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 allow_file_spec_access 293 Allow speculative faults on file backed pages. 294 Speculative faults are enabled only for those vm_ops 295 that implement and return true for allow_speculation 296 callback. 297 298 allow_mismatched_32bit_el0 [ARM64] 299 Allow execve() of 32-bit applications and setting of the 300 PER_LINUX32 personality on systems where only a strict 301 subset of the CPUs support 32-bit EL0. When this 302 parameter is present, the set of CPUs supporting 32-bit 303 EL0 is indicated by /sys/devices/system/cpu/aarch32_el0 304 and hot-unplug operations may be restricted. 305 306 amd_iommu= [HW,X86-64] 307 Pass parameters to the AMD IOMMU driver in the system. 308 Possible values are: 309 fullflush - enable flushing of IO/TLB entries when 310 they are unmapped. Otherwise they are 311 flushed before they will be reused, which 312 is a lot of faster 313 off - do not initialize any AMD IOMMU found in 314 the system 315 force_isolation - Force device isolation for all 316 devices. The IOMMU driver is not 317 allowed anymore to lift isolation 318 requirements as needed. This option 319 does not override iommu=pt 320 321 amd_iommu_dump= [HW,X86-64] 322 Enable AMD IOMMU driver option to dump the ACPI table 323 for AMD IOMMU. With this option enabled, AMD IOMMU 324 driver will print ACPI tables for AMD IOMMU during 325 IOMMU initialization. 326 327 amd_iommu_intr= [HW,X86-64] 328 Specifies one of the following AMD IOMMU interrupt 329 remapping modes: 330 legacy - Use legacy interrupt remapping mode. 331 vapic - Use virtual APIC mode, which allows IOMMU 332 to inject interrupts directly into guest. 333 This mode requires kvm-amd.avic=1. 334 (Default when IOMMU HW support is present.) 335 336 amijoy.map= [HW,JOY] Amiga joystick support 337 Map of devices attached to JOY0DAT and JOY1DAT 338 Format: <a>,<b> 339 See also Documentation/input/joydev/joystick.rst 340 341 analog.map= [HW,JOY] Analog joystick and gamepad support 342 Specifies type or capabilities of an analog joystick 343 connected to one of 16 gameports 344 Format: <type1>,<type2>,..<type16> 345 346 apc= [HW,SPARC] 347 Power management functions (SPARCstation-4/5 + deriv.) 348 Format: noidle 349 Disable APC CPU standby support. SPARCstation-Fox does 350 not play well with APC CPU idle - disable it if you have 351 APC and your system crashes randomly. 352 353 apic= [APIC,X86] Advanced Programmable Interrupt Controller 354 Change the output verbosity while booting 355 Format: { quiet (default) | verbose | debug } 356 Change the amount of debugging information output 357 when initialising the APIC and IO-APIC components. 358 For X86-32, this can also be used to specify an APIC 359 driver name. 360 Format: apic=driver_name 361 Examples: apic=bigsmp 362 363 apic_extnmi= [APIC,X86] External NMI delivery setting 364 Format: { bsp (default) | all | none } 365 bsp: External NMI is delivered only to CPU 0 366 all: External NMIs are broadcast to all CPUs as a 367 backup of CPU 0 368 none: External NMI is masked for all CPUs. This is 369 useful so that a dump capture kernel won't be 370 shot down by NMI 371 372 autoconf= [IPV6] 373 See Documentation/networking/ipv6.rst. 374 375 show_lapic= [APIC,X86] Advanced Programmable Interrupt Controller 376 Limit apic dumping. The parameter defines the maximal 377 number of local apics being dumped. Also it is possible 378 to set it to "all" by meaning -- no limit here. 379 Format: { 1 (default) | 2 | ... | all }. 380 The parameter valid if only apic=debug or 381 apic=verbose is specified. 382 Example: apic=debug show_lapic=all 383 384 apm= [APM] Advanced Power Management 385 See header of arch/x86/kernel/apm_32.c. 386 387 arcrimi= [HW,NET] ARCnet - "RIM I" (entirely mem-mapped) cards 388 Format: <io>,<irq>,<nodeID> 389 390 arm64.nobti [ARM64] Unconditionally disable Branch Target 391 Identification support 392 393 arm64.nopauth [ARM64] Unconditionally disable Pointer Authentication 394 support 395 396 arm64.nomte [ARM64] Unconditionally disable Memory Tagging Extension 397 support 398 399 ataflop= [HW,M68k] 400 401 atarimouse= [HW,MOUSE] Atari Mouse 402 403 atkbd.extra= [HW] Enable extra LEDs and keys on IBM RapidAccess, 404 EzKey and similar keyboards 405 406 atkbd.reset= [HW] Reset keyboard during initialization 407 408 atkbd.set= [HW] Select keyboard code set 409 Format: <int> (2 = AT (default), 3 = PS/2) 410 411 atkbd.scroll= [HW] Enable scroll wheel on MS Office and similar 412 keyboards 413 414 atkbd.softraw= [HW] Choose between synthetic and real raw mode 415 Format: <bool> (0 = real, 1 = synthetic (default)) 416 417 atkbd.softrepeat= [HW] 418 Use software keyboard repeat 419 420 audit= [KNL] Enable the audit sub-system 421 Format: { "0" | "1" | "off" | "on" } 422 0 | off - kernel audit is disabled and can not be 423 enabled until the next reboot 424 unset - kernel audit is initialized but disabled and 425 will be fully enabled by the userspace auditd. 426 1 | on - kernel audit is initialized and partially 427 enabled, storing at most audit_backlog_limit 428 messages in RAM until it is fully enabled by the 429 userspace auditd. 430 Default: unset 431 432 audit_backlog_limit= [KNL] Set the audit queue size limit. 433 Format: <int> (must be >=0) 434 Default: 64 435 436 bau= [X86_UV] Enable the BAU on SGI UV. The default 437 behavior is to disable the BAU (i.e. bau=0). 438 Format: { "0" | "1" } 439 0 - Disable the BAU. 440 1 - Enable the BAU. 441 unset - Disable the BAU. 442 443 baycom_epp= [HW,AX25] 444 Format: <io>,<mode> 445 446 baycom_par= [HW,AX25] BayCom Parallel Port AX.25 Modem 447 Format: <io>,<mode> 448 See header of drivers/net/hamradio/baycom_par.c. 449 450 baycom_ser_fdx= [HW,AX25] 451 BayCom Serial Port AX.25 Modem (Full Duplex Mode) 452 Format: <io>,<irq>,<mode>[,<baud>] 453 See header of drivers/net/hamradio/baycom_ser_fdx.c. 454 455 baycom_ser_hdx= [HW,AX25] 456 BayCom Serial Port AX.25 Modem (Half Duplex Mode) 457 Format: <io>,<irq>,<mode> 458 See header of drivers/net/hamradio/baycom_ser_hdx.c. 459 460 blkdevparts= Manual partition parsing of block device(s) for 461 embedded devices based on command line input. 462 See Documentation/block/cmdline-partition.rst 463 464 boot_delay= Milliseconds to delay each printk during boot. 465 Values larger than 10 seconds (10000) are changed to 466 no delay (0). 467 Format: integer 468 469 bootconfig [KNL] 470 Extended command line options can be added to an initrd 471 and this will cause the kernel to look for it. 472 473 See Documentation/admin-guide/bootconfig.rst 474 475 bert_disable [ACPI] 476 Disable BERT OS support on buggy BIOSes. 477 478 bgrt_disable [ACPI][X86] 479 Disable BGRT to avoid flickering OEM logo. 480 481 bttv.card= [HW,V4L] bttv (bt848 + bt878 based grabber cards) 482 bttv.radio= Most important insmod options are available as 483 kernel args too. 484 bttv.pll= See Documentation/admin-guide/media/bttv.rst 485 bttv.tuner= 486 487 bulk_remove=off [PPC] This parameter disables the use of the pSeries 488 firmware feature for flushing multiple hpte entries 489 at a time. 490 491 c101= [NET] Moxa C101 synchronous serial card 492 493 cachesize= [BUGS=X86-32] Override level 2 CPU cache size detection. 494 Sometimes CPU hardware bugs make them report the cache 495 size incorrectly. The kernel will attempt work arounds 496 to fix known problems, but for some CPUs it is not 497 possible to determine what the correct size should be. 498 This option provides an override for these situations. 