page attribute table extension of 3300 pagetables) support. 3301 3302 nopcid [X86-64] Disable the PCID cpu feature. 3303 3304 norandmaps Don't use address space randomization. Equivalent to 3305 echo 0 > /proc/sys/kernel/randomize_va_space 3306 3307 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3308 with UP alternatives 3309 3310 nordrand [X86] Disable kernel use of the RDRAND and 3311 RDSEED instructions even if they are supported 3312 by the processor. RDRAND and RDSEED are still 3313 available to user space applications. 3314 3315 noresume [SWSUSP] Disables resume and restores original swap 3316 space. 3317 3318 no-scroll [VGA] Disables scrollback. 3319 This is required for the Braillex ib80-piezo Braille 3320 reader made by F.H. Papenmeier (Germany). 3321 3322 nosbagart [IA-64] 3323 3324 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3325 3326 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3327 and disable the IO APIC. legacy for "maxcpus=0". 3328 3329 nosoftlockup [KNL] Disable the soft-lockup detector. 3330 3331 nosync [HW,M68K] Disables sync negotiation for all devices. 3332 3333 nowatchdog [KNL] Disable both lockup detectors, i.e. 3334 soft-lockup and NMI watchdog (hard-lockup). 3335 3336 nowb [ARM] 3337 3338 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3339 3340 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3341 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3342 Some features depend on CPU0. Known dependencies are: 3343 1. Resume from suspend/hibernate depends on CPU0. 3344 Suspend/hibernate will fail if CPU0 is offline and you 3345 need to online CPU0 before suspend/hibernate. 3346 2. PIC interrupts also depend on CPU0. CPU0 can't be 3347 removed if a PIC interrupt is detected. 3348 It's said poweroff/reboot may depend on CPU0 on some 3349 machines although I haven't seen such issues so far 3350 after CPU0 is offline on a few tested machines. 3351 If the dependencies are under your control, you can 3352 turn on cpu0_hotplug. 3353 3354 nps_mtm_hs_ctr= [KNL,ARC] 3355 This parameter sets the maximum duration, in 3356 cycles, each HW thread of the CTOP can run 3357 without interruptions, before HW switches it. 3358 The actual maximum duration is 16 times this 3359 parameter's value. 3360 Format: integer between 1 and 255 3361 Default: 255 3362 3363 nptcg= [IA-64] Override max number of concurrent global TLB 3364 purges which is reported from either PAL_VM_SUMMARY or 3365 SAL PALO. 3366 3367 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3368 could support. nr_cpus=n : n >= 1 limits the kernel to 3369 support 'n' processors. It could be larger than the 3370 number of already plugged CPU during bootup, later in 3371 runtime you can physically add extra cpu until it reaches 3372 n. So during boot up some boot time memory for per-cpu 3373 variables need be pre-allocated for later physical cpu 3374 hot plugging. 3375 3376 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3377 3378 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing. 3379 Allowed values are enable and disable 3380 3381 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3382 'node', 'default' can be specified 3383 This can be set from sysctl after boot. 3384 See Documentation/admin-guide/sysctl/vm.rst for details. 3385 3386 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3387 See Documentation/core-api/debugging-via-ohci1394.rst for more 3388 info. 3389 3390 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3391 Rather than timing out after 20 ms if an EC 3392 command is not properly ACKed, override the length 3393 of the timeout. We have interrupts disabled while 3394 waiting for the ACK, so if this is set too high 3395 interrupts *may* be lost! 3396 3397 omap_mux= [OMAP] Override bootloader pin multiplexing. 3398 Format: <mux_mode0.mode_name=value>... 3399 For example, to override I2C bus2: 3400 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3401 3402 oprofile.timer= [HW] 3403 Use timer interrupt instead of performance counters 3404 3405 oprofile.cpu_type= Force an oprofile cpu type 3406 This might be useful if you have an older oprofile 3407 userland or if you want common events. 3408 Format: { arch_perfmon } 3409 arch_perfmon: [X86] Force use of architectural 3410 perfmon on Intel CPUs instead of the 3411 CPU specific event set. 3412 timer: [X86] Force use of architectural NMI 3413 timer mode (see also oprofile.timer 3414 for generic hr timer mode) 3415 3416 oops=panic Always panic on oopses. Default is to just kill the 3417 process, but there is a small probability of 3418 deadlocking the machine. 3419 This will also cause panics on machine check exceptions. 3420 Useful together with panic=30 to trigger a reboot. 3421 3422 page_alloc.shuffle= 3423 [KNL] Boolean flag to control whether the page allocator 3424 should randomize its free lists. The randomization may 3425 be automatically enabled if the kernel detects it is 3426 running on a platform with a direct-mapped memory-side 3427 cache, and this parameter can be used to 3428 override/disable that behavior. The state of the flag 3429 can be read from sysfs at: 3430 /sys/module/page_alloc/parameters/shuffle. 3431 3432 page_owner= [KNL] Boot-time page_owner enabling option. 3433 Storage of the information about who allocated 3434 each page is disabled in default. With this switch, 3435 we can turn it on. 3436 on: enable the feature 3437 3438 page_poison= [KNL] Boot-time parameter changing the state of 3439 poisoning on the buddy allocator, available with 3440 CONFIG_PAGE_POISONING=y. 3441 off: turn off poisoning (default) 3442 on: turn on poisoning 3443 3444 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3445 timeout > 0: seconds before rebooting 3446 timeout = 0: wait forever 3447 timeout < 0: reboot immediately 3448 Format: <timeout> 3449 3450 panic_print= Bitmask for printing system info when panic happens. 3451 User can chose combination of the following bits: 3452 bit 0: print all tasks info 3453 bit 1: print system memory info 3454 bit 2: print timer info 3455 bit 3: print locks info if CONFIG_LOCKDEP is on 3456 bit 4: print ftrace buffer 3457 bit 5: print all printk messages in buffer 3458 3459 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 3460 Format: <hex>[,nousertaint] 3461 Hexadecimal bitmask representing the set of TAINT flags 3462 that will cause the kernel to panic when add_taint() is 3463 called with any of the flags in this set. 3464 The optional switch "nousertaint" can be utilized to 3465 prevent userspace forced crashes by writing to sysctl 3466 /proc/sys/kernel/tainted any flagset matching with the 3467 bitmask set on panic_on_taint. 3468 See Documentation/admin-guide/tainted-kernels.rst for 3469 extra details on the taint flags that users can pick 3470 to compose the bitmask to assign to panic_on_taint. 3471 3472 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3473 on a WARN(). 3474 3475 crash_kexec_post_notifiers 3476 Run kdump after running panic-notifiers and dumping 3477 kmsg. This only for the users who doubt kdump always 3478 succeeds in any situation. 3479 Note that this also increases risks of kdump failure, 3480 because some panic notifiers can make the crashed 3481 kernel more unstable. 3482 3483 parkbd.port= [HW] Parallel port number the keyboard adapter is 3484 connected to, default is 0. 3485 Format: <parport#> 3486 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3487 0 for XT, 1 for AT (default is AT). 3488 Format: <mode> 3489 3490 parport= [HW,PPT] Specify parallel ports. 0 disables. 3491 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3492 Use 'auto' to force the driver to use any 3493 IRQ/DMA settings detected (the default is to 3494 ignore detected IRQ/DMA settings because of 3495 possible conflicts). You can specify the base 3496 address, IRQ, and DMA settings; IRQ and DMA 3497 should be numbers, or 'auto' (for using detected 3498 settings on that particular port), or 'nofifo' 3499 (to avoid using a FIFO even if it is detected). 3500 Parallel ports are assigned in the order they 3501 are specified on the command line, starting 3502 with parport0. 3503 3504 parport_init_mode= [HW,PPT] 3505 Configure VIA parallel port to operate in 3506 a specific mode. This is necessary on Pegasos 3507 computer where firmware has no options for setting 3508 up parallel port mode and sets it to spp. 3509 Currently this function knows 686a and 8231 chips. 3510 Format: [spp|ps2|epp|ecp|ecpepp] 3511 3512 pause_on_oops= 3513 Halt all CPUs after the first oops has been printed for 3514 the specified number of seconds. This is to be used if 3515 your oopses keep scrolling off the screen. 3516 3517 pcbit= [HW,ISDN] 3518 3519 pcd. [PARIDE] 3520 See header of drivers/block/paride/pcd.c. 3521 See also Documentation/admin-guide/blockdev/paride.rst. 3522 3523 pci=option[,option...] [PCI] various PCI subsystem options. 3524 3525 Some options herein operate on a specific device 3526 or a set of devices (<pci_dev>). These are 3527 specified in one of the following formats: 3528 3529 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3530 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3531 3532 Note: the first format specifies a PCI 3533 bus/device/function address which may change 3534 if new hardware is inserted, if motherboard 3535 firmware changes, or due to changes caused 3536 by other kernel parameters. If the 3537 domain is left unspecified, it is 3538 taken to be zero. Optionally, a path 3539 to a device through multiple device/function 3540 addresses can be specified after the base 3541 address (this is more robust against 3542 renumbering issues). The second format 3543 selects devices using IDs from the 3544 configuration space which may match multiple 3545 devices in the system. 3546 3547 earlydump dump PCI config space before the kernel 3548 changes anything 3549 off [X86] don't probe for the PCI bus 3550 bios [X86-32] force use of PCI BIOS, don't access 3551 the hardware directly. Use this if your machine 3552 has a non-standard PCI host bridge. 3553 nobios [X86-32] disallow use of PCI BIOS, only direct 3554 hardware access methods are allowed. Use this 3555 if you experience crashes upon bootup and you 3556 suspect they are caused by the BIOS. 3557 conf1 [X86] Force use of PCI Configuration Access 3558 Mechanism 1 (config address in IO port 0xCF8, 3559 data in IO port 0xCFC, both 32-bit). 3560 conf2 [X86] Force use of PCI Configuration Access 3561 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3562 the function, IO port 0xCFA, also 8-bit, sets 3563 bus number. The config space is then accessed 3564 through ports 0xC000-0xCFFF). 3565 See http://wiki.osdev.org/PCI for more info 3566 on the configuration access mechanisms. 3567 noaer [PCIE] If the PCIEAER kernel config parameter is 3568 enabled, this kernel boot option can be used to 3569 disable the use of PCIE advanced error reporting. 3570 nodomains [PCI] Disable support for multiple PCI 3571 root domains (aka PCI segments, in ACPI-speak). 3572 nommconf [X86] Disable use of MMCONFIG for PCI 3573 Configuration 3574 check_enable_amd_mmconf [X86] check for and enable 3575 properly configured MMIO access to PCI 3576 config space on AMD family 10h CPU 3577 nomsi [MSI] If the PCI_MSI kernel config parameter is 3578 enabled, this kernel boot option can be used to 3579 disable the use of MSI interrupts system-wide. 3580 noioapicquirk [APIC] Disable all boot interrupt quirks. 3581 Safety option to keep boot IRQs enabled. This 3582 should never be necessary. 3583 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3584 primary IO-APIC for bridges that cannot disable 3585 boot IRQs. This fixes a source of spurious IRQs 3586 when the system masks IRQs. 3587 noioapicreroute [APIC] Disable workaround that uses the 3588 boot IRQ equivalent of an IRQ that connects to 3589 a chipset where boot IRQs cannot be disabled. 3590 The opposite of ioapicreroute. 