499 500 carrier_timeout= 501 [NET] Specifies amount of time (in seconds) that 502 the kernel should wait for a network carrier. By default 503 it waits 120 seconds. 504 505 ca_keys= [KEYS] This parameter identifies a specific key(s) on 506 the system trusted keyring to be used for certificate 507 trust validation. 508 format: { id:<keyid> | builtin } 509 510 cca= [MIPS] Override the kernel pages' cache coherency 511 algorithm. Accepted values range from 0 to 7 512 inclusive. See arch/mips/include/asm/pgtable-bits.h 513 for platform specific values (SB1, Loongson3 and 514 others). 515 516 ccw_timeout_log [S390] 517 See Documentation/s390/common_io.rst for details. 518 519 cgroup_disable= [KNL] Disable a particular controller or optional feature 520 Format: {name of the controller(s) or feature(s) to disable} 521 The effects of cgroup_disable=foo are: 522 - foo isn't auto-mounted if you mount all cgroups in 523 a single hierarchy 524 - foo isn't visible as an individually mountable 525 subsystem 526 - if foo is an optional feature then the feature is 527 disabled and corresponding cgroup files are not 528 created 529 {Currently only "memory" controller deal with this and 530 cut the overhead, others just disable the usage. So 531 only cgroup_disable=memory is actually worthy} 532 Specifying "pressure" disables per-cgroup pressure 533 stall information accounting feature 534 535 cgroup_no_v1= [KNL] Disable cgroup controllers and named hierarchies in v1 536 Format: { { controller | "all" | "named" } 537 [,{ controller | "all" | "named" }...] } 538 Like cgroup_disable, but only applies to cgroup v1; 539 the blacklisted controllers remain available in cgroup2. 540 "all" blacklists all controllers and "named" disables 541 named mounts. Specifying both "all" and "named" disables 542 all v1 hierarchies. 543 544 cgroup.memory= [KNL] Pass options to the cgroup memory controller. 545 Format: <string> 546 nosocket -- Disable socket memory accounting. 547 nokmem -- Disable kernel memory accounting. 548 549 checkreqprot [SELINUX] Set initial checkreqprot flag value. 550 Format: { "0" | "1" } 551 See security/selinux/Kconfig help text. 552 0 -- check protection applied by kernel (includes 553 any implied execute protection). 554 1 -- check protection requested by application. 555 Default value is set via a kernel config option. 556 Value can be changed at runtime via 557 /sys/fs/selinux/checkreqprot. 558 Setting checkreqprot to 1 is deprecated. 559 560 cio_ignore= [S390] 561 See Documentation/s390/common_io.rst for details. 562 clk_ignore_unused 563 [CLK] 564 Prevents the clock framework from automatically gating 565 clocks that have not been explicitly enabled by a Linux 566 device driver but are enabled in hardware at reset or 567 by the bootloader/firmware. Note that this does not 568 force such clocks to be always-on nor does it reserve 569 those clocks in any way. This parameter is useful for 570 debug and development, but should not be needed on a 571 platform with proper driver support. For more 572 information, see Documentation/driver-api/clk.rst. 573 574 clock= [BUGS=X86-32, HW] gettimeofday clocksource override. 575 [Deprecated] 576 Forces specified clocksource (if available) to be used 577 when calculating gettimeofday(). If specified 578 clocksource is not available, it defaults to PIT. 579 Format: { pit | tsc | cyclone | pmtmr } 580 581 clocksource= Override the default clocksource 582 Format: <string> 583 Override the default clocksource and use the clocksource 584 with the name specified. 585 Some clocksource names to choose from, depending on 586 the platform: 587 [all] jiffies (this is the base, fallback clocksource) 588 [ACPI] acpi_pm 589 [ARM] imx_timer1,OSTS,netx_timer,mpu_timer2, 590 pxa_timer,timer3,32k_counter,timer0_1 591 [X86-32] pit,hpet,tsc; 592 scx200_hrt on Geode; cyclone on IBM x440 593 [MIPS] MIPS 594 [PARISC] cr16 595 [S390] tod 596 [SH] SuperH 597 [SPARC64] tick 598 [X86-64] hpet,tsc 599 600 clocksource.arm_arch_timer.evtstrm= 601 [ARM,ARM64] 602 Format: <bool> 603 Enable/disable the eventstream feature of the ARM 604 architected timer so that code using WFE-based polling 605 loops can be debugged more effectively on production 606 systems. 607 608 clocksource.max_cswd_read_retries= [KNL] 609 Number of clocksource_watchdog() retries due to 610 external delays before the clock will be marked 611 unstable. Defaults to three retries, that is, 612 four attempts to read the clock under test. 613 614 clearcpuid=BITNUM[,BITNUM...] [X86] 615 Disable CPUID feature X for the kernel. See 616 arch/x86/include/asm/cpufeatures.h for the valid bit 617 numbers. Note the Linux specific bits are not necessarily 618 stable over kernel options, but the vendor specific 619 ones should be. 620 Also note that user programs calling CPUID directly 621 or using the feature without checking anything 622 will still see it. This just prevents it from 623 being used by the kernel or shown in /proc/cpuinfo. 624 Also note the kernel might malfunction if you disable 625 some critical bits. 626 627 cma=nn[MG]@[start[MG][-end[MG]]] 628 [KNL,CMA] 629 Sets the size of kernel global memory area for 630 contiguous memory allocations and optionally the 631 placement constraint by the physical address range of 632 memory allocations. A value of 0 disables CMA 633 altogether. For more information, see 634 kernel/dma/contiguous.c 635 636 cma_pernuma=nn[MG] 637 [ARM64,KNL] 638 Sets the size of kernel per-numa memory area for 639 contiguous memory allocations. A value of 0 disables 640 per-numa CMA altogether. And If this option is not 641 specificed, the default value is 0. 642 With per-numa CMA enabled, DMA users on node nid will 643 first try to allocate buffer from the pernuma area 644 which is located in node nid, if the allocation fails, 645 they will fallback to the global default memory area. 646 647 cmo_free_hint= [PPC] Format: { yes | no } 648 Specify whether pages are marked as being inactive 649 when they are freed. This is used in CMO environments 650 to determine OS memory pressure for page stealing by 651 a hypervisor. 652 Default: yes 653 654 coherent_pool=nn[KMG] [ARM,KNL] 655 Sets the size of memory pool for coherent, atomic dma 656 allocations, by default set to 256K. 657 658 com20020= [HW,NET] ARCnet - COM20020 chipset 659 Format: 660 <io>[,<irq>[,<nodeID>[,<backplane>[,<ckp>[,<timeout>]]]]] 661 662 com90io= [HW,NET] ARCnet - COM90xx chipset (IO-mapped buffers) 663 Format: <io>[,<irq>] 664 665 com90xx= [HW,NET] 666 ARCnet - COM90xx chipset (memory-mapped buffers) 667 Format: <io>[,<irq>[,<memstart>]] 668 669 condev= [HW,S390] console device 670 conmode= 671 672 console= [KNL] Output console device and options. 673 674 tty<n> Use the virtual console device <n>. 675 676 ttyS<n>[,options] 677 ttyUSB0[,options] 678 Use the specified serial port. The options are of 679 the form "bbbbpnf", where "bbbb" is the baud rate, 680 "p" is parity ("n", "o", or "e"), "n" is number of 681 bits, and "f" is flow control ("r" for RTS or 682 omit it). Default is "9600n8". 683 684 See Documentation/admin-guide/serial-console.rst for more 685 information. See 686 Documentation/networking/netconsole.rst for an 687 alternative. 688 689 uart[8250],io,<addr>[,options] 690 uart[8250],mmio,<addr>[,options] 691 uart[8250],mmio16,<addr>[,options] 692 uart[8250],mmio32,<addr>[,options] 693 uart[8250],0x<addr>[,options] 694 Start an early, polled-mode console on the 8250/16550 695 UART at the specified I/O port or MMIO address, 696 switching to the matching ttyS device later. 