3591 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3592 routing table. These calls are known to be buggy 3593 on several machines and they hang the machine 3594 when used, but on other computers it's the only 3595 way to get the interrupt routing table. Try 3596 this option if the kernel is unable to allocate 3597 IRQs or discover secondary PCI buses on your 3598 motherboard. 3599 rom [X86] Assign address space to expansion ROMs. 3600 Use with caution as certain devices share 3601 address decoders between ROMs and other 3602 resources. 3603 norom [X86] Do not assign address space to 3604 expansion ROMs that do not already have 3605 BIOS assigned address ranges. 3606 nobar [X86] Do not assign address space to the 3607 BARs that weren't assigned by the BIOS. 3608 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3609 assigned automatically to PCI devices. You can 3610 make the kernel exclude IRQs of your ISA cards 3611 this way. 3612 pirqaddr=0xAAAAA [X86] Specify the physical address 3613 of the PIRQ table (normally generated 3614 by the BIOS) if it is outside the 3615 F0000h-100000h range. 3616 lastbus=N [X86] Scan all buses thru bus #N. Can be 3617 useful if the kernel is unable to find your 3618 secondary buses and you want to tell it 3619 explicitly which ones they are. 3620 assign-busses [X86] Always assign all PCI bus 3621 numbers ourselves, overriding 3622 whatever the firmware may have done. 3623 usepirqmask [X86] Honor the possible IRQ mask stored 3624 in the BIOS $PIR table. This is needed on 3625 some systems with broken BIOSes, notably 3626 some HP Pavilion N5400 and Omnibook XE3 3627 notebooks. This will have no effect if ACPI 3628 IRQ routing is enabled. 3629 noacpi [X86] Do not use ACPI for IRQ routing 3630 or for PCI scanning. 3631 use_crs [X86] Use PCI host bridge window information 3632 from ACPI. On BIOSes from 2008 or later, this 3633 is enabled by default. If you need to use this, 3634 please report a bug. 3635 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3636 If you need to use this, please report a bug. 3637 routeirq Do IRQ routing for all PCI devices. 3638 This is normally done in pci_enable_device(), 3639 so this option is a temporary workaround 3640 for broken drivers that don't call it. 3641 skip_isa_align [X86] do not align io start addr, so can 3642 handle more pci cards 3643 noearly [X86] Don't do any early type 1 scanning. 3644 This might help on some broken boards which 3645 machine check when some devices' config space 3646 is read. But various workarounds are disabled 3647 and some IOMMU drivers will not work. 3648 bfsort Sort PCI devices into breadth-first order. 3649 This sorting is done to get a device 3650 order compatible with older (<= 2.4) kernels. 3651 nobfsort Don't sort PCI devices into breadth-first order. 3652 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3653 tuning and use the BIOS-configured MPS defaults. 3654 pcie_bus_safe Set every device's MPS to the largest value 3655 supported by all devices below the root complex. 3656 pcie_bus_perf Set device MPS to the largest allowable MPS 3657 based on its parent bus. Also set MRRS (Max 3658 Read Request Size) to the largest supported 3659 value (no larger than the MPS that the device 3660 or bus can support) for best performance. 3661 pcie_bus_peer2peer Set every device's MPS to 128B, which 3662 every device is guaranteed to support. This 3663 configuration allows peer-to-peer DMA between 3664 any pair of devices, possibly at the cost of 3665 reduced performance. This also guarantees 3666 that hot-added devices will work. 3667 cbiosize=nn[KMG] The fixed amount of bus space which is 3668 reserved for the CardBus bridge's IO window. 3669 The default value is 256 bytes. 3670 cbmemsize=nn[KMG] The fixed amount of bus space which is 3671 reserved for the CardBus bridge's memory 3672 window. The default value is 64 megabytes. 3673 resource_alignment= 3674 Format: 3675 [<order of align>@]<pci_dev>[; ...] 3676 Specifies alignment and device to reassign 3677 aligned memory resources. How to 3678 specify the device is described above. 3679 If <order of align> is not specified, 3680 PAGE_SIZE is used as alignment. 3681 A PCI-PCI bridge can be specified if resource 3682 windows need to be expanded. 3683 To specify the alignment for several 3684 instances of a device, the PCI vendor, 3685 device, subvendor, and subdevice may be 3686 specified, e.g., 12@pci:8086:9c22:103c:198f 3687 for 4096-byte alignment. 3688 ecrc= Enable/disable PCIe ECRC (transaction layer 3689 end-to-end CRC checking). 3690 bios: Use BIOS/firmware settings. This is the 3691 the default. 3692 off: Turn ECRC off 3693 on: Turn ECRC on. 3694 hpiosize=nn[KMG] The fixed amount of bus space which is 3695 reserved for hotplug bridge's IO window. 3696 Default size is 256 bytes. 3697 hpmmiosize=nn[KMG] The fixed amount of bus space which is 3698 reserved for hotplug bridge's MMIO window. 3699 Default size is 2 megabytes. 3700 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 3701 reserved for hotplug bridge's MMIO_PREF window. 3702 Default size is 2 megabytes. 3703 hpmemsize=nn[KMG] The fixed amount of bus space which is 3704 reserved for hotplug bridge's MMIO and 3705 MMIO_PREF window. 3706 Default size is 2 megabytes. 3707 hpbussize=nn The minimum amount of additional bus numbers 3708 reserved for buses below a hotplug bridge. 3709 Default is 1. 3710 realloc= Enable/disable reallocating PCI bridge resources 3711 if allocations done by BIOS are too small to 3712 accommodate resources required by all child 3713 devices. 3714 off: Turn realloc off 3715 on: Turn realloc on 3716 realloc same as realloc=on 3717 noari do not use PCIe ARI. 3718 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3719 do not use PCIe ATS (and IOMMU device IOTLB). 3720 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3721 only look for one device below a PCIe downstream 3722 port. 3723 big_root_window Try to add a big 64bit memory window to the PCIe 3724 root complex on AMD CPUs. Some GFX hardware 3725 can resize a BAR to allow access to all VRAM. 3726 Adding the window is slightly risky (it may 3727 conflict with unreported devices), so this 3728 taints the kernel. 3729 disable_acs_redir=<pci_dev>[; ...] 3730 Specify one or more PCI devices (in the format 3731 specified above) separated by semicolons. 3732 Each device specified will have the PCI ACS 3733 redirect capabilities forced off which will 3734 allow P2P traffic between devices through 3735 bridges without forcing it upstream. Note: 3736 this removes isolation between devices and 3737 may put more devices in an IOMMU group. 3738 force_floating [S390] Force usage of floating interrupts. 3739 nomio [S390] Do not use MIO instructions. 3740 norid [S390] ignore the RID field and force use of 3741 one PCI domain per PCI function 3742 3743 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3744 Management. 3745 off Disable ASPM. 3746 force Enable ASPM even on devices that claim not to support it. 3747 WARNING: Forcing ASPM on may cause system lockups. 3748 3749 pcie_ports= [PCIE] PCIe port services handling: 3750 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3751 even if the platform doesn't give the OS permission to 3752 use them. This may cause conflicts if the platform 3753 also tries to use these services. 3754 dpc-native Use native PCIe service for DPC only. May 3755 cause conflicts if firmware uses AER or DPC. 3756 compat Disable native PCIe services (PME, AER, DPC, PCIe 3757 hotplug). 3758 3759 pcie_port_pm= [PCIE] PCIe port power management handling: 3760 off Disable power management of all PCIe ports 3761 force Forcibly enable power management of all PCIe ports 3762 3763 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 3764 nomsi Do not use MSI for native PCIe PME signaling (this makes 3765 all PCIe root ports use INTx for all services). 3766 3767 pcmv= [HW,PCMCIA] BadgePAD 4 3768 3769 pd_ignore_unused 3770 [PM] 3771 Keep all power-domains already enabled by bootloader on, 3772 even if no driver has claimed them. This is useful 3773 for debug and development, but should not be 3774 needed on a platform with proper driver support. 3775 3776 pd. [PARIDE] 3777 See Documentation/admin-guide/blockdev/paride.rst. 3778 3779 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 3780 boot time. 3781 Format: { 0 | 1 } 3782 See arch/parisc/kernel/pdc_chassis.c 3783 3784 percpu_alloc= Select which percpu first chunk allocator to use. 3785 Currently supported values are "embed" and "page". 3786 Archs may support subset or none of the selections. 3787 See comments in mm/percpu.c for details on each 3788 allocator. This parameter is primarily for debugging 3789 and performance comparison. 3790 3791 pf. [PARIDE] 3792 See Documentation/admin-guide/blockdev/paride.rst. 3793 3794 pg. [PARIDE] 3795 See Documentation/admin-guide/blockdev/paride.rst. 3796 3797 pirq= [SMP,APIC] Manual mp-table setup 3798 See Documentation/x86/i386/IO-APIC.rst. 3799 3800 plip= [PPT,NET] Parallel port network link 3801 Format: { parport<nr> | timid | 0 } 3802 See also Documentation/admin-guide/parport.rst. 3803 3804 pmtmr= [X86] Manual setup of pmtmr I/O Port. 3805 Override pmtimer IOPort with a hex value. 3806 e.g. pmtmr=0x508 3807 3808 pm_debug_messages [SUSPEND,KNL] 3809 Enable suspend/resume debug messages during boot up. 3810 3811 pnp.debug=1 [PNP] 3812 Enable PNP debug messages (depends on the 3813 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 3814 via /sys/module/pnp/parameters/debug. We always show 3815 current resource usage; turning this on also shows 3816 possible settings and some assignment information. 3817 3818 pnpacpi= [ACPI] 3819 { off } 3820 3821 pnpbios= [ISAPNP] 3822 { on | off | curr | res | no-curr | no-res } 3823 3824 pnp_reserve_irq= 3825 [ISAPNP] Exclude IRQs for the autoconfiguration 3826 3827 pnp_reserve_dma= 3828 [ISAPNP] Exclude DMAs for the autoconfiguration 3829 3830 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 3831 Ranges are in pairs (I/O port base and size). 3832 3833 pnp_reserve_mem= 3834 [ISAPNP] Exclude memory regions for the 3835 autoconfiguration. 3836 Ranges are in pairs (memory base and size). 3837 3838 ports= [IP_VS_FTP] IPVS ftp helper module 3839 Default is 21. 3840 Up to 8 (IP_VS_APP_MAX_PORTS) ports 3841 may be specified. 3842 Format: <port>,<port>.... 3843 3844 powersave=off [PPC] This option disables power saving features. 3845 It specifically disables cpuidle and sets the 3846 platform machine description specific power_save 3847 function to NULL. On Idle the CPU just reduces 3848 execution priority. 3849 3850 ppc_strict_facility_enable 3851 [PPC] This option catches any kernel floating point, 3852 Altivec, VSX and SPE outside of regions specifically 3853 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 3854 There is some performance impact when enabling this. 3855 3856 ppc_tm= [PPC] 3857 Format: {"off"} 3858 Disable Hardware Transactional Memory 3859 3860 print-fatal-signals= 3861 [KNL] debug: print fatal signals 3862 3863 If enabled, warn about various signal handling 3864 related application anomalies: too many signals, 3865 too many POSIX.1 timers, fatal signals causing a 3866 coredump - etc. 3867 3868 If you hit the warning due to signal overflow, 3869 you might want to try "ulimit -i unlimited". 3870 3871 default: off. 3872 3873 printk.always_kmsg_dump= 3874 Trigger kmsg_dump for cases other than kernel oops or 3875 panics 3876 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3877 default: disabled 3878 3879 printk.devkmsg={on,off,ratelimit} 3880 Control writing to /dev/kmsg. 3881 on - unlimited logging to /dev/kmsg from userspace 3882 off - logging to /dev/kmsg disabled 3883 ratelimit - ratelimit the logging 3884 Default: ratelimit 3885 3886 printk.time= Show timing data prefixed to each printk message line 3887 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 3888 3889 processor.max_cstate= [HW,ACPI] 3890 Limit processor to maximum C-state 3891 max_cstate=9 overrides any DMI blacklist limit. 3892 3893 processor.nocst [HW,ACPI] 3894 Ignore the _CST method to determine C-states, 3895 instead using the legacy FADT method 3896 3897 profile= [KNL] Enable kernel profiling via /proc/profile 3898 Format: [<profiletype>,]<number> 3899 Param: <profiletype>: "schedule", "sleep", or "kvm" 3900 [defaults to kernel profiling] 3901 Param: "schedule" - profile schedule points. 3902 Param: "sleep" - profile D-state sleeping (millisecs). 3903 Requires CONFIG_SCHEDSTATS 3904 Param: "kvm" - profile VM exits. 3905 Param: <number> - step/bucket size as a power of 2 for 3906 statistical time based profiling. 3907 3908 prompt_ramdisk= [RAM] List of RAM disks to prompt for floppy disk 3909 before loading. 3910 See Documentation/admin-guide/blockdev/ramdisk.rst. 3911 3912 prot_virt= [S390] enable hosting protected virtual machines 3913 isolated from the hypervisor (if hardware supports 3914 that). 3915 Format: <bool> 3916 3917 psi= [KNL] Enable or disable pressure stall information 3918 tracking. 3919 Format: <bool> 3920 3921 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 3922 probe for; one of (bare|imps|exps|lifebook|any). 3923 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 3924 per second. 3925 psmouse.resetafter= [HW,MOUSE] 3926 Try to reset the device after so many bad packets 3927 (0 = never). 3928 psmouse.resolution= 3929 [HW,MOUSE] Set desired mouse resolution, in dpi. 3930 psmouse.smartscroll= 3931 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 3932 0 = disabled, 1 = enabled (default). 3933 3934 pstore.backend= Specify the name of the pstore backend to use 3935 3936 pt. [PARIDE] 3937 See Documentation/admin-guide/blockdev/paride.rst. 3938 3939 pti= [X86_64] Control Page Table Isolation of user and 3940 kernel address spaces. Disabling this feature 3941 removes hardening, but improves performance of 3942 system calls and interrupts. 3943 3944 on - unconditionally enable 3945 off - unconditionally disable 3946 auto - kernel detects whether your CPU model is 3947 vulnerable to issues that PTI mitigates 3948 3949 Not specifying this option is equivalent to pti=auto. 3950 3951 nopti [X86_64] 3952 Equivalent to pti=off 3953 3954 pty.legacy_count= 3955 [KNL] Number of legacy pty's. Overwrites compiled-in 3956 default number. 3957 3958 quiet [KNL] Disable most log messages 3959 3960 r128= [HW,DRM] 3961 3962 raid= [HW,RAID] 3963 See Documentation/admin-guide/md.rst. 3964 3965 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 3966 See Documentation/admin-guide/blockdev/ramdisk.rst. 3967 3968 random.trust_cpu={on,off} 3969 [KNL] Enable or disable trusting the use of the 3970 CPU's random number generator (if available) to 3971 fully seed the kernel's CRNG. Default is controlled 3972 by CONFIG_RANDOM_TRUST_CPU. 3973 3974 ras=option[,option,...] [KNL] RAS-specific options 3975 3976 cec_disable [X86] 3977 Disable the Correctable Errors Collector, 3978 see CONFIG_RAS_CEC help text. 3979 3980 rcu_nocbs= [KNL] 3981 The argument is a cpu list, as described above, 3982 except that the string "all" can be used to 3983 specify every CPU on the system. 3984 3985 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 3986 the specified list of CPUs to be no-callback CPUs. 3987 Invocation of these CPUs' RCU callbacks will be 3988 offloaded to "rcuox/N" kthreads created for that 3989 purpose, where "x" is "p" for RCU-preempt, and 3990 "s" for RCU-sched, and "N" is the CPU number. 3991 This reduces OS jitter on the offloaded CPUs, 3992 which can be useful for HPC and real-time 3993 workloads. It can also improve energy efficiency 3994 for asymmetric multiprocessors. 3995 3996 rcu_nocb_poll [KNL] 3997 Rather than requiring that offloaded CPUs 3998 (specified by rcu_nocbs= above) explicitly 3999 awaken the corresponding "rcuoN" kthreads, 4000 make these kthreads poll for callbacks. 4001 This improves the real-time response for the 4002 offloaded CPUs by relieving them of the need to 4003 wake up the corresponding kthread, but degrades 4004 energy efficiency by requiring that the kthreads 4005 periodically wake up to do the polling. 4006 4007 rcutree.blimit= [KNL] 4008 Set maximum number of finished RCU callbacks to 4009 process in one batch. 4010 4011 rcutree.dump_tree= [KNL] 4012 Dump the structure of the rcu_node combining tree 4013 out at early boot. This is used for diagnostic 4014 purposes, to verify correct tree setup. 4015 4016 rcutree.gp_cleanup_delay= [KNL] 4017 Set the number of jiffies to delay each step of 4018 RCU grace-period cleanup. 4019 4020 rcutree.gp_init_delay= [KNL] 4021 Set the number of jiffies to delay each step of 4022 RCU grace-period initialization. 4023 4024 rcutree.gp_preinit_delay= [KNL] 4025 Set the number of jiffies to delay each step of 4026 RCU grace-period pre-initialization, that is, 4027 the propagation of recent CPU-hotplug changes up 4028 the rcu_node combining tree. 4029 4030 rcutree.use_softirq= [KNL] 4031 If set to zero, move all RCU_SOFTIRQ processing to 4032 per-CPU rcuc kthreads. Defaults to a non-zero 4033 value, meaning that RCU_SOFTIRQ is used by default. 4034 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4035 4036 rcutree.rcu_fanout_exact= [KNL] 4037 Disable autobalancing of the rcu_node combining 4038 tree. This is used by rcutorture, and might 4039 possibly be useful for architectures having high 4040 cache-to-cache transfer latencies. 4041 4042 rcutree.rcu_fanout_leaf= [KNL] 4043 Change the number of CPUs assigned to each 4044 leaf rcu_node structure. Useful for very 4045 large systems, which will choose the value 64, 4046 and for NUMA systems with large remote-access 4047 latencies, which will choose a value aligned 4048 with the appropriate hardware boundaries. 4049 4050 rcutree.rcu_min_cached_objs= [KNL] 4051 Minimum number of objects which are cached and 4052 maintained per one CPU. Object size is equal 4053 to PAGE_SIZE. The cache allows to reduce the 4054 pressure to page allocator, also it makes the 4055 whole algorithm to behave better in low memory 4056 condition. 4057 4058 rcutree.jiffies_till_first_fqs= [KNL] 4059 Set delay from grace-period initialization to 4060 first attempt to force quiescent states. 4061 Units are jiffies, minimum value is zero, 4062 and maximum value is HZ. 4063 4064 rcutree.jiffies_till_next_fqs= [KNL] 4065 Set delay between subsequent attempts to force 4066 quiescent states. Units are jiffies, minimum 4067 value is one, and maximum value is HZ. 4068 4069 rcutree.jiffies_till_sched_qs= [KNL] 4070 Set required age in jiffies for a 4071 given grace period before RCU starts 4072 soliciting quiescent-state help from 4073 rcu_note_context_switch() and cond_resched(). 4074 If not specified, the kernel will calculate 4075 a value based on the most recent settings 4076 of rcutree.jiffies_till_first_fqs 4077 and rcutree.jiffies_till_next_fqs. 4078 This calculated value may be viewed in 4079 rcutree.jiffies_to_sched_qs. Any attempt to set 4080 rcutree.jiffies_to_sched_qs will be cheerfully 4081 overwritten. 4082 4083 rcutree.kthread_prio= [KNL,BOOT] 4084 Set the SCHED_FIFO priority of the RCU per-CPU 4085 kthreads (rcuc/N). This value is also used for 4086 the priority of the RCU boost threads (rcub/N) 4087 and for the RCU grace-period kthreads (rcu_bh, 4088 rcu_preempt, and rcu_sched). If RCU_BOOST is 4089 set, valid values are 1-99 and the default is 1 4090 (the least-favored priority). Otherwise, when 4091 RCU_BOOST is not set, valid values are 0-99 and 4092 the default is zero (non-realtime operation). 4093 4094 rcutree.rcu_nocb_gp_stride= [KNL] 4095 Set the number of NOCB callback kthreads in 4096 each group, which defaults to the square root 4097 of the number of CPUs. Larger numbers reduce 4098 the wakeup overhead on the global grace-period 4099 kthread, but increases that same overhead on 4100 each group's NOCB grace-period kthread. 4101 4102 rcutree.qhimark= [KNL] 4103 Set threshold of queued RCU callbacks beyond which 4104 batch limiting is disabled. 4105 4106 rcutree.qlowmark= [KNL] 4107 Set threshold of queued RCU callbacks below which 4108 batch limiting is re-enabled. 4109 4110 rcutree.qovld= [KNL] 4111 Set threshold of queued RCU callbacks beyond which 4112 RCU's force-quiescent-state scan will aggressively 4113 enlist help from cond_resched() and sched IPIs to 4114 help CPUs more quickly reach quiescent states. 4115 Set to less than zero to make this be set based 4116 on rcutree.qhimark at boot time and to zero to 4117 disable more aggressive help enlistment. 4118 4119 rcutree.rcu_idle_gp_delay= [KNL] 4120 Set wakeup interval for idle CPUs that have 4121 RCU callbacks (RCU_FAST_NO_HZ=y). 4122 4123 rcutree.rcu_idle_lazy_gp_delay= [KNL] 4124 Set wakeup interval for idle CPUs that have 4125 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y). 4126 Lazy RCU callbacks are those which RCU can 4127 prove do nothing more than free memory. 4128 4129 rcutree.rcu_kick_kthreads= [KNL] 4130 Cause the grace-period kthread to get an extra 4131 wake_up() if it sleeps three times longer than 4132 it should at force-quiescent-state time. 4133 This wake_up() will be accompanied by a 4134 WARN_ONCE() splat and an ftrace_dump(). 4135 4136 rcutree.sysrq_rcu= [KNL] 4137 Commandeer a sysrq key to dump out Tree RCU's 4138 rcu_node tree with an eye towards determining 4139 why a new grace period has not yet started. 4140 4141 rcuperf.gp_async= [KNL] 4142 Measure performance of asynchronous 4143 grace-period primitives such as call_rcu(). 4144 4145 rcuperf.gp_async_max= [KNL] 4146 Specify the maximum number of outstanding 4147 callbacks per writer thread. When a writer 4148 thread exceeds this limit, it invokes the 4149 corresponding flavor of rcu_barrier() to allow 4150 previously posted callbacks to drain. 4151 4152 rcuperf.gp_exp= [KNL] 4153 Measure performance of expedited synchronous 4154 grace-period primitives. 4155 4156 rcuperf.holdoff= [KNL] 4157 Set test-start holdoff period. The purpose of 4158 this parameter is to delay the start of the 4159 test until boot completes in order to avoid 4160 interference. 4161 4162 rcuperf.kfree_rcu_test= [KNL] 4163 Set to measure performance of kfree_rcu() flooding. 4164 4165 rcuperf.kfree_nthreads= [KNL] 4166 The number of threads running loops of kfree_rcu(). 4167 4168 rcuperf.kfree_alloc_num= [KNL] 4169 Number of allocations and frees done in an iteration. 4170 4171 rcuperf.kfree_loops= [KNL] 4172 Number of loops doing rcuperf.kfree_alloc_num number 4173 of allocations and frees. 4174 4175 rcuperf.nreaders= [KNL] 4176 Set number of RCU readers. The value -1 selects 4177 N, where N is the number of CPUs. A value 4178 "n" less than -1 selects N-n+1, where N is again 4179 the number of CPUs. For example, -2 selects N 4180 (the number of CPUs), -3 selects N+1, and so on. 4181 A value of "n" less than or equal to -N selects 4182 a single reader. 4183 4184 rcuperf.nwriters= [KNL] 4185 Set number of RCU writers. The values operate 4186 the same as for rcuperf.nreaders. 4187 N, where N is the number of CPUs 4188 4189 rcuperf.perf_type= [KNL] 4190 Specify the RCU implementation to test. 4191 4192 rcuperf.shutdown= [KNL] 4193 Shut the system down after performance tests 4194 complete. This is useful for hands-off automated 4195 testing. 4196 4197 rcuperf.verbose= [KNL] 4198 Enable additional printk() statements. 4199 4200 rcuperf.writer_holdoff= [KNL] 4201 Write-side holdoff between grace periods, 4202 in microseconds. The default of zero says 4203 no holdoff. 4204 4205 rcutorture.fqs_duration= [KNL] 4206 Set duration of force_quiescent_state bursts 4207 in microseconds. 4208 4209 rcutorture.fqs_holdoff= [KNL] 4210 Set holdoff time within force_quiescent_state bursts 4211 in microseconds. 4212 4213 rcutorture.fqs_stutter= [KNL] 4214 Set wait time between force_quiescent_state bursts 4215 in seconds. 4216 4217 rcutorture.fwd_progress= [KNL] 4218 Enable RCU grace-period forward-progress testing 4219 for the types of RCU supporting this notion. 4220 4221 rcutorture.fwd_progress_div= [KNL] 4222 Specify the fraction of a CPU-stall-warning 4223 period to do tight-loop forward-progress testing. 