697 MMIO inter-register address stride is either 8-bit 698 (mmio), 16-bit (mmio16), or 32-bit (mmio32). 699 If none of [io|mmio|mmio16|mmio32], <addr> is assumed 700 to be equivalent to 'mmio'. 'options' are specified in 701 the same format described for ttyS above; if unspecified, 702 the h/w is not re-initialized. 703 704 hvc<n> Use the hypervisor console device <n>. This is for 705 both Xen and PowerPC hypervisors. 706 707 If the device connected to the port is not a TTY but a braille 708 device, prepend "brl," before the device type, for instance 709 console=brl,ttyS0 710 For now, only VisioBraille is supported. 711 712 console_msg_format= 713 [KNL] Change console messages format 714 default 715 By default we print messages on consoles in 716 "[time stamp] text\n" format (time stamp may not be 717 printed, depending on CONFIG_PRINTK_TIME or 718 `printk_time' param). 719 syslog 720 Switch to syslog format: "<%u>[time stamp] text\n" 721 IOW, each message will have a facility and loglevel 722 prefix. The format is similar to one used by syslog() 723 syscall, or to executing "dmesg -S --raw" or to reading 724 from /proc/kmsg. 725 726 consoleblank= [KNL] The console blank (screen saver) timeout in 727 seconds. A value of 0 disables the blank timer. 728 Defaults to 0. 729 730 coredump_filter= 731 [KNL] Change the default value for 732 /proc/<pid>/coredump_filter. 733 See also Documentation/filesystems/proc.rst. 734 735 coresight_cpu_debug.enable 736 [ARM,ARM64] 737 Format: <bool> 738 Enable/disable the CPU sampling based debugging. 739 0: default value, disable debugging 740 1: enable debugging at boot time 741 742 cpuidle.off=1 [CPU_IDLE] 743 disable the cpuidle sub-system 744 745 cpuidle.governor= 746 [CPU_IDLE] Name of the cpuidle governor to use. 747 748 cpufreq.off=1 [CPU_FREQ] 749 disable the cpufreq sub-system 750 751 cpufreq.default_governor= 752 [CPU_FREQ] Name of the default cpufreq governor or 753 policy to use. This governor must be registered in the 754 kernel before the cpufreq driver probes. 755 756 cpu_init_udelay=N 757 [X86] Delay for N microsec between assert and de-assert 758 of APIC INIT to start processors. This delay occurs 759 on every CPU online, such as boot, and resume from suspend. 760 Default: 10000 761 762 cpcihp_generic= [HW,PCI] Generic port I/O CompactPCI driver 763 Format: 764 <first_slot>,<last_slot>,<port>,<enum_bit>[,<debug>] 765 766 crashkernel=size[KMG][@offset[KMG]] 767 [KNL] Using kexec, Linux can switch to a 'crash kernel' 768 upon panic. This parameter reserves the physical 769 memory region [offset, offset + size] for that kernel 770 image. If '@offset' is omitted, then a suitable offset 771 is selected automatically. 772 [KNL, X86-64] Select a region under 4G first, and 773 fall back to reserve region above 4G when '@offset' 774 hasn't been specified. 775 See Documentation/admin-guide/kdump/kdump.rst for further details. 776 777 crashkernel=range1:size1[,range2:size2,...][@offset] 778 [KNL] Same as above, but depends on the memory 779 in the running system. The syntax of range is 780 start-[end] where start and end are both 781 a memory unit (amount[KMG]). See also 782 Documentation/admin-guide/kdump/kdump.rst for an example. 783 784 crashkernel=size[KMG],high 785 [KNL, X86-64] range could be above 4G. Allow kernel 786 to allocate physical memory region from top, so could 787 be above 4G if system have more than 4G ram installed. 788 Otherwise memory region will be allocated below 4G, if 789 available. 790 It will be ignored if crashkernel=X is specified. 791 crashkernel=size[KMG],low 792 [KNL, X86-64] range under 4G. When crashkernel=X,high 793 is passed, kernel could allocate physical memory region 794 above 4G, that cause second kernel crash on system 795 that require some amount of low memory, e.g. swiotlb 796 requires at least 64M+32K low memory, also enough extra 797 low memory is needed to make sure DMA buffers for 32-bit 798 devices won't run out. Kernel would try to allocate at 799 at least 256M below 4G automatically. 800 This one let user to specify own low range under 4G 801 for second kernel instead. 802 0: to disable low allocation. 803 It will be ignored when crashkernel=X,high is not used 804 or memory reserved is below 4G. 805 806 cryptomgr.notests 807 [KNL] Disable crypto self-tests 808 809 cs89x0_dma= [HW,NET] 810 Format: <dma> 811 812 cs89x0_media= [HW,NET] 813 Format: { rj45 | aui | bnc } 814 815 dasd= [HW,NET] 816 See header of drivers/s390/block/dasd_devmap.c. 817 818 db9.dev[2|3]= [HW,JOY] Multisystem joystick support via parallel port 819 (one device per port) 820 Format: <port#>,<type> 821 See also Documentation/input/devices/joystick-parport.rst 822 823 ddebug_query= [KNL,DYNAMIC_DEBUG] Enable debug messages at early boot 824 time. See 825 Documentation/admin-guide/dynamic-debug-howto.rst for 826 details. Deprecated, see dyndbg. 827 828 debug [KNL] Enable kernel debugging (events log level). 829 830 debug_boot_weak_hash 831 [KNL] Enable printing [hashed] pointers early in the 832 boot sequence. If enabled, we use a weak hash instead 833 of siphash to hash pointers. Use this option if you are 834 seeing instances of '(___ptrval___)') and need to see a 835 value (hashed pointer) instead. Cryptographically 836 insecure, please do not use on production kernels. 837 838 debug_locks_verbose= 839 [KNL] verbose self-tests 840 Format=<0|1> 841 Print debugging info while doing the locking API 842 self-tests. 843 We default to 0 (no extra messages), setting it to 844 1 will print _a lot_ more information - normally 845 only useful to kernel developers. 846 847 debug_objects [KNL] Enable object debugging 848 849 no_debug_objects 850 [KNL] Disable object debugging 851 852 debug_guardpage_minorder= 853 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this 854 parameter allows control of the order of pages that will 855 be intentionally kept free (and hence protected) by the 856 buddy allocator. Bigger value increase the probability 857 of catching random memory corruption, but reduce the 858 amount of memory for normal system use. The maximum 859 possible value is MAX_ORDER/2. Setting this parameter 860 to 1 or 2 should be enough to identify most random 861 memory corruption problems caused by bugs in kernel or 862 driver code when a CPU writes to (or reads from) a 863 random memory location. Note that there exists a class 864 of memory corruptions problems caused by buggy H/W or 865 F/W or by drivers badly programing DMA (basically when 866 memory is written at bus level and the CPU MMU is 867 bypassed) which are not detectable by 868 CONFIG_DEBUG_PAGEALLOC, hence this option will not help 869 tracking down these problems. 870 871 debug_pagealloc= 872 [KNL] When CONFIG_DEBUG_PAGEALLOC is set, this parameter 873 enables the feature at boot time. By default, it is 874 disabled and the system will work mostly the same as a 875 kernel built without CONFIG_DEBUG_PAGEALLOC. 876 Note: to get most of debug_pagealloc error reports, it's 877 useful to also enable the page_owner functionality. 878 on: enable the feature 879 880 debugfs= [KNL] This parameter enables what is exposed to userspace 881 and debugfs internal clients. 882 Format: { on, no-mount, off } 883 on: All functions are enabled. 884 no-mount: 885 Filesystem is not registered but kernel clients can 886 access APIs and a crashkernel can be used to read 887 its content. There is nothing to mount. 