4224 4225 rcutorture.fwd_progress_holdoff= [KNL] 4226 Number of seconds to wait between successive 4227 forward-progress tests. 4228 4229 rcutorture.fwd_progress_need_resched= [KNL] 4230 Enclose cond_resched() calls within checks for 4231 need_resched() during tight-loop forward-progress 4232 testing. 4233 4234 rcutorture.gp_cond= [KNL] 4235 Use conditional/asynchronous update-side 4236 primitives, if available. 4237 4238 rcutorture.gp_exp= [KNL] 4239 Use expedited update-side primitives, if available. 4240 4241 rcutorture.gp_normal= [KNL] 4242 Use normal (non-expedited) asynchronous 4243 update-side primitives, if available. 4244 4245 rcutorture.gp_sync= [KNL] 4246 Use normal (non-expedited) synchronous 4247 update-side primitives, if available. If all 4248 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4249 rcutorture.gp_normal=, and rcutorture.gp_sync= 4250 are zero, rcutorture acts as if is interpreted 4251 they are all non-zero. 4252 4253 rcutorture.n_barrier_cbs= [KNL] 4254 Set callbacks/threads for rcu_barrier() testing. 4255 4256 rcutorture.nfakewriters= [KNL] 4257 Set number of concurrent RCU writers. These just 4258 stress RCU, they don't participate in the actual 4259 test, hence the "fake". 4260 4261 rcutorture.nreaders= [KNL] 4262 Set number of RCU readers. The value -1 selects 4263 N-1, where N is the number of CPUs. A value 4264 "n" less than -1 selects N-n-2, where N is again 4265 the number of CPUs. For example, -2 selects N 4266 (the number of CPUs), -3 selects N+1, and so on. 4267 4268 rcutorture.object_debug= [KNL] 4269 Enable debug-object double-call_rcu() testing. 4270 4271 rcutorture.onoff_holdoff= [KNL] 4272 Set time (s) after boot for CPU-hotplug testing. 4273 4274 rcutorture.onoff_interval= [KNL] 4275 Set time (jiffies) between CPU-hotplug operations, 4276 or zero to disable CPU-hotplug testing. 4277 4278 rcutorture.read_exit= [KNL] 4279 Set the number of read-then-exit kthreads used 4280 to test the interaction of RCU updaters and 4281 task-exit processing. 4282 4283 rcutorture.read_exit_burst= [KNL] 4284 The number of times in a given read-then-exit 4285 episode that a set of read-then-exit kthreads 4286 is spawned. 4287 4288 rcutorture.read_exit_delay= [KNL] 4289 The delay, in seconds, between successive 4290 read-then-exit testing episodes. 4291 4292 rcutorture.shuffle_interval= [KNL] 4293 Set task-shuffle interval (s). Shuffling tasks 4294 allows some CPUs to go into dyntick-idle mode 4295 during the rcutorture test. 4296 4297 rcutorture.shutdown_secs= [KNL] 4298 Set time (s) after boot system shutdown. This 4299 is useful for hands-off automated testing. 4300 4301 rcutorture.stall_cpu= [KNL] 4302 Duration of CPU stall (s) to test RCU CPU stall 4303 warnings, zero to disable. 4304 4305 rcutorture.stall_cpu_block= [KNL] 4306 Sleep while stalling if set. This will result 4307 in warnings from preemptible RCU in addition 4308 to any other stall-related activity. 4309 4310 rcutorture.stall_cpu_holdoff= [KNL] 4311 Time to wait (s) after boot before inducing stall. 4312 4313 rcutorture.stall_cpu_irqsoff= [KNL] 4314 Disable interrupts while stalling if set. 4315 4316 rcutorture.stall_gp_kthread= [KNL] 4317 Duration (s) of forced sleep within RCU 4318 grace-period kthread to test RCU CPU stall 4319 warnings, zero to disable. If both stall_cpu 4320 and stall_gp_kthread are specified, the 4321 kthread is starved first, then the CPU. 4322 4323 rcutorture.stat_interval= [KNL] 4324 Time (s) between statistics printk()s. 4325 4326 rcutorture.stutter= [KNL] 4327 Time (s) to stutter testing, for example, specifying 4328 five seconds causes the test to run for five seconds, 4329 wait for five seconds, and so on. This tests RCU's 4330 ability to transition abruptly to and from idle. 4331 4332 rcutorture.test_boost= [KNL] 4333 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4334 "Maybe" means test if the RCU implementation 4335 under test support RCU priority boosting. 4336 4337 rcutorture.test_boost_duration= [KNL] 4338 Duration (s) of each individual boost test. 4339 4340 rcutorture.test_boost_interval= [KNL] 4341 Interval (s) between each boost test. 4342 4343 rcutorture.test_no_idle_hz= [KNL] 4344 Test RCU's dyntick-idle handling. See also the 4345 rcutorture.shuffle_interval parameter. 4346 4347 rcutorture.torture_type= [KNL] 4348 Specify the RCU implementation to test. 4349 4350 rcutorture.verbose= [KNL] 4351 Enable additional printk() statements. 4352 4353 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4354 Dump ftrace buffer after reporting RCU CPU 4355 stall warning. 4356 4357 rcupdate.rcu_cpu_stall_suppress= [KNL] 4358 Suppress RCU CPU stall warning messages. 4359 4360 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 4361 Suppress RCU CPU stall warning messages and 4362 rcutorture writer stall warnings that occur 4363 during early boot, that is, during the time 4364 before the init task is spawned. 4365 4366 rcupdate.rcu_cpu_stall_timeout= [KNL] 4367 Set timeout for RCU CPU stall warning messages. 4368 4369 rcupdate.rcu_expedited= [KNL] 4370 Use expedited grace-period primitives, for 4371 example, synchronize_rcu_expedited() instead 4372 of synchronize_rcu(). This reduces latency, 4373 but can increase CPU utilization, degrade 4374 real-time latency, and degrade energy efficiency. 4375 No effect on CONFIG_TINY_RCU kernels. 4376 4377 rcupdate.rcu_normal= [KNL] 4378 Use only normal grace-period primitives, 4379 for example, synchronize_rcu() instead of 4380 synchronize_rcu_expedited(). This improves 4381 real-time latency, CPU utilization, and 4382 energy efficiency, but can expose users to 4383 increased grace-period latency. This parameter 4384 overrides rcupdate.rcu_expedited. No effect on 4385 CONFIG_TINY_RCU kernels. 4386 4387 rcupdate.rcu_normal_after_boot= [KNL] 4388 Once boot has completed (that is, after 4389 rcu_end_inkernel_boot() has been invoked), use 4390 only normal grace-period primitives. No effect 4391 on CONFIG_TINY_RCU kernels. 4392 4393 rcupdate.rcu_task_ipi_delay= [KNL] 4394 Set time in jiffies during which RCU tasks will 4395 avoid sending IPIs, starting with the beginning 4396 of a given grace period. Setting a large 4397 number avoids disturbing real-time workloads, 4398 but lengthens grace periods. 4399 4400 rcupdate.rcu_task_stall_timeout= [KNL] 4401 Set timeout in jiffies for RCU task stall warning 4402 messages. Disable with a value less than or equal 4403 to zero. 4404 4405 rcupdate.rcu_self_test= [KNL] 4406 Run the RCU early boot self tests 4407 4408 rdinit= [KNL] 4409 Format: <full_path> 4410 Run specified binary instead of /init from the ramdisk, 4411 used for early userspace startup. See initrd. 4412 4413 rdrand= [X86] 4414 force - Override the decision by the kernel to hide the 4415 advertisement of RDRAND support (this affects 4416 certain AMD processors because of buggy BIOS 4417 support, specifically around the suspend/resume 4418 path). 4419 4420 rdt= [HW,X86,RDT] 4421 Turn on/off individual RDT features. List is: 4422 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4423 mba. 4424 E.g. to turn on cmt and turn off mba use: 4425 rdt=cmt,!mba 4426 4427 reboot= [KNL] 4428 Format (x86 or x86_64): 4429 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4430 [[,]s[mp]#### \ 4431 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4432 [[,]f[orce] 4433 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4434 (prefix with 'panic_' to set mode for panic 4435 reboot only), 4436 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4437 reboot_force is either force or not specified, 4438 reboot_cpu is s[mp]#### with #### being the processor 4439 to be used for rebooting. 4440 4441 refscale.holdoff= [KNL] 4442 Set test-start holdoff period. The purpose of 4443 this parameter is to delay the start of the 4444 test until boot completes in order to avoid 4445 interference. 4446 4447 refscale.loops= [KNL] 4448 Set the number of loops over the synchronization 4449 primitive under test. Increasing this number 4450 reduces noise due to loop start/end overhead, 4451 but the default has already reduced the per-pass 4452 noise to a handful of picoseconds on ca. 2020 4453 x86 laptops. 4454 4455 refscale.nreaders= [KNL] 4456 Set number of readers. The default value of -1 4457 selects N, where N is roughly 75% of the number 4458 of CPUs. A value of zero is an interesting choice. 4459 4460 refscale.nruns= [KNL] 4461 Set number of runs, each of which is dumped onto 4462 the console log. 4463 4464 refscale.readdelay= [KNL] 4465 Set the read-side critical-section duration, 4466 measured in microseconds. 4467 4468 refscale.scale_type= [KNL] 4469 Specify the read-protection implementation to test. 4470 4471 refscale.shutdown= [KNL] 4472 Shut down the system at the end of the performance 4473 test. This defaults to 1 (shut it down) when 4474 rcuperf is built into the kernel and to 0 (leave 4475 it running) when rcuperf is built as a module. 4476 4477 refscale.verbose= [KNL] 4478 Enable additional printk() statements. 4479 4480 relax_domain_level= 4481 [KNL, SMP] Set scheduler's default relax_domain_level. 4482 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4483 4484 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4485 Format: <base1>,<size1>[,<base2>,<size2>,...] 4486 Reserve I/O ports or memory so the kernel won't use 4487 them. If <base> is less than 0x10000, the region 4488 is assumed to be I/O ports; otherwise it is memory. 4489 4490 reservetop= [X86-32] 4491 Format: nn[KMG] 4492 Reserves a hole at the top of the kernel virtual 4493 address space. 4494 4495 reservelow= [X86] 4496 Format: nn[K] 4497 Set the amount of memory to reserve for BIOS at 4498 the bottom of the address space. 4499 4500 reset_devices [KNL] Force drivers to reset the underlying device 4501 during initialization. 4502 4503 resume= [SWSUSP] 4504 Specify the partition device for software suspend 4505 Format: 4506 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4507 4508 resume_offset= [SWSUSP] 4509 Specify the offset from the beginning of the partition 4510 given by "resume=" at which the swap header is located, 4511 in <PAGE_SIZE> units (needed only for swap files). 4512 See Documentation/power/swsusp-and-swap-files.rst 4513 4514 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4515 read the resume files 4516 4517 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4518 Useful for devices that are detected asynchronously 4519 (e.g. USB and MMC devices). 4520 4521 hibernate= [HIBERNATION] 4522 noresume Don't check if there's a hibernation image 4523 present during boot. 4524 nocompress Don't compress/decompress hibernation images. 4525 no Disable hibernation and resume. 4526 protect_image Turn on image protection during restoration 4527 (that will set all pages holding image data 4528 during restoration read-only). 4529 4530 retain_initrd [RAM] Keep initrd memory after extraction 4531 4532 rfkill.default_state= 4533 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4534 etc. communication is blocked by default. 4535 1 Unblocked. 4536 4537 rfkill.master_switch_mode= 4538 0 The "airplane mode" button does nothing. 4539 1 The "airplane mode" button toggles between everything 4540 blocked and the previous configuration. 4541 2 The "airplane mode" button toggles between everything 4542 blocked and everything unblocked. 4543 4544 rhash_entries= [KNL,NET] 4545 Set number of hash buckets for route cache 4546 4547 ring3mwait=disable 4548 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4549 CPUs. 4550 4551 ro [KNL] Mount root device read-only on boot 4552 4553 rodata= [KNL] 4554 on Mark read-only kernel memory as read-only (default). 4555 off Leave read-only kernel memory writable for debugging. 4556 4557 rockchip.usb_uart 4558 Enable the uart passthrough on the designated usb port 4559 on Rockchip SoCs. When active, the signals of the 4560 debug-uart get routed to the D+ and D- pins of the usb 4561 port and the regular usb controller gets disabled. 4562 4563 root= [KNL] Root filesystem 4564 See name_to_dev_t comment in init/do_mounts.c. 4565 4566 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4567 mount the root filesystem 4568 4569 rootflags= [KNL] Set root filesystem mount option string 4570 4571 rootfstype= [KNL] Set root filesystem type 4572 4573 rootwait [KNL] Wait (indefinitely) for root device to show up. 4574 Useful for devices that are detected asynchronously 4575 (e.g. USB and MMC devices). 4576 4577 rproc_mem=nn[KMG][@address] 4578 [KNL,ARM,CMA] Remoteproc physical memory block. 4579 Memory area to be used by remote processor image, 4580 managed by CMA. 4581 4582 rw [KNL] Mount root device read-write on boot 4583 4584 S [KNL] Run init in single mode 4585 4586 s390_iommu= [HW,S390] 4587 Set s390 IOTLB flushing mode 4588 strict 4589 With strict flushing every unmap operation will result in 4590 an IOTLB flush. Default is lazy flushing before reuse, 4591 which is faster. 4592 4593 sa1100ir [NET] 4594 See drivers/net/irda/sa1100_ir.c. 4595 4596 sbni= [NET] Granch SBNI12 leased line adapter 4597 4598 sched_debug [KNL] Enables verbose scheduler debug messages. 4599 4600 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4601 Allowed values are enable and disable. This feature 4602 incurs a small amount of overhead in the scheduler 4603 but is useful for debugging and performance tuning. 4604 4605 sched_thermal_decay_shift= 4606 [KNL, SMP] Set a decay shift for scheduler thermal 4607 pressure signal. Thermal pressure signal follows the 4608 default decay period of other scheduler pelt 4609 signals(usually 32 ms but configurable). Setting 4610 sched_thermal_decay_shift will left shift the decay 4611 period for the thermal pressure signal by the shift 4612 value. 4613 i.e. with the default pelt decay period of 32 ms 4614 sched_thermal_decay_shift thermal pressure decay pr 4615 1 64 ms 4616 2 128 ms 4617 and so on. 4618 Format: integer between 0 and 10 4619 Default is 0. 4620 4621 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 4622 xtime_lock contention on larger systems, and/or RCU lock 4623 contention on all systems with CONFIG_MAXSMP set. 4624 Format: { "0" | "1" } 4625 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 4626 1 -- enable. 4627 Note: increases power consumption, thus should only be 4628 enabled if running jitter sensitive (HPC/RT) workloads. 4629 4630 security= [SECURITY] Choose a legacy "major" security module to 4631 enable at boot. This has been deprecated by the 4632 "lsm=" parameter. 4633 4634 selinux= [SELINUX] Disable or enable SELinux at boot time. 4635 Format: { "0" | "1" } 4636 See security/selinux/Kconfig help text. 4637 0 -- disable. 4638 1 -- enable. 4639 Default value is 1. 4640 4641 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 4642 Format: { "0" | "1" } 4643 See security/apparmor/Kconfig help text 4644 0 -- disable. 4645 1 -- enable. 4646 Default value is set via kernel config option. 4647 4648 serialnumber [BUGS=X86-32] 4649 4650 shapers= [NET] 4651 Maximal number of shapers. 4652 4653 simeth= [IA-64] 4654 simscsi= 4655 4656 slram= [HW,MTD] 4657 4658 slab_nomerge [MM] 4659 Disable merging of slabs with similar size. May be 4660 necessary if there is some reason to distinguish 4661 allocs to different slabs, especially in hardened 4662 environments where the risk of heap overflows and 4663 layout control by attackers can usually be 4664 frustrated by disabling merging. This will reduce 4665 most of the exposure of a heap attack to a single 4666 cache (risks via metadata attacks are mostly 4667 unchanged). Debug options disable merging on their 4668 own. 4669 For more information see Documentation/vm/slub.rst. 4670 4671 slab_max_order= [MM, SLAB] 4672 Determines the maximum allowed order for slabs. 4673 A high setting may cause OOMs due to memory 4674 fragmentation. Defaults to 1 for systems with 4675 more than 32MB of RAM, 0 otherwise. 4676 4677 slub_debug[=options[,slabs]] [MM, SLUB] 4678 Enabling slub_debug allows one to determine the 4679 culprit if slab objects become corrupted. Enabling 4680 slub_debug can create guard zones around objects and 4681 may poison objects when not in use. Also tracks the 4682 last alloc / free. For more information see 4683 Documentation/vm/slub.rst. 4684 4685 slub_memcg_sysfs= [MM, SLUB] 4686 Determines whether to enable sysfs directories for 4687 memory cgroup sub-caches. 1 to enable, 0 to disable. 4688 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON. 4689 Enabling this can lead to a very high number of debug 4690 directories and files being created under 4691 /sys/kernel/slub. 4692 4693 slub_max_order= [MM, SLUB] 4694 Determines the maximum allowed order for slabs. 4695 A high setting may cause OOMs due to memory 4696 fragmentation. For more information see 4697 Documentation/vm/slub.rst. 4698 4699 slub_min_objects= [MM, SLUB] 4700 The minimum number of objects per slab. SLUB will 4701 increase the slab order up to slub_max_order to 4702 generate a sufficiently large slab able to contain 4703 the number of objects indicated. The higher the number 4704 of objects the smaller the overhead of tracking slabs 4705 and the less frequently locks need to be acquired. 4706 For more information see Documentation/vm/slub.rst. 4707 4708 slub_min_order= [MM, SLUB] 4709 Determines the minimum page order for slabs. Must be 4710 lower than slub_max_order. 4711 For more information see Documentation/vm/slub.rst. 4712 4713 slub_nomerge [MM, SLUB] 4714 Same with slab_nomerge. This is supported for legacy. 4715 See slab_nomerge for more information. 4716 4717 smart2= [HW] 4718 Format: <io1>[,<io2>[,...,<io8>]] 4719 4720 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 4721 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 4722 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 4723 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 4724 smsc-ircc2.ircc_irq= [HW] IRQ line 4725 smsc-ircc2.ircc_dma= [HW] DMA channel 4726 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 4727 0: Toshiba Satellite 1800 (GP data pin select) 4728 1: Fast pin select (default) 4729 2: ATC IRMode 4730 4731 smt [KNL,S390] Set the maximum number of threads (logical 4732 CPUs) to use per physical CPU on systems capable of 4733 symmetric multithreading (SMT). Will be capped to the 4734 actual hardware limit. 4735 Format: <integer> 4736 Default: -1 (no limit) 4737 4738 softlockup_panic= 4739 [KNL] Should the soft-lockup detector generate panics. 4740 Format: 0 | 1 4741 4742 A value of 1 instructs the soft-lockup detector 4743 to panic the machine when a soft-lockup occurs. It is 4744 also controlled by the kernel.softlockup_panic sysctl 4745 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 4746 respective build-time switch to that functionality. 4747 4748 softlockup_all_cpu_backtrace= 4749 [KNL] Should the soft-lockup detector generate 4750 backtraces on all cpus. 4751 Format: 0 | 1 4752 4753 sonypi.*= [HW] Sony Programmable I/O Control Device driver 4754 See Documentation/admin-guide/laptops/sonypi.rst 4755 4756 spectre_v2= [X86] Control mitigation of Spectre variant 2 4757 (indirect branch speculation) vulnerability. 4758 The default operation protects the kernel from 4759 user space attacks. 4760 4761 on - unconditionally enable, implies 4762 spectre_v2_user=on 4763 off - unconditionally disable, implies 4764 spectre_v2_user=off 4765 auto - kernel detects whether your CPU model is 4766 vulnerable 4767 4768 Selecting 'on' will, and 'auto' may, choose a 4769 mitigation method at run time according to the 4770 CPU, the available microcode, the setting of the 4771 CONFIG_RETPOLINE configuration option, and the 4772 compiler with which the kernel was built. 4773 4774 Selecting 'on' will also enable the mitigation 4775 against user space to user space task attacks. 4776 4777 Selecting 'off' will disable both the kernel and 4778 the user space protections. 4779 4780 Specific mitigations can also be selected manually: 4781 4782 retpoline - replace indirect branches 4783 retpoline,generic - google's original retpoline 4784 retpoline,amd - AMD-specific minimal thunk 4785 4786 Not specifying this option is equivalent to 4787 spectre_v2=auto. 4788 4789 spectre_v2_user= 4790 [X86] Control mitigation of Spectre variant 2 4791 (indirect branch speculation) vulnerability between 4792 user space tasks 4793 4794 on - Unconditionally enable mitigations. Is 4795 enforced by spectre_v2=on 4796 4797 off - Unconditionally disable mitigations. Is 4798 enforced by spectre_v2=off 4799 4800 prctl - Indirect branch speculation is enabled, 4801 but mitigation can be enabled via prctl 4802 per thread. The mitigation control state 4803 is inherited on fork. 4804 4805 prctl,ibpb 4806 - Like "prctl" above, but only STIBP is 4807 controlled per thread. IBPB is issued 4808 always when switching between different user 4809 space processes. 4810 4811 seccomp 4812 - Same as "prctl" above, but all seccomp 4813 threads will enable the mitigation unless 4814 they explicitly opt out. 4815 4816 seccomp,ibpb 4817 - Like "seccomp" above, but only STIBP is 4818 controlled per thread. IBPB is issued 4819 always when switching between different 4820 user space processes. 4821 4822 auto - Kernel selects the mitigation depending on 4823 the available CPU features and vulnerability. 4824 4825 Default mitigation: 4826 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 4827 4828 Not specifying this option is equivalent to 4829 spectre_v2_user=auto. 4830 4831 spec_store_bypass_disable= 4832 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 4833 (Speculative Store Bypass vulnerability) 4834 4835 Certain CPUs are vulnerable to an exploit against a 4836 a common industry wide performance optimization known 4837 as "Speculative Store Bypass" in which recent stores 4838 to the same memory location may not be observed by 4839 later loads during speculative execution. The idea 4840 is that such stores are unlikely and that they can 4841 be detected prior to instruction retirement at the 4842 end of a particular speculation execution window. 4843 4844 In vulnerable processors, the speculatively forwarded 4845 store can be used in a cache side channel attack, for 4846 example to read memory to which the attacker does not 4847 directly have access (e.g. inside sandboxed code). 4848 4849 This parameter controls whether the Speculative Store 4850 Bypass optimization is used. 4851 4852 On x86 the options are: 4853 4854 on - Unconditionally disable Speculative Store Bypass 4855 off - Unconditionally enable Speculative Store Bypass 4856 auto - Kernel detects whether the CPU model contains an 4857 implementation of Speculative Store Bypass and 4858 picks the most appropriate mitigation. If the 4859 CPU is not vulnerable, "off" is selected. If the 4860 CPU is vulnerable the default mitigation is 4861 architecture and Kconfig dependent. See below. 4862 prctl - Control Speculative Store Bypass per thread 4863 via prctl. Speculative Store Bypass is enabled 4864 for a process by default. The state of the control 4865 is inherited on fork. 4866 seccomp - Same as "prctl" above, but all seccomp threads 4867 will disable SSB unless they explicitly opt out. 4868 4869 Default mitigations: 4870 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 4871 4872 On powerpc the options are: 4873 4874 on,auto - On Power8 and Power9 insert a store-forwarding 4875 barrier on kernel entry and exit. On Power7 4876 perform a software flush on kernel entry and 4877 exit. 4878 off - No action. 4879 4880 Not specifying this option is equivalent to 4881 spec_store_bypass_disable=auto. 4882 4883 spia_io_base= [HW,MTD] 4884 spia_fio_base= 4885 spia_pedr= 4886 spia_peddr= 4887 4888 split_lock_detect= 4889 [X86] Enable split lock detection 4890 4891 When enabled (and if hardware support is present), atomic 4892 instructions that access data across cache line 4893 boundaries will result in an alignment check exception. 