888 off: Filesystem is not registered and clients 889 get a -EPERM as result when trying to register files 890 or directories within debugfs. 891 This is equivalent of the runtime functionality if 892 debugfs was not enabled in the kernel at all. 893 Default value is set in build-time with a kernel configuration. 894 895 debugpat [X86] Enable PAT debugging 896 897 decnet.addr= [HW,NET] 898 Format: <area>[,<node>] 899 See also Documentation/networking/decnet.rst. 900 901 default_hugepagesz= 902 [HW] The size of the default HugeTLB page. This is 903 the size represented by the legacy /proc/ hugepages 904 APIs. In addition, this is the default hugetlb size 905 used for shmget(), mmap() and mounting hugetlbfs 906 filesystems. If not specified, defaults to the 907 architecture's default huge page size. Huge page 908 sizes are architecture dependent. See also 909 Documentation/admin-guide/mm/hugetlbpage.rst. 910 Format: size[KMG] 911 912 deferred_probe_timeout= 913 [KNL] Debugging option to set a timeout in seconds for 914 deferred probe to give up waiting on dependencies to 915 probe. Only specific dependencies (subsystems or 916 drivers) that have opted in will be ignored. A timeout of 0 917 will timeout at the end of initcalls. This option will also 918 dump out devices still on the deferred probe list after 919 retrying. 920 921 dfltcc= [HW,S390] 922 Format: { on | off | def_only | inf_only | always } 923 on: s390 zlib hardware support for compression on 924 level 1 and decompression (default) 925 off: No s390 zlib hardware support 926 def_only: s390 zlib hardware support for deflate 927 only (compression on level 1) 928 inf_only: s390 zlib hardware support for inflate 929 only (decompression) 930 always: Same as 'on' but ignores the selected compression 931 level always using hardware support (used for debugging) 932 933 dhash_entries= [KNL] 934 Set number of hash buckets for dentry cache. 935 936 disable_1tb_segments [PPC] 937 Disables the use of 1TB hash page table segments. This 938 causes the kernel to fall back to 256MB segments which 939 can be useful when debugging issues that require an SLB 940 miss to occur. 941 942 disable_dma32= [KNL] 943 Dynamically disable ZONE_DMA32 on kernels compiled with 944 CONFIG_ZONE_DMA32=y. 945 946 stress_slb [PPC] 947 Limits the number of kernel SLB entries, and flushes 948 them frequently to increase the rate of SLB faults 949 on kernel addresses. 950 951 disable= [IPV6] 952 See Documentation/networking/ipv6.rst. 953 954 hardened_usercopy= 955 [KNL] Under CONFIG_HARDENED_USERCOPY, whether 956 hardening is enabled for this boot. Hardened 957 usercopy checking is used to protect the kernel 958 from reading or writing beyond known memory 959 allocation boundaries as a proactive defense 960 against bounds-checking flaws in the kernel's 961 copy_to_user()/copy_from_user() interface. 962 on Perform hardened usercopy checks (default). 963 off Disable hardened usercopy checks. 964 965 disable_radix [PPC] 966 Disable RADIX MMU mode on POWER9 967 968 radix_hcall_invalidate=on [PPC/PSERIES] 969 Disable RADIX GTSE feature and use hcall for TLB 970 invalidate. 971 972 disable_tlbie [PPC] 973 Disable TLBIE instruction. Currently does not work 974 with KVM, with HASH MMU, or with coherent accelerators. 975 976 disable_cpu_apicid= [X86,APIC,SMP] 977 Format: <int> 978 The number of initial APIC ID for the 979 corresponding CPU to be disabled at boot, 980 mostly used for the kdump 2nd kernel to 981 disable BSP to wake up multiple CPUs without 982 causing system reset or hang due to sending 983 INIT from AP to BSP. 984 985 perf_v4_pmi= [X86,INTEL] 986 Format: <bool> 987 Disable Intel PMU counter freezing feature. 988 The feature only exists starting from 989 Arch Perfmon v4 (Skylake and newer). 990 991 disable_ddw [PPC/PSERIES] 992 Disable Dynamic DMA Window support. Use this 993 to workaround buggy firmware. 994 995 disable_ipv6= [IPV6] 996 See Documentation/networking/ipv6.rst. 997 998 disable_mtrr_cleanup [X86] 999 The kernel tries to adjust MTRR layout from continuous 1000 to discrete, to make X server driver able to add WB 1001 entry later. This parameter disables that. 1002 1003 disable_mtrr_trim [X86, Intel and AMD only] 1004 By default the kernel will trim any uncacheable 1005 memory out of your available memory pool based on 1006 MTRR settings. This parameter disables that behavior, 1007 possibly causing your machine to run very slowly. 1008 1009 disable_timer_pin_1 [X86] 1010 Disable PIN 1 of APIC timer 1011 Can be useful to work around chipset bugs. 1012 1013 dis_ucode_ldr [X86] Disable the microcode loader. 1014 1015 dma_debug=off If the kernel is compiled with DMA_API_DEBUG support, 1016 this option disables the debugging code at boot. 1017 1018 dma_debug_entries=<number> 1019 This option allows to tune the number of preallocated 1020 entries for DMA-API debugging code. One entry is 1021 required per DMA-API allocation. Use this if the 1022 DMA-API debugging code disables itself because the 1023 architectural default is too low. 1024 1025 dma_debug_driver=<driver_name> 1026 With this option the DMA-API debugging driver 1027 filter feature can be enabled at boot time. Just 1028 pass the driver to filter for as the parameter. 1029 The filter can be disabled or changed to another 1030 driver later using sysfs. 1031 1032 driver_async_probe= [KNL] 1033 List of driver names to be probed asynchronously. 1034 Format: <driver_name1>,<driver_name2>... 1035 1036 drm.edid_firmware=[<connector>:]<file>[,[<connector>:]<file>] 1037 Broken monitors, graphic adapters, KVMs and EDIDless 1038 panels may send no or incorrect EDID data sets. 1039 This parameter allows to specify an EDID data sets 1040 in the /lib/firmware directory that are used instead. 1041 Generic built-in EDID data sets are used, if one of 1042 edid/1024x768.bin, edid/1280x1024.bin, 1043 edid/1680x1050.bin, or edid/1920x1080.bin is given 1044 and no file with the same name exists. Details and 1045 instructions how to build your own EDID data are 1046 available in Documentation/admin-guide/edid.rst. An EDID 1047 data set will only be used for a particular connector, 1048 if its name and a colon are prepended to the EDID 1049 name. Each connector may use a unique EDID data 1050 set by separating the files with a comma. An EDID 1051 data set with no connector name will be used for 1052 any connectors not explicitly specified. 1053 1054 dscc4.setup= [NET] 1055 1056 dt_cpu_ftrs= [PPC] 1057 Format: {"off" | "known"} 1058 Control how the dt_cpu_ftrs device-tree binding is 1059 used for CPU feature discovery and setup (if it 1060 exists). 1061 off: Do not use it, fall back to legacy cpu table. 1062 known: Do not pass through unknown features to guests 1063 or userspace, only those that the kernel is aware of. 1064 1065 dump_apple_properties [X86] 1066 Dump name and content of EFI device properties on 1067 x86 Macs. Useful for driver authors to determine 1068 what data is available or for reverse-engineering. 1069 1070 dyndbg[="val"] [KNL,DYNAMIC_DEBUG] 1071 <module>.dyndbg[="val"] 1072 Enable debug messages at boot time. See 1073 Documentation/admin-guide/dynamic-debug-howto.rst 1074 for details. 1075 1076 nopku [X86] Disable Memory Protection Keys CPU feature found 1077 in some Intel CPUs. 1078 1079 <module>.async_probe [KNL] 1080 Enable asynchronous probe on this module. 1081 1082 early_ioremap_debug [KNL] 1083 Enable debug messages in early_ioremap support. This 1084 is useful for tracking down temporary early mappings 1085 which are not unmapped. 1086 1087 earlycon= [KNL] Output early console device and options. 1088 1089 When used with no options, the early console is 1090 determined by stdout-path property in device tree's 1091 chosen node or the ACPI SPCR table if supported by 1092 the platform. 1093 1094 cdns,<addr>[,options] 1095 Start an early, polled-mode console on a Cadence 1096 (xuartps) serial port at the specified address. Only 1097 supported option is baud rate. If baud rate is not 1098 specified, the serial port must already be setup and 1099 configured. 1100 1101 uart[8250],io,<addr>[,options] 1102 uart[8250],mmio,<addr>[,options] 1103 uart[8250],mmio32,<addr>[,options] 1104 uart[8250],mmio32be,<addr>[,options] 1105 uart[8250],0x<addr>[,options] 1106 Start an early, polled-mode console on the 8250/16550 1107 UART at the specified I/O port or MMIO address. 1108 MMIO inter-register address stride is either 8-bit 1109 (mmio) or 32-bit (mmio32 or mmio32be). 1110 If none of [io|mmio|mmio32|mmio32be], <addr> is assumed 1111 to be equivalent to 'mmio'. 'options' are specified 1112 in the same format described for "console=ttyS<n>"; if 1113 unspecified, the h/w is not initialized. 1114 1115 pl011,<addr> 1116 pl011,mmio32,<addr> 1117 Start an early, polled-mode console on a pl011 serial 1118 port at the specified address. The pl011 serial port 1119 must already be setup and configured. Options are not 1120 yet supported. If 'mmio32' is specified, then only 1121 the driver will use only 32-bit accessors to read/write 1122 the device registers. 1123 1124 meson,<addr> 1125 Start an early, polled-mode console on a meson serial 1126 port at the specified address. The serial port must 1127 already be setup and configured. Options are not yet 1128 supported. 1129 1130 msm_serial,<addr> 1131 Start an early, polled-mode console on an msm serial 1132 port at the specified address. The serial port 1133 must already be setup and configured. Options are not 1134 yet supported. 1135 1136 msm_serial_dm,<addr> 1137 Start an early, polled-mode console on an msm serial 1138 dm port at the specified address. The serial port 1139 must already be setup and configured. Options are not 1140 yet supported. 1141 1142 owl,<addr> 1143 Start an early, polled-mode console on a serial port 1144 of an Actions Semi SoC, such as S500 or S900, at the 1145 specified address. The serial port must already be 1146 setup and configured. Options are not yet supported. 1147 1148 rda,<addr> 1149 Start an early, polled-mode console on a serial port 1150 of an RDA Micro SoC, such as RDA8810PL, at the 1151 specified address. The serial port must already be 1152 setup and configured. Options are not yet supported. 1153 1154 sbi 1155 Use RISC-V SBI (Supervisor Binary Interface) for early 1156 console. 1157 1158 smh Use ARM semihosting calls for early console. 1159 1160 s3c2410,<addr> 1161 s3c2412,<addr> 1162 s3c2440,<addr> 1163 s3c6400,<addr> 1164 s5pv210,<addr> 1165 exynos4210,<addr> 1166 Use early console provided by serial driver available 1167 on Samsung SoCs, requires selecting proper type and 1168 a correct base address of the selected UART port. The 1169 serial port must already be setup and configured. 1170 Options are not yet supported. 1171 1172 lantiq,<addr> 1173 Start an early, polled-mode console on a lantiq serial 1174 (lqasc) port at the specified address. The serial port 1175 must already be setup and configured. Options are not 1176 yet supported. 1177 1178 lpuart,<addr> 1179 lpuart32,<addr> 1180 Use early console provided by Freescale LP UART driver 1181 found on Freescale Vybrid and QorIQ LS1021A processors. 1182 A valid base address must be provided, and the serial 1183 port must already be setup and configured. 1184 1185 ec_imx21,<addr> 1186 ec_imx6q,<addr> 1187 Start an early, polled-mode, output-only console on the 1188 Freescale i.MX UART at the specified address. The UART 1189 must already be setup and configured. 1190 1191 ar3700_uart,<addr> 1192 Start an early, polled-mode console on the 1193 Armada 3700 serial port at the specified 1194 address. The serial port must already be setup 1195 and configured. Options are not yet supported. 1196 1197 qcom_geni,<addr> 1198 Start an early, polled-mode console on a Qualcomm 1199 Generic Interface (GENI) based serial port at the 1200 specified address. The serial port must already be 1201 setup and configured. Options are not yet supported. 1202 1203 efifb,[options] 1204 Start an early, unaccelerated console on the EFI 1205 memory mapped framebuffer (if available). On cache 1206 coherent non-x86 systems that use system memory for 1207 the framebuffer, pass the 'ram' option so that it is 1208 mapped with the correct attributes. 1209 1210 linflex,<addr> 1211 Use early console provided by Freescale LINFlexD UART 1212 serial driver for NXP S32V234 SoCs. A valid base 1213 address must be provided, and the serial port must 1214 already be setup and configured. 1215 1216 earlyprintk= [X86,SH,ARM,M68k,S390] 1217 earlyprintk=vga 1218 earlyprintk=sclp 1219 earlyprintk=xen 1220 earlyprintk=serial[,ttySn[,baudrate]] 1221 earlyprintk=serial[,0x...[,baudrate]] 1222 earlyprintk=ttySn[,baudrate] 1223 earlyprintk=dbgp[debugController#] 1224 earlyprintk=pciserial[,force],bus:device.function[,baudrate] 1225 earlyprintk=xdbc[xhciController#] 1226 1227 earlyprintk is useful when the kernel crashes before 1228 the normal console is initialized. It is not enabled by 1229 default because it has some cosmetic problems. 1230 1231 Append ",keep" to not disable it when the real console 1232 takes over. 1233 1234 Only one of vga, efi, serial, or usb debug port can 1235 be used at a time. 1236 1237 Currently only ttyS0 and ttyS1 may be specified by 1238 name. Other I/O ports may be explicitly specified 1239 on some architectures (x86 and arm at least) by 1240 replacing ttySn with an I/O port address, like this: 1241 earlyprintk=serial,0x1008,115200 1242 You can find the port for a given device in 1243 /proc/tty/driver/serial: 1244 2: uart:ST16650V2 port:00001008 irq:18 ... 1245 1246 Interaction with the standard serial driver is not 1247 very good. 1248 1249 The VGA and EFI output is eventually overwritten by 1250 the real console. 1251 1252 The xen output can only be used by Xen PV guests. 1253 1254 The sclp output can only be used on s390. 1255 1256 The optional "force" to "pciserial" enables use of a 1257 PCI device even when its classcode is not of the 1258 UART class. 1259 1260 edac_report= [HW,EDAC] Control how to report EDAC event 1261 Format: {"on" | "off" | "force"} 1262 on: enable EDAC to report H/W event. May be overridden 1263 by other higher priority error reporting module. 1264 off: disable H/W event reporting through EDAC. 1265 force: enforce the use of EDAC to report H/W event. 1266 default: on. 1267 1268 ekgdboc= [X86,KGDB] Allow early kernel console debugging 1269 ekgdboc=kbd 1270 1271 This is designed to be used in conjunction with 1272 the boot argument: earlyprintk=vga 1273 1274 This parameter works in place of the kgdboc parameter 1275 but can only be used if the backing tty is available 1276 very early in the boot process. For early debugging 1277 via a serial port see kgdboc_earlycon instead. 1278 1279 edd= [EDD] 1280 Format: {"off" | "on" | "skip[mbr]"} 1281 1282 efi= [EFI] 1283 Format: { "debug", "disable_early_pci_dma", 1284 "nochunk", "noruntime", "nosoftreserve", 1285 "novamap", "no_disable_early_pci_dma" } 1286 debug: enable misc debug output. 