4894 4895 off - not enabled 4896 4897 warn - the kernel will emit rate limited warnings 4898 about applications triggering the #AC 4899 exception. This mode is the default on CPUs 4900 that supports split lock detection. 4901 4902 fatal - the kernel will send SIGBUS to applications 4903 that trigger the #AC exception. 4904 4905 If an #AC exception is hit in the kernel or in 4906 firmware (i.e. not while executing in user mode) 4907 the kernel will oops in either "warn" or "fatal" 4908 mode. 4909 4910 srbds= [X86,INTEL] 4911 Control the Special Register Buffer Data Sampling 4912 (SRBDS) mitigation. 4913 4914 Certain CPUs are vulnerable to an MDS-like 4915 exploit which can leak bits from the random 4916 number generator. 4917 4918 By default, this issue is mitigated by 4919 microcode. However, the microcode fix can cause 4920 the RDRAND and RDSEED instructions to become 4921 much slower. Among other effects, this will 4922 result in reduced throughput from /dev/urandom. 4923 4924 The microcode mitigation can be disabled with 4925 the following option: 4926 4927 off: Disable mitigation and remove 4928 performance impact to RDRAND and RDSEED 4929 4930 srcutree.counter_wrap_check [KNL] 4931 Specifies how frequently to check for 4932 grace-period sequence counter wrap for the 4933 srcu_data structure's ->srcu_gp_seq_needed field. 4934 The greater the number of bits set in this kernel 4935 parameter, the less frequently counter wrap will 4936 be checked for. Note that the bottom two bits 4937 are ignored. 4938 4939 srcutree.exp_holdoff [KNL] 4940 Specifies how many nanoseconds must elapse 4941 since the end of the last SRCU grace period for 4942 a given srcu_struct until the next normal SRCU 4943 grace period will be considered for automatic 4944 expediting. Set to zero to disable automatic 4945 expediting. 4946 4947 ssbd= [ARM64,HW] 4948 Speculative Store Bypass Disable control 4949 4950 On CPUs that are vulnerable to the Speculative 4951 Store Bypass vulnerability and offer a 4952 firmware based mitigation, this parameter 4953 indicates how the mitigation should be used: 4954 4955 force-on: Unconditionally enable mitigation for 4956 for both kernel and userspace 4957 force-off: Unconditionally disable mitigation for 4958 for both kernel and userspace 4959 kernel: Always enable mitigation in the 4960 kernel, and offer a prctl interface 4961 to allow userspace to register its 4962 interest in being mitigated too. 4963 4964 stack_guard_gap= [MM] 4965 override the default stack gap protection. The value 4966 is in page units and it defines how many pages prior 4967 to (for stacks growing down) resp. after (for stacks 4968 growing up) the main stack are reserved for no other 4969 mapping. Default value is 256 pages. 4970 4971 stacktrace [FTRACE] 4972 Enabled the stack tracer on boot up. 4973 4974 stacktrace_filter=[function-list] 4975 [FTRACE] Limit the functions that the stack tracer 4976 will trace at boot up. function-list is a comma separated 4977 list of functions. This list can be changed at run 4978 time by the stack_trace_filter file in the debugfs 4979 tracing directory. Note, this enables stack tracing 4980 and the stacktrace above is not needed. 4981 4982 sti= [PARISC,HW] 4983 Format: <num> 4984 Set the STI (builtin display/keyboard on the HP-PARISC 4985 machines) console (graphic card) which should be used 4986 as the initial boot-console. 4987 See also comment in drivers/video/console/sticore.c. 4988 4989 sti_font= [HW] 4990 See comment in drivers/video/console/sticore.c. 4991 4992 stifb= [HW] 4993 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 4994 4995 sunrpc.min_resvport= 4996 sunrpc.max_resvport= 4997 [NFS,SUNRPC] 4998 SunRPC servers often require that client requests 4999 originate from a privileged port (i.e. a port in the 5000 range 0 < portnr < 1024). 5001 An administrator who wishes to reserve some of these 5002 ports for other uses may adjust the range that the 5003 kernel's sunrpc client considers to be privileged 5004 using these two parameters to set the minimum and 5005 maximum port values. 5006 5007 sunrpc.svc_rpc_per_connection_limit= 5008 [NFS,SUNRPC] 5009 Limit the number of requests that the server will 5010 process in parallel from a single connection. 5011 The default value is 0 (no limit). 5012 5013 sunrpc.pool_mode= 5014 [NFS] 5015 Control how the NFS server code allocates CPUs to 5016 service thread pools. Depending on how many NICs 5017 you have and where their interrupts are bound, this 5018 option will affect which CPUs will do NFS serving. 5019 Note: this parameter cannot be changed while the 5020 NFS server is running. 5021 5022 auto the server chooses an appropriate mode 5023 automatically using heuristics 5024 global a single global pool contains all CPUs 5025 percpu one pool for each CPU 5026 pernode one pool for each NUMA node (equivalent 5027 to global on non-NUMA machines) 5028 5029 sunrpc.tcp_slot_table_entries= 5030 sunrpc.udp_slot_table_entries= 5031 [NFS,SUNRPC] 5032 Sets the upper limit on the number of simultaneous 5033 RPC calls that can be sent from the client to a 5034 server. Increasing these values may allow you to 5035 improve throughput, but will also increase the 5036 amount of memory reserved for use by the client. 5037 5038 suspend.pm_test_delay= 5039 [SUSPEND] 5040 Sets the number of seconds to remain in a suspend test 5041 mode before resuming the system (see 5042 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 5043 is set. Default value is 5. 5044 5045 svm= [PPC] 5046 Format: { on | off | y | n | 1 | 0 } 5047 This parameter controls use of the Protected 5048 Execution Facility on pSeries. 5049 5050 swapaccount=[0|1] 5051 [KNL] Enable accounting of swap in memory resource 5052 controller if no parameter or 1 is given or disable 5053 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 5054 5055 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 5056 Format: { <int> | force | noforce } 5057 <int> -- Number of I/O TLB slabs 5058 force -- force using of bounce buffers even if they 5059 wouldn't be automatically used by the kernel 5060 noforce -- Never use bounce buffers (for debugging) 5061 5062 switches= [HW,M68k] 5063 5064 sysctl.*= [KNL] 5065 Set a sysctl parameter, right before loading the init 5066 process, as if the value was written to the respective 5067 /proc/sys/... file. Both '.' and '/' are recognized as 5068 separators. Unrecognized parameters and invalid values 5069 are reported in the kernel log. Sysctls registered 5070 later by a loaded module cannot be set this way. 5071 Example: sysctl.vm.swappiness=40 5072 5073 sysfs.deprecated=0|1 [KNL] 5074 Enable/disable old style sysfs layout for old udev 5075 on older distributions. When this option is enabled 5076 very new udev will not work anymore. When this option 5077 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 5078 in older udev will not work anymore. 5079 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 5080 the kernel configuration. 5081 5082 sysrq_always_enabled 5083 [KNL] 5084 Ignore sysrq setting - this boot parameter will 5085 neutralize any effect of /proc/sys/kernel/sysrq. 5086 Useful for debugging. 5087 5088 tcpmhash_entries= [KNL,NET] 5089 Set the number of tcp_metrics_hash slots. 5090 Default value is 8192 or 16384 depending on total 5091 ram pages. This is used to specify the TCP metrics 5092 cache size. See Documentation/networking/ip-sysctl.rst 5093 "tcp_no_metrics_save" section for more details. 5094 5095 tdfx= [HW,DRM] 5096 5097 test_suspend= [SUSPEND][,N] 5098 Specify "mem" (for Suspend-to-RAM) or "standby" (for 5099 standby suspend) or "freeze" (for suspend type freeze) 5100 as the system sleep state during system startup with 5101 the optional capability to repeat N number of times. 5102 The system is woken from this state using a 5103 wakeup-capable RTC alarm. 5104 5105 thash_entries= [KNL,NET] 5106 Set number of hash buckets for TCP connection 5107 5108 thermal.act= [HW,ACPI] 5109 -1: disable all active trip points in all thermal zones 5110 <degrees C>: override all lowest active trip points 5111 5112 thermal.crt= [HW,ACPI] 5113 -1: disable all critical trip points in all thermal zones 5114 <degrees C>: override all critical trip points 5115 5116 thermal.nocrt= [HW,ACPI] 5117 Set to disable actions on ACPI thermal zone 5118 critical and hot trip points. 5119 5120 thermal.off= [HW,ACPI] 5121 1: disable ACPI thermal control 5122 5123 thermal.psv= [HW,ACPI] 5124 -1: disable all passive trip points 5125 <degrees C>: override all passive trip points to this 5126 value 5127 5128 thermal.tzp= [HW,ACPI] 5129 Specify global default ACPI thermal zone polling rate 5130 <deci-seconds>: poll all this frequency 5131 0: no polling (default) 5132 5133 threadirqs [KNL] 5134 Force threading of all interrupt handlers except those 5135 marked explicitly IRQF_NO_THREAD. 5136 5137 topology= [S390] 5138 Format: {off | on} 5139 Specify if the kernel should make use of the cpu 5140 topology information if the hardware supports this. 5141 The scheduler will make use of this information and 5142 e.g. base its process migration decisions on it. 5143 Default is on. 5144 5145 topology_updates= [KNL, PPC, NUMA] 5146 Format: {off} 5147 Specify if the kernel should ignore (off) 5148 topology updates sent by the hypervisor to this 5149 LPAR. 5150 5151 torture.disable_onoff_at_boot= [KNL] 5152 Prevent the CPU-hotplug component of torturing 5153 until after init has spawned. 5154 5155 torture.ftrace_dump_at_shutdown= [KNL] 5156 Dump the ftrace buffer at torture-test shutdown, 5157 even if there were no errors. This can be a 5158 very costly operation when many torture tests 5159 are running concurrently, especially on systems 5160 with rotating-rust storage. 5161 5162 tp720= [HW,PS2] 5163 5164 tpm_suspend_pcr=[HW,TPM] 5165 Format: integer pcr id 5166 Specify that at suspend time, the tpm driver 5167 should extend the specified pcr with zeros, 5168 as a workaround for some chips which fail to 5169 flush the last written pcr on TPM_SaveState. 5170 This will guarantee that all the other pcrs 5171 are saved. 5172 5173 trace_buf_size=nn[KMG] 5174 [FTRACE] will set tracing buffer size on each cpu. 5175 5176 trace_event=[event-list] 5177 [FTRACE] Set and start specified trace events in order 5178 to facilitate early boot debugging. The event-list is a 5179 comma separated list of trace events to enable. See 5180 also Documentation/trace/events.rst 5181 5182 trace_options=[option-list] 5183 [FTRACE] Enable or disable tracer options at boot. 5184 The option-list is a comma delimited list of options 5185 that can be enabled or disabled just as if you were 5186 to echo the option name into 5187 5188 /sys/kernel/debug/tracing/trace_options 5189 5190 For example, to enable stacktrace option (to dump the 5191 stack trace of each event), add to the command line: 5192 5193 trace_options=stacktrace 5194 5195 See also Documentation/trace/ftrace.rst "trace options" 5196 section. 5197 5198 tp_printk[FTRACE] 5199 Have the tracepoints sent to printk as well as the 5200 tracing ring buffer. This is useful for early boot up 5201 where the system hangs or reboots and does not give the 5202 option for reading the tracing buffer or performing a 5203 ftrace_dump_on_oops. 5204 5205 To turn off having tracepoints sent to printk, 5206 echo 0 > /proc/sys/kernel/tracepoint_printk 5207 Note, echoing 1 into this file without the 5208 tracepoint_printk kernel cmdline option has no effect. 5209 5210 ** CAUTION ** 5211 5212 Having tracepoints sent to printk() and activating high 5213 frequency tracepoints such as irq or sched, can cause 5214 the system to live lock. 5215 5216 traceoff_on_warning 5217 [FTRACE] enable this option to disable tracing when a 5218 warning is hit. This turns off "tracing_on". Tracing can 5219 be enabled again by echoing '1' into the "tracing_on" 5220 file located in /sys/kernel/debug/tracing/ 5221 5222 This option is useful, as it disables the trace before 5223 the WARNING dump is called, which prevents the trace to 5224 be filled with content caused by the warning output. 