1287 disable_early_pci_dma: disable the busmaster bit on all 1288 PCI bridges while in the EFI boot stub. 1289 nochunk: disable reading files in "chunks" in the EFI 1290 boot stub, as chunking can cause problems with some 1291 firmware implementations. 1292 noruntime : disable EFI runtime services support 1293 nosoftreserve: The EFI_MEMORY_SP (Specific Purpose) 1294 attribute may cause the kernel to reserve the 1295 memory range for a memory mapping driver to 1296 claim. Specify efi=nosoftreserve to disable this 1297 reservation and treat the memory by its base type 1298 (i.e. EFI_CONVENTIONAL_MEMORY / "System RAM"). 1299 novamap: do not call SetVirtualAddressMap(). 1300 no_disable_early_pci_dma: Leave the busmaster bit set 1301 on all PCI bridges while in the EFI boot stub 1302 1303 efi_no_storage_paranoia [EFI; X86] 1304 Using this parameter you can use more than 50% of 1305 your efi variable storage. Use this parameter only if 1306 you are really sure that your UEFI does sane gc and 1307 fulfills the spec otherwise your board may brick. 1308 1309 efi_fake_mem= nn[KMG]@ss[KMG]:aa[,nn[KMG]@ss[KMG]:aa,..] [EFI; X86] 1310 Add arbitrary attribute to specific memory range by 1311 updating original EFI memory map. 1312 Region of memory which aa attribute is added to is 1313 from ss to ss+nn. 1314 1315 If efi_fake_mem=2G@4G:0x10000,2G@0x10a0000000:0x10000 1316 is specified, EFI_MEMORY_MORE_RELIABLE(0x10000) 1317 attribute is added to range 0x100000000-0x180000000 and 1318 0x10a0000000-0x1120000000. 1319 1320 If efi_fake_mem=8G@9G:0x40000 is specified, the 1321 EFI_MEMORY_SP(0x40000) attribute is added to 1322 range 0x240000000-0x43fffffff. 1323 1324 Using this parameter you can do debugging of EFI memmap 1325 related features. For example, you can do debugging of 1326 Address Range Mirroring feature even if your box 1327 doesn't support it, or mark specific memory as 1328 "soft reserved". 1329 1330 efivar_ssdt= [EFI; X86] Name of an EFI variable that contains an SSDT 1331 that is to be dynamically loaded by Linux. If there are 1332 multiple variables with the same name but with different 1333 vendor GUIDs, all of them will be loaded. See 1334 Documentation/admin-guide/acpi/ssdt-overlays.rst for details. 1335 1336 1337 eisa_irq_edge= [PARISC,HW] 1338 See header of drivers/parisc/eisa.c. 1339 1340 elanfreq= [X86-32] 1341 See comment before function elanfreq_setup() in 1342 arch/x86/kernel/cpu/cpufreq/elanfreq.c. 1343 1344 elfcorehdr=[size[KMG]@]offset[KMG] [IA64,PPC,SH,X86,S390] 1345 Specifies physical address of start of kernel core 1346 image elf header and optionally the size. Generally 1347 kexec loader will pass this option to capture kernel. 1348 See Documentation/admin-guide/kdump/kdump.rst for details. 1349 1350 enable_mtrr_cleanup [X86] 1351 The kernel tries to adjust MTRR layout from continuous 1352 to discrete, to make X server driver able to add WB 1353 entry later. This parameter enables that. 1354 1355 enable_timer_pin_1 [X86] 1356 Enable PIN 1 of APIC timer 1357 Can be useful to work around chipset bugs 1358 (in particular on some ATI chipsets). 1359 The kernel tries to set a reasonable default. 1360 1361 enforcing [SELINUX] Set initial enforcing status. 1362 Format: {"0" | "1"} 1363 See security/selinux/Kconfig help text. 1364 0 -- permissive (log only, no denials). 1365 1 -- enforcing (deny and log). 1366 Default value is 0. 1367 Value can be changed at runtime via 1368 /sys/fs/selinux/enforce. 1369 1370 erst_disable [ACPI] 1371 Disable Error Record Serialization Table (ERST) 1372 support. 1373 1374 ether= [HW,NET] Ethernet cards parameters 1375 This option is obsoleted by the "netdev=" option, which 1376 has equivalent usage. See its documentation for details. 1377 1378 evm= [EVM] 1379 Format: { "fix" } 1380 Permit 'security.evm' to be updated regardless of 1381 current integrity status. 1382 1383 failslab= 1384 fail_usercopy= 1385 fail_page_alloc= 1386 fail_make_request=[KNL] 1387 General fault injection mechanism. 1388 Format: <interval>,<probability>,<space>,<times> 1389 See also Documentation/fault-injection/. 1390 1391 fb_tunnels= [NET] 1392 Format: { initns | none } 1393 See Documentation/admin-guide/sysctl/net.rst for 1394 fb_tunnels_only_for_init_ns 1395 1396 floppy= [HW] 1397 See Documentation/admin-guide/blockdev/floppy.rst. 1398 1399 force_pal_cache_flush 1400 [IA-64] Avoid check_sal_cache_flush which may hang on 1401 buggy SAL_CACHE_FLUSH implementations. Using this 1402 parameter will force ia64_sal_cache_flush to call 1403 ia64_pal_cache_flush instead of SAL_CACHE_FLUSH. 1404 1405 forcepae [X86-32] 1406 Forcefully enable Physical Address Extension (PAE). 1407 Many Pentium M systems disable PAE but may have a 1408 functionally usable PAE implementation. 1409 Warning: use of this parameter will taint the kernel 1410 and may cause unknown problems. 1411 1412 ftrace=[tracer] 1413 [FTRACE] will set and start the specified tracer 1414 as early as possible in order to facilitate early 1415 boot debugging. 1416 1417 ftrace_dump_on_oops[=orig_cpu] 1418 [FTRACE] will dump the trace buffers on oops. 1419 If no parameter is passed, ftrace will dump 1420 buffers of all CPUs, but if you pass orig_cpu, it will 1421 dump only the buffer of the CPU that triggered the 1422 oops. 1423 1424 ftrace_filter=[function-list] 1425 [FTRACE] Limit the functions traced by the function 1426 tracer at boot up. function-list is a comma separated 1427 list of functions. This list can be changed at run 1428 time by the set_ftrace_filter file in the debugfs 1429 tracing directory. 1430 1431 ftrace_notrace=[function-list] 1432 [FTRACE] Do not trace the functions specified in 1433 function-list. This list can be changed at run time 1434 by the set_ftrace_notrace file in the debugfs 1435 tracing directory. 1436 1437 ftrace_graph_filter=[function-list] 1438 [FTRACE] Limit the top level callers functions traced 1439 by the function graph tracer at boot up. 1440 function-list is a comma separated list of functions 1441 that can be changed at run time by the 1442 set_graph_function file in the debugfs tracing directory. 1443 1444 ftrace_graph_notrace=[function-list] 1445 [FTRACE] Do not trace from the functions specified in 1446 function-list. This list is a comma separated list of 1447 functions that can be changed at run time by the 1448 set_graph_notrace file in the debugfs tracing directory. 1449 1450 ftrace_graph_max_depth=<uint> 1451 [FTRACE] Used with the function graph tracer. This is 1452 the max depth it will trace into a function. This value 1453 can be changed at run time by the max_graph_depth file 1454 in the tracefs tracing directory. default: 0 (no limit) 1455 1456 fw_devlink= [KNL] Create device links between consumer and supplier 1457 devices by scanning the firmware to infer the 1458 consumer/supplier relationships. This feature is 1459 especially useful when drivers are loaded as modules as 1460 it ensures proper ordering of tasks like device probing 1461 (suppliers first, then consumers), supplier boot state 1462 clean up (only after all consumers have probed), 1463 suspend/resume & runtime PM (consumers first, then 1464 suppliers). 1465 Format: { off | permissive | on | rpm } 1466 off -- Don't create device links from firmware info. 1467 permissive -- Create device links from firmware info 1468 but use it only for ordering boot state clean 1469 up (sync_state() calls). 1470 on -- Create device links from firmware info and use it 1471 to enforce probe and suspend/resume ordering. 