5225 5226 This option can also be set at run time via the sysctl 5227 option: kernel/traceoff_on_warning 5228 5229 transparent_hugepage= 5230 [KNL] 5231 Format: [always|madvise|never] 5232 Can be used to control the default behavior of the system 5233 with respect to transparent hugepages. 5234 See Documentation/admin-guide/mm/transhuge.rst 5235 for more details. 5236 5237 tsc= Disable clocksource stability checks for TSC. 5238 Format: <string> 5239 [x86] reliable: mark tsc clocksource as reliable, this 5240 disables clocksource verification at runtime, as well 5241 as the stability checks done at bootup. Used to enable 5242 high-resolution timer mode on older hardware, and in 5243 virtualized environment. 5244 [x86] noirqtime: Do not use TSC to do irq accounting. 5245 Used to run time disable IRQ_TIME_ACCOUNTING on any 5246 platforms where RDTSC is slow and this accounting 5247 can add overhead. 5248 [x86] unstable: mark the TSC clocksource as unstable, this 5249 marks the TSC unconditionally unstable at bootup and 5250 avoids any further wobbles once the TSC watchdog notices. 5251 [x86] nowatchdog: disable clocksource watchdog. Used 5252 in situations with strict latency requirements (where 5253 interruptions from clocksource watchdog are not 5254 acceptable). 5255 5256 tsc_early_khz= [X86] Skip early TSC calibration and use the given 5257 value instead. Useful when the early TSC frequency discovery 5258 procedure is not reliable, such as on overclocked systems 5259 with CPUID.16h support and partial CPUID.15h support. 5260 Format: <unsigned int> 5261 5262 tsx= [X86] Control Transactional Synchronization 5263 Extensions (TSX) feature in Intel processors that 5264 support TSX control. 5265 5266 This parameter controls the TSX feature. The options are: 5267 5268 on - Enable TSX on the system. Although there are 5269 mitigations for all known security vulnerabilities, 5270 TSX has been known to be an accelerator for 5271 several previous speculation-related CVEs, and 5272 so there may be unknown security risks associated 5273 with leaving it enabled. 5274 5275 off - Disable TSX on the system. (Note that this 5276 option takes effect only on newer CPUs which are 5277 not vulnerable to MDS, i.e., have 5278 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 5279 the new IA32_TSX_CTRL MSR through a microcode 5280 update. This new MSR allows for the reliable 5281 deactivation of the TSX functionality.) 5282 5283 auto - Disable TSX if X86_BUG_TAA is present, 5284 otherwise enable TSX on the system. 5285 5286 Not specifying this option is equivalent to tsx=off. 5287 5288 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5289 for more details. 5290 5291 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 5292 Abort (TAA) vulnerability. 5293 5294 Similar to Micro-architectural Data Sampling (MDS) 5295 certain CPUs that support Transactional 5296 Synchronization Extensions (TSX) are vulnerable to an 5297 exploit against CPU internal buffers which can forward 5298 information to a disclosure gadget under certain 5299 conditions. 5300 5301 In vulnerable processors, the speculatively forwarded 5302 data can be used in a cache side channel attack, to 5303 access data to which the attacker does not have direct 5304 access. 5305 5306 This parameter controls the TAA mitigation. The 5307 options are: 5308 5309 full - Enable TAA mitigation on vulnerable CPUs 5310 if TSX is enabled. 5311 5312 full,nosmt - Enable TAA mitigation and disable SMT on 5313 vulnerable CPUs. If TSX is disabled, SMT 5314 is not disabled because CPU is not 5315 vulnerable to cross-thread TAA attacks. 5316 off - Unconditionally disable TAA mitigation 5317 5318 On MDS-affected machines, tsx_async_abort=off can be 5319 prevented by an active MDS mitigation as both vulnerabilities 5320 are mitigated with the same mechanism so in order to disable 5321 this mitigation, you need to specify mds=off too. 5322 5323 Not specifying this option is equivalent to 5324 tsx_async_abort=full. On CPUs which are MDS affected 5325 and deploy MDS mitigation, TAA mitigation is not 5326 required and doesn't provide any additional 5327 mitigation. 5328 5329 For details see: 5330 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5331 5332 turbografx.map[2|3]= [HW,JOY] 5333 TurboGraFX parallel port interface 5334 Format: 5335 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 5336 See also Documentation/input/devices/joystick-parport.rst 5337 5338 udbg-immortal [PPC] When debugging early kernel crashes that 5339 happen after console_init() and before a proper 5340 console driver takes over, this boot options might 5341 help "seeing" what's going on. 5342 5343 uhash_entries= [KNL,NET] 5344 Set number of hash buckets for UDP/UDP-Lite connections 5345 5346 uhci-hcd.ignore_oc= 5347 [USB] Ignore overcurrent events (default N). 5348 Some badly-designed motherboards generate lots of 5349 bogus events, for ports that aren't wired to 5350 anything. Set this parameter to avoid log spamming. 5351 Note that genuine overcurrent events won't be 5352 reported either. 5353 5354 unknown_nmi_panic 5355 [X86] Cause panic on unknown NMI. 5356 5357 usbcore.authorized_default= 5358 [USB] Default USB device authorization: 5359 (default -1 = authorized except for wireless USB, 5360 0 = not authorized, 1 = authorized, 2 = authorized 5361 if device connected to internal port) 5362 5363 usbcore.autosuspend= 5364 [USB] The autosuspend time delay (in seconds) used 5365 for newly-detected USB devices (default 2). This 5366 is the time required before an idle device will be 5367 autosuspended. Devices for which the delay is set 5368 to a negative value won't be autosuspended at all. 5369 5370 usbcore.usbfs_snoop= 5371 [USB] Set to log all usbfs traffic (default 0 = off). 5372 5373 usbcore.usbfs_snoop_max= 5374 [USB] Maximum number of bytes to snoop in each URB 5375 (default = 65536). 5376 5377 usbcore.blinkenlights= 5378 [USB] Set to cycle leds on hubs (default 0 = off). 5379 5380 usbcore.old_scheme_first= 5381 [USB] Start with the old device initialization 5382 scheme (default 0 = off). 5383 5384 usbcore.usbfs_memory_mb= 5385 [USB] Memory limit (in MB) for buffers allocated by 5386 usbfs (default = 16, 0 = max = 2047). 5387 5388 usbcore.use_both_schemes= 5389 [USB] Try the other device initialization scheme 5390 if the first one fails (default 1 = enabled). 5391 5392 usbcore.initial_descriptor_timeout= 5393 [USB] Specifies timeout for the initial 64-byte 5394 USB_REQ_GET_DESCRIPTOR request in milliseconds 5395 (default 5000 = 5.0 seconds). 5396 5397 usbcore.nousb [USB] Disable the USB subsystem 5398 5399 usbcore.quirks= 5400 [USB] A list of quirk entries to augment the built-in 5401 usb core quirk list. List entries are separated by 5402 commas. Each entry has the form 5403 VendorID:ProductID:Flags. The IDs are 4-digit hex 5404 numbers and Flags is a set of letters. Each letter 5405 will change the built-in quirk; setting it if it is 5406 clear and clearing it if it is set. The letters have 5407 the following meanings: 5408 a = USB_QUIRK_STRING_FETCH_255 (string 5409 descriptors must not be fetched using 5410 a 255-byte read); 5411 b = USB_QUIRK_RESET_RESUME (device can't resume 5412 correctly so reset it instead); 5413 c = USB_QUIRK_NO_SET_INTF (device can't handle 5414 Set-Interface requests); 5415 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5416 handle its Configuration or Interface 5417 strings); 5418 e = USB_QUIRK_RESET (device can't be reset 5419 (e.g morph devices), don't use reset); 5420 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5421 more interface descriptions than the 5422 bNumInterfaces count, and can't handle 5423 talking to these interfaces); 5424 g = USB_QUIRK_DELAY_INIT (device needs a pause 5425 during initialization, after we read 5426 the device descriptor); 5427 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5428 high speed and super speed interrupt 5429 endpoints, the USB 2.0 and USB 3.0 spec 5430 require the interval in microframes (1 5431 microframe = 125 microseconds) to be 5432 calculated as interval = 2 ^ 5433 (bInterval-1). 5434 Devices with this quirk report their 5435 bInterval as the result of this 5436 calculation instead of the exponent 5437 variable used in the calculation); 5438 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5439 handle device_qualifier descriptor 5440 requests); 5441 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5442 generates spurious wakeup, ignore 5443 remote wakeup capability); 5444 k = USB_QUIRK_NO_LPM (device can't handle Link 5445 Power Management); 5446 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5447 (Device reports its bInterval as linear 5448 frames instead of the USB 2.0 5449 calculation); 5450 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5451 to be disconnected before suspend to 5452 prevent spurious wakeup); 5453 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5454 pause after every control message); 5455 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5456 delay after resetting its port); 5457 Example: quirks=0781:5580:bk,0a5c:5834:gij 5458 5459 usbhid.mousepoll= 5460 [USBHID] The interval which mice are to be polled at. 5461 5462 usbhid.jspoll= 5463 [USBHID] The interval which joysticks are to be polled at. 5464 5465 usbhid.kbpoll= 5466 [USBHID] The interval which keyboards are to be polled at. 5467 5468 usb-storage.delay_use= 5469 [UMS] The delay in seconds before a new device is 5470 scanned for Logical Units (default 1). 5471 5472 usb-storage.quirks= 5473 [UMS] A list of quirks entries to supplement or 5474 override the built-in unusual_devs list. List 5475 entries are separated by commas. Each entry has 5476 the form VID:PID:Flags where VID and PID are Vendor 5477 and Product ID values (4-digit hex numbers) and 5478 Flags is a set of characters, each corresponding 5479 to a common usb-storage quirk flag as follows: 5480 a = SANE_SENSE (collect more than 18 bytes 5481 of sense data, not on uas); 5482 b = BAD_SENSE (don't collect more than 18 5483 bytes of sense data, not on uas); 5484 c = FIX_CAPACITY (decrease the reported 5485 device capacity by one sector); 5486 d = NO_READ_DISC_INFO (don't use 5487 READ_DISC_INFO command, not on uas); 5488 e = NO_READ_CAPACITY_16 (don't use 5489 READ_CAPACITY_16 command); 5490 f = NO_REPORT_OPCODES (don't use report opcodes 5491 command, uas only); 5492 g = MAX_SECTORS_240 (don't transfer more than 5493 240 sectors at a time, uas only); 5494 h = CAPACITY_HEURISTICS (decrease the 5495 reported device capacity by one 5496 sector if the number is odd); 5497 i = IGNORE_DEVICE (don't bind to this 5498 device); 5499 j = NO_REPORT_LUNS (don't use report luns 5500 command, uas only); 5501 l = NOT_LOCKABLE (don't try to lock and 5502 unlock ejectable media, not on uas); 5503 m = MAX_SECTORS_64 (don't transfer more 5504 than 64 sectors = 32 KB at a time, 5505 not on uas); 5506 n = INITIAL_READ10 (force a retry of the 5507 initial READ(10) command, not on uas); 5508 o = CAPACITY_OK (accept the capacity 5509 reported by the device, not on uas); 5510 p = WRITE_CACHE (the device cache is ON 5511 by default, not on uas); 5512 r = IGNORE_RESIDUE (the device reports 5513 bogus residue values, not on uas); 5514 s = SINGLE_LUN (the device has only one 5515 Logical Unit); 5516 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5517 commands, uas only); 5518 u = IGNORE_UAS (don't bind to the uas driver); 5519 w = NO_WP_DETECT (don't test whether the 5520 medium is write-protected). 5521 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5522 even if the device claims no cache, 5523 not on uas) 5524 Example: quirks=0419:aaf5:rl,0421:0433:rc 5525 5526 user_debug= [KNL,ARM] 5527 Format: <int> 5528 See arch/arm/Kconfig.debug help text. 5529 1 - undefined instruction events 5530 2 - system calls 5531 4 - invalid data aborts 5532 8 - SIGSEGV faults 5533 16 - SIGBUS faults 5534 Example: user_debug=31 5535 5536 userpte= 5537 [X86] Flags controlling user PTE allocations. 