1472 rpm -- Like "on", but also use to order runtime PM. 1473 1474 fw_devlink.strict=<bool> 1475 [KNL] Treat all inferred dependencies as mandatory 1476 dependencies. This only applies for fw_devlink=on|rpm. 1477 Format: <bool> 1478 1479 gamecon.map[2|3]= 1480 [HW,JOY] Multisystem joystick and NES/SNES/PSX pad 1481 support via parallel port (up to 5 devices per port) 1482 Format: <port#>,<pad1>,<pad2>,<pad3>,<pad4>,<pad5> 1483 See also Documentation/input/devices/joystick-parport.rst 1484 1485 gamma= [HW,DRM] 1486 1487 gart_fix_e820= [X86-64] disable the fix e820 for K8 GART 1488 Format: off | on 1489 default: on 1490 1491 gcov_persist= [GCOV] When non-zero (default), profiling data for 1492 kernel modules is saved and remains accessible via 1493 debugfs, even when the module is unloaded/reloaded. 1494 When zero, profiling data is discarded and associated 1495 debugfs files are removed at module unload time. 1496 1497 goldfish [X86] Enable the goldfish android emulator platform. 1498 Don't use this when you are not running on the 1499 android emulator 1500 1501 gpt [EFI] Forces disk with valid GPT signature but 1502 invalid Protective MBR to be treated as GPT. If the 1503 primary GPT is corrupted, it enables the backup/alternate 1504 GPT to be used instead. 1505 1506 grcan.enable0= [HW] Configuration of physical interface 0. Determines 1507 the "Enable 0" bit of the configuration register. 1508 Format: 0 | 1 1509 Default: 0 1510 grcan.enable1= [HW] Configuration of physical interface 1. Determines 1511 the "Enable 0" bit of the configuration register. 1512 Format: 0 | 1 1513 Default: 0 1514 grcan.select= [HW] Select which physical interface to use. 1515 Format: 0 | 1 1516 Default: 0 1517 grcan.txsize= [HW] Sets the size of the tx buffer. 1518 Format: <unsigned int> such that (txsize & ~0x1fffc0) == 0. 1519 Default: 1024 1520 grcan.rxsize= [HW] Sets the size of the rx buffer. 1521 Format: <unsigned int> such that (rxsize & ~0x1fffc0) == 0. 1522 Default: 1024 1523 1524 gpio-mockup.gpio_mockup_ranges 1525 [HW] Sets the ranges of gpiochip of for this device. 1526 Format: <start1>,<end1>,<start2>,<end2>... 1527 1528 hardlockup_all_cpu_backtrace= 1529 [KNL] Should the hard-lockup detector generate 1530 backtraces on all cpus. 1531 Format: 0 | 1 1532 1533 hashdist= [KNL,NUMA] Large hashes allocated during boot 1534 are distributed across NUMA nodes. Defaults on 1535 for 64-bit NUMA, off otherwise. 1536 Format: 0 | 1 (for off | on) 1537 1538 hcl= [IA-64] SGI's Hardware Graph compatibility layer 1539 1540 hd= [EIDE] (E)IDE hard drive subsystem geometry 1541 Format: <cyl>,<head>,<sect> 1542 1543 hest_disable [ACPI] 1544 Disable Hardware Error Source Table (HEST) support; 1545 corresponding firmware-first mode error processing 1546 logic will be disabled. 1547 1548 highmem=nn[KMG] [KNL,BOOT] forces the highmem zone to have an exact 1549 size of <nn>. This works even on boxes that have no 1550 highmem otherwise. This also works to reduce highmem 1551 size on bigger boxes. 1552 1553 highres= [KNL] Enable/disable high resolution timer mode. 1554 Valid parameters: "on", "off" 1555 Default: "on" 1556 1557 hlt [BUGS=ARM,SH] 1558 1559 hpet= [X86-32,HPET] option to control HPET usage 1560 Format: { enable (default) | disable | force | 1561 verbose } 1562 disable: disable HPET and use PIT instead 1563 force: allow force enabled of undocumented chips (ICH4, 1564 VIA, nVidia) 1565 verbose: show contents of HPET registers during setup 1566 1567 hpet_mmap= [X86, HPET_MMAP] Allow userspace to mmap HPET 1568 registers. Default set by CONFIG_HPET_MMAP_DEFAULT. 1569 1570 hugetlb_cma= [HW] The size of a cma area used for allocation 1571 of gigantic hugepages. 1572 Format: nn[KMGTPE] 1573 1574 Reserve a cma area of given size and allocate gigantic 1575 hugepages using the cma allocator. If enabled, the 1576 boot-time allocation of gigantic hugepages is skipped. 1577 1578 hugepages= [HW] Number of HugeTLB pages to allocate at boot. 1579 If this follows hugepagesz (below), it specifies 1580 the number of pages of hugepagesz to be allocated. 1581 If this is the first HugeTLB parameter on the command 1582 line, it specifies the number of pages to allocate for 1583 the default huge page size. See also 1584 Documentation/admin-guide/mm/hugetlbpage.rst. 1585 Format: <integer> 1586 1587 hugepagesz= 1588 [HW] The size of the HugeTLB pages. This is used in 1589 conjunction with hugepages (above) to allocate huge 1590 pages of a specific size at boot. The pair 1591 hugepagesz=X hugepages=Y can be specified once for 1592 each supported huge page size. Huge page sizes are 1593 architecture dependent. See also 1594 Documentation/admin-guide/mm/hugetlbpage.rst. 1595 Format: size[KMG] 1596 1597 hung_task_panic= 1598 [KNL] Should the hung task detector generate panics. 1599 Format: 0 | 1 1600 1601 A value of 1 instructs the kernel to panic when a 1602 hung task is detected. The default value is controlled 1603 by the CONFIG_BOOTPARAM_HUNG_TASK_PANIC build-time 1604 option. The value selected by this boot parameter can 1605 be changed later by the kernel.hung_task_panic sysctl. 1606 1607 hvc_iucv= [S390] Number of z/VM IUCV hypervisor console (HVC) 1608 terminal devices. Valid values: 0..8 1609 hvc_iucv_allow= [S390] Comma-separated list of z/VM user IDs. 1610 If specified, z/VM IUCV HVC accepts connections 1611 from listed z/VM user IDs only. 1612 1613 hvc_dcc.enable= [ARM,ARM64] Enable DCC driver at runtime. For GKI, 1614 disabled at runtime by default to prevent 1615 crashes in devices which do not support DCC. 1616 1617 hv_nopvspin [X86,HYPER_V] Disables the paravirt spinlock optimizations 1618 which allow the hypervisor to 'idle' the 1619 guest on lock contention. 1620 1621 keep_bootcon [KNL] 1622 Do not unregister boot console at start. This is only 1623 useful for debugging when something happens in the window 1624 between unregistering the boot console and initializing 1625 the real console. 1626 1627 i2c_bus= [HW] Override the default board specific I2C bus speed 1628 or register an additional I2C bus that is not 1629 registered from board initialization code. 1630 Format: 1631 <bus_id>,<clkrate> 1632 1633 i8042.debug [HW] Toggle i8042 debug mode 1634 i8042.unmask_kbd_data 1635 [HW] Enable printing of interrupt data from the KBD port 1636 (disabled by default, and as a pre-condition 1637 requires that i8042.debug=1 be enabled) 1638 i8042.direct [HW] Put keyboard port into non-translated mode 1639 i8042.dumbkbd [HW] Pretend that controller can only read data from 1640 keyboard and cannot control its state 1641 (Don't attempt to blink the leds) 1642 i8042.noaux [HW] Don't check for auxiliary (== mouse) port 1643 i8042.nokbd [HW] Don't check/create keyboard port 1644 i8042.noloop [HW] Disable the AUX Loopback command while probing 1645 for the AUX port 1646 i8042.nomux [HW] Don't check presence of an active multiplexing 1647 controller 1648 i8042.nopnp [HW] Don't use ACPIPnP / PnPBIOS to discover KBD/AUX 1649 controllers 1650 i8042.notimeout [HW] Ignore timeout condition signalled by controller 1651 i8042.reset [HW] Reset the controller during init, cleanup and 1652 suspend-to-ram transitions, only during s2r 1653 transitions, or never reset 1654 Format: { 1 | Y | y | 0 | N | n } 1655 1, Y, y: always reset controller 1656 0, N, n: don't ever reset controller 1657 Default: only on s2r transitions on x86; most other 1658 architectures force reset to be always executed 1659 i8042.unlock [HW] Unlock (ignore) the keylock 1660 i8042.kbdreset [HW] Reset device connected to KBD port 1661 i8042.probe_defer 1662 [HW] Allow deferred probing upon i8042 probe errors 1663 1664 i810= [HW,DRM] 1665 1666 i8k.ignore_dmi [HW] Continue probing hardware even if DMI data 1667 indicates that the driver is running on unsupported 1668 hardware. 