5538 5539 nohigh = do not allocate PTE pages in 5540 HIGHMEM regardless of setting 5541 of CONFIG_HIGHPTE. 5542 5543 vdso= [X86,SH] 5544 On X86_32, this is an alias for vdso32=. Otherwise: 5545 5546 vdso=1: enable VDSO (the default) 5547 vdso=0: disable VDSO mapping 5548 5549 vdso32= [X86] Control the 32-bit vDSO 5550 vdso32=1: enable 32-bit VDSO 5551 vdso32=0 or vdso32=2: disable 32-bit VDSO 5552 5553 See the help text for CONFIG_COMPAT_VDSO for more 5554 details. If CONFIG_COMPAT_VDSO is set, the default is 5555 vdso32=0; otherwise, the default is vdso32=1. 5556 5557 For compatibility with older kernels, vdso32=2 is an 5558 alias for vdso32=0. 5559 5560 Try vdso32=0 if you encounter an error that says: 5561 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 5562 5563 vector= [IA-64,SMP] 5564 vector=percpu: enable percpu vector domain 5565 5566 video= [FB] Frame buffer configuration 5567 See Documentation/fb/modedb.rst. 5568 5569 video.brightness_switch_enabled= [0,1] 5570 If set to 1, on receiving an ACPI notify event 5571 generated by hotkey, video driver will adjust brightness 5572 level and then send out the event to user space through 5573 the allocated input device; If set to 0, video driver 5574 will only send out the event without touching backlight 5575 brightness level. 5576 default: 1 5577 5578 virtio_mmio.device= 5579 [VMMIO] Memory mapped virtio (platform) device. 5580 5581 <size>@<baseaddr>:<irq>[:<id>] 5582 where: 5583 <size> := size (can use standard suffixes 5584 like K, M and G) 5585 <baseaddr> := physical base address 5586 <irq> := interrupt number (as passed to 5587 request_irq()) 5588 <id> := (optional) platform device id 5589 example: 5590 virtio_mmio.device=1K@0x100b0000:48:7 5591 5592 Can be used multiple times for multiple devices. 5593 5594 vga= [BOOT,X86-32] Select a particular video mode 5595 See Documentation/x86/boot.rst and 5596 Documentation/admin-guide/svga.rst. 5597 Use vga=ask for menu. 5598 This is actually a boot loader parameter; the value is 5599 passed to the kernel using a special protocol. 5600 5601 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 5602 May slow down system boot speed, especially when 5603 enabled on systems with a large amount of memory. 5604 All options are enabled by default, and this 5605 interface is meant to allow for selectively 5606 enabling or disabling specific virtual memory 5607 debugging features. 5608 5609 Available options are: 5610 P Enable page structure init time poisoning 5611 - Disable all of the above options 5612 5613 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 5614 size of <nn>. This can be used to increase the 5615 minimum size (128MB on x86). It can also be used to 5616 decrease the size and leave more room for directly 5617 mapped kernel RAM. 5618 5619 vmcp_cma=nn[MG] [KNL,S390] 5620 Sets the memory size reserved for contiguous memory 5621 allocations for the vmcp device driver. 5622 5623 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 5624 Format: <command> 5625 5626 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 5627 Format: <command> 5628 5629 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 5630 Format: <command> 5631 5632 vsyscall= [X86-64] 5633 Controls the behavior of vsyscalls (i.e. calls to 5634 fixed addresses of 0xffffffffff600x00 from legacy 5635 code). Most statically-linked binaries and older 5636 versions of glibc use these calls. Because these 5637 functions are at fixed addresses, they make nice 5638 targets for exploits that can control RIP. 5639 5640 emulate [default] Vsyscalls turn into traps and are 5641 emulated reasonably safely. The vsyscall 5642 page is readable. 5643 5644 xonly Vsyscalls turn into traps and are 5645 emulated reasonably safely. The vsyscall 5646 page is not readable. 5647 5648 none Vsyscalls don't work at all. This makes 5649 them quite hard to use for exploits but 5650 might break your system. 5651 5652 vt.color= [VT] Default text color. 5653 Format: 0xYX, X = foreground, Y = background. 5654 Default: 0x07 = light gray on black. 5655 5656 vt.cur_default= [VT] Default cursor shape. 5657 Format: 0xCCBBAA, where AA, BB, and CC are the same as 5658 the parameters of the <Esc>[?A;B;Cc escape sequence; 5659 see VGA-softcursor.txt. Default: 2 = underline. 5660 5661 vt.default_blu= [VT] 5662 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 5663 Change the default blue palette of the console. 5664 This is a 16-member array composed of values 5665 ranging from 0-255. 5666 5667 vt.default_grn= [VT] 5668 Format: <green0>,<green1>,<green2>,...,<green15> 5669 Change the default green palette of the console. 5670 This is a 16-member array composed of values 5671 ranging from 0-255. 5672 5673 vt.default_red= [VT] 5674 Format: <red0>,<red1>,<red2>,...,<red15> 5675 Change the default red palette of the console. 5676 This is a 16-member array composed of values 5677 ranging from 0-255. 5678 5679 vt.default_utf8= 5680 [VT] 5681 Format=<0|1> 5682 Set system-wide default UTF-8 mode for all tty's. 5683 Default is 1, i.e. UTF-8 mode is enabled for all 5684 newly opened terminals. 5685 5686 vt.global_cursor_default= 5687 [VT] 5688 Format=<-1|0|1> 5689 Set system-wide default for whether a cursor 5690 is shown on new VTs. Default is -1, 5691 i.e. cursors will be created by default unless 5692 overridden by individual drivers. 0 will hide 5693 cursors, 1 will display them. 5694 5695 vt.italic= [VT] Default color for italic text; 0-15. 5696 Default: 2 = green. 5697 5698 vt.underline= [VT] Default color for underlined text; 0-15. 5699 Default: 3 = cyan. 5700 5701 watchdog timers [HW,WDT] For information on watchdog timers, 5702 see Documentation/watchdog/watchdog-parameters.rst 5703 or other driver-specific files in the 5704 Documentation/watchdog/ directory. 5705 5706 watchdog_thresh= 5707 [KNL] 5708 Set the hard lockup detector stall duration 5709 threshold in seconds. The soft lockup detector 5710 threshold is set to twice the value. A value of 0 5711 disables both lockup detectors. Default is 10 5712 seconds. 5713 5714 workqueue.watchdog_thresh= 5715 If CONFIG_WQ_WATCHDOG is configured, workqueue can 5716 warn stall conditions and dump internal state to 5717 help debugging. 0 disables workqueue stall 5718 detection; otherwise, it's the stall threshold 5719 duration in seconds. The default value is 30 and 5720 it can be updated at runtime by writing to the 5721 corresponding sysfs file. 5722 5723 workqueue.disable_numa 5724 By default, all work items queued to unbound 5725 workqueues are affine to the NUMA nodes they're 5726 issued on, which results in better behavior in 5727 general. If NUMA affinity needs to be disabled for 5728 whatever reason, this option can be used. Note 5729 that this also can be controlled per-workqueue for 5730 workqueues visible under /sys/bus/workqueue/. 5731 5732 workqueue.power_efficient 5733 Per-cpu workqueues are generally preferred because 5734 they show better performance thanks to cache 5735 locality; unfortunately, per-cpu workqueues tend to 5736 be more power hungry than unbound workqueues. 5737 5738 Enabling this makes the per-cpu workqueues which 5739 were observed to contribute significantly to power 5740 consumption unbound, leading to measurably lower 5741 power usage at the cost of small performance 5742 overhead. 5743 5744 The default value of this parameter is determined by 5745 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 5746 5747 workqueue.debug_force_rr_cpu 5748 Workqueue used to implicitly guarantee that work 5749 items queued without explicit CPU specified are put 5750 on the local CPU. This guarantee is no longer true 5751 and while local CPU is still preferred work items 5752 may be put on foreign CPUs. This debug option 5753 forces round-robin CPU selection to flush out 5754 usages which depend on the now broken guarantee. 5755 When enabled, memory and cache locality will be 5756 impacted. 5757 5758 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 5759 default x2apic cluster mode on platforms 5760 supporting x2apic. 5761 5762 x86_intel_mid_timer= [X86-32,APBT] 5763 Choose timer option for x86 Intel MID platform. 5764 Two valid options are apbt timer only and lapic timer 5765 plus one apbt timer for broadcast timer. 5766 x86_intel_mid_timer=apbt_only | lapic_and_apbt 5767 5768 xen_512gb_limit [KNL,X86-64,XEN] 5769 Restricts the kernel running paravirtualized under Xen 5770 to use only up to 512 GB of RAM. The reason to do so is 5771 crash analysis tools and Xen tools for doing domain 5772 save/restore/migration must be enabled to handle larger 5773 domains. 5774 5775 xen_emul_unplug= [HW,X86,XEN] 5776 Unplug Xen emulated devices 5777 Format: [unplug0,][unplug1] 5778 ide-disks -- unplug primary master IDE devices 5779 aux-ide-disks -- unplug non-primary-master IDE devices 5780 nics -- unplug network devices 5781 all -- unplug all emulated devices (NICs and IDE disks) 5782 unnecessary -- unplugging emulated devices is 5783 unnecessary even if the host did not respond to 5784 the unplug protocol 5785 never -- do not unplug even if version check succeeds 5786 5787 xen_legacy_crash [X86,XEN] 5788 Crash from Xen panic notifier, without executing late 5789 panic() code such as dumping handler. 5790 5791 xen_nopvspin [X86,XEN] 5792 Disables the ticketlock slowpath using Xen PV 5793 optimizations. 5794 5795 xen_nopv [X86] 5796 Disables the PV optimizations forcing the HVM guest to 5797 run as generic HVM guest with no PV drivers. 5798 This option is obsoleted by the "nopv" option, which 5799 has equivalent effect for XEN platform. 5800 5801 xen_scrub_pages= [XEN] 5802 Boolean option to control scrubbing pages before giving them back 5803 to Xen, for use by other domains. Can be also changed at runtime 5804 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 5805 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 5806 5807 xen_timer_slop= [X86-64,XEN] 5808 Set the timer slop (in nanoseconds) for the virtual Xen 5809 timers (default is 100000). This adjusts the minimum 5810 delta of virtualized Xen timers, where lower values 5811 improve timer resolution at the expense of processing 5812 more timer interrupts. 5813 5814 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 5815 Disables the PV optimizations forcing the guest to run 5816 as generic guest with no PV drivers. Currently support 5817 XEN HVM, KVM, HYPER_V and VMWARE guest. 5818 5819 xirc2ps_cs= [NET,PCMCIA] 5820 Format: 5821 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 5822 5823 xive= [PPC] 5824 By default on POWER9 and above, the kernel will 5825 natively use the XIVE interrupt controller. This option 5826 allows the fallback firmware mode to be used: 5827 5828 off Fallback to firmware control of XIVE interrupt 5829 controller on both pseries and powernv 5830 platforms. Only useful on POWER9 and above. 5831 5832 xhci-hcd.quirks [USB,KNL] 5833 A hex value specifying bitmask with supplemental xhci 5834 host controller quirks. Meaning of each bit can be 5835 consulted in header drivers/usb/host/xhci.h. 5836 5837 xmon [PPC] 5838 Format: { early | on | rw | ro | off } 5839 Controls if xmon debugger is enabled. Default is off. 5840 Passing only "xmon" is equivalent to "xmon=early". 5841 early Call xmon as early as possible on boot; xmon 5842 debugger is called from setup_arch(). 5843 on xmon debugger hooks will be installed so xmon 5844 is only called on a kernel crash. Default mode, 5845 i.e. either "ro" or "rw" mode, is controlled 5846 with CONFIG_XMON_DEFAULT_RO_MODE. 5847 rw xmon debugger hooks will be installed so xmon 5848 is called only on a kernel crash, mode is write, 5849 meaning SPR registers, memory and, other data 5850 can be written using xmon commands. 5851 ro same as "rw" option above but SPR registers, 5852 memory, and other data can't be written using 5853 xmon commands. 5854 off xmon is disabled. 5855