1669 i8k.force [HW] Activate i8k driver even if SMM BIOS signature 1670 does not match list of supported models. 1671 i8k.power_status 1672 [HW] Report power status in /proc/i8k 1673 (disabled by default) 1674 i8k.restricted [HW] Allow controlling fans only if SYS_ADMIN 1675 capability is set. 1676 1677 i915.invert_brightness= 1678 [DRM] Invert the sense of the variable that is used to 1679 set the brightness of the panel backlight. Normally a 1680 brightness value of 0 indicates backlight switched off, 1681 and the maximum of the brightness value sets the backlight 1682 to maximum brightness. If this parameter is set to 0 1683 (default) and the machine requires it, or this parameter 1684 is set to 1, a brightness value of 0 sets the backlight 1685 to maximum brightness, and the maximum of the brightness 1686 value switches the backlight off. 1687 -1 -- never invert brightness 1688 0 -- machine default 1689 1 -- force brightness inversion 1690 1691 icn= [HW,ISDN] 1692 Format: <io>[,<membase>[,<icn_id>[,<icn_id2>]]] 1693 1694 ide-core.nodma= [HW] (E)IDE subsystem 1695 Format: =0.0 to prevent dma on hda, =0.1 hdb =1.0 hdc 1696 .vlb_clock .pci_clock .noflush .nohpa .noprobe .nowerr 1697 .cdrom .chs .ignore_cable are additional options 1698 See Documentation/ide/ide.rst. 1699 1700 ide-generic.probe-mask= [HW] (E)IDE subsystem 1701 Format: <int> 1702 Probe mask for legacy ISA IDE ports. Depending on 1703 platform up to 6 ports are supported, enabled by 1704 setting corresponding bits in the mask to 1. The 1705 default value is 0x0, which has a special meaning. 1706 On systems that have PCI, it triggers scanning the 1707 PCI bus for the first and the second port, which 1708 are then probed. On systems without PCI the value 1709 of 0x0 enables probing the two first ports as if it 1710 was 0x3. 1711 1712 ide-pci-generic.all-generic-ide [HW] (E)IDE subsystem 1713 Claim all unknown PCI IDE storage controllers. 1714 1715 idle= [X86] 1716 Format: idle=poll, idle=halt, idle=nomwait 1717 Poll forces a polling idle loop that can slightly 1718 improve the performance of waking up a idle CPU, but 1719 will use a lot of power and make the system run hot. 1720 Not recommended. 1721 idle=halt: Halt is forced to be used for CPU idle. 1722 In such case C2/C3 won't be used again. 1723 idle=nomwait: Disable mwait for CPU C-states 1724 1725 ieee754= [MIPS] Select IEEE Std 754 conformance mode 1726 Format: { strict | legacy | 2008 | relaxed } 1727 Default: strict 1728 1729 Choose which programs will be accepted for execution 1730 based on the IEEE 754 NaN encoding(s) supported by 1731 the FPU and the NaN encoding requested with the value 1732 of an ELF file header flag individually set by each 1733 binary. Hardware implementations are permitted to 1734 support either or both of the legacy and the 2008 NaN 1735 encoding mode. 1736 1737 Available settings are as follows: 1738 strict accept binaries that request a NaN encoding 1739 supported by the FPU 1740 legacy only accept legacy-NaN binaries, if supported 1741 by the FPU 1742 2008 only accept 2008-NaN binaries, if supported 1743 by the FPU 1744 relaxed accept any binaries regardless of whether 1745 supported by the FPU 1746 1747 The FPU emulator is always able to support both NaN 1748 encodings, so if no FPU hardware is present or it has 1749 been disabled with 'nofpu', then the settings of 1750 'legacy' and '2008' strap the emulator accordingly, 1751 'relaxed' straps the emulator for both legacy-NaN and 1752 2008-NaN, whereas 'strict' enables legacy-NaN only on 1753 legacy processors and both NaN encodings on MIPS32 or 1754 MIPS64 CPUs. 1755 1756 The setting for ABS.fmt/NEG.fmt instruction execution 1757 mode generally follows that for the NaN encoding, 1758 except where unsupported by hardware. 1759 1760 ignore_loglevel [KNL] 1761 Ignore loglevel setting - this will print /all/ 1762 kernel messages to the console. Useful for debugging. 1763 We also add it as printk module parameter, so users 1764 could change it dynamically, usually by 1765 /sys/module/printk/parameters/ignore_loglevel. 1766 1767 ignore_rlimit_data 1768 Ignore RLIMIT_DATA setting for data mappings, 1769 print warning at first misuse. Can be changed via 1770 /sys/module/kernel/parameters/ignore_rlimit_data. 1771 1772 ihash_entries= [KNL] 1773 Set number of hash buckets for inode cache. 1774 1775 ima_appraise= [IMA] appraise integrity measurements 1776 Format: { "off" | "enforce" | "fix" | "log" } 1777 default: "enforce" 1778 1779 ima_appraise_tcb [IMA] Deprecated. Use ima_policy= instead. 1780 The builtin appraise policy appraises all files 1781 owned by uid=0. 1782 1783 ima_canonical_fmt [IMA] 1784 Use the canonical format for the binary runtime 1785 measurements, instead of host native format. 1786 1787 ima_hash= [IMA] 1788 Format: { md5 | sha1 | rmd160 | sha256 | sha384 1789 | sha512 | ... } 1790 default: "sha1" 1791 1792 The list of supported hash algorithms is defined 1793 in crypto/hash_info.h. 1794 1795 ima_policy= [IMA] 1796 The builtin policies to load during IMA setup. 1797 Format: "tcb | appraise_tcb | secure_boot | 1798 fail_securely" 1799 1800 The "tcb" policy measures all programs exec'd, files 1801 mmap'd for exec, and all files opened with the read 1802 mode bit set by either the effective uid (euid=0) or 1803 uid=0. 1804 1805 The "appraise_tcb" policy appraises the integrity of 1806 all files owned by root. 1807 1808 The "secure_boot" policy appraises the integrity 1809 of files (eg. kexec kernel image, kernel modules, 1810 firmware, policy, etc) based on file signatures. 1811 1812 The "fail_securely" policy forces file signature 1813 verification failure also on privileged mounted 1814 filesystems with the SB_I_UNVERIFIABLE_SIGNATURE 1815 flag. 1816 1817 ima_tcb [IMA] Deprecated. Use ima_policy= instead. 1818 Load a policy which meets the needs of the Trusted 1819 Computing Base. This means IMA will measure all 1820 programs exec'd, files mmap'd for exec, and all files 1821 opened for read by uid=0. 1822 1823 ima_template= [IMA] 1824 Select one of defined IMA measurements template formats. 1825 Formats: { "ima" | "ima-ng" | "ima-sig" } 1826 Default: "ima-ng" 1827 1828 ima_template_fmt= 1829 [IMA] Define a custom template format. 1830 Format: { "field1|...|fieldN" } 1831 1832 ima.ahash_minsize= [IMA] Minimum file size for asynchronous hash usage 1833 Format: <min_file_size> 1834 Set the minimal file size for using asynchronous hash. 1835 If left unspecified, ahash usage is disabled. 1836 1837 ahash performance varies for different data sizes on 1838 different crypto accelerators. This option can be used 1839 to achieve the best performance for a particular HW. 1840 1841 ima.ahash_bufsize= [IMA] Asynchronous hash buffer size 1842 Format: <bufsize> 1843 Set hashing buffer size. Default: 4k. 1844 1845 ahash performance varies for different chunk sizes on 1846 different crypto accelerators. This option can be used 1847 to achieve best performance for particular HW. 1848 1849 init= [KNL] 1850 Format: <full_path> 1851 Run specified binary instead of /sbin/init as init 1852 process. 1853 1854 initcall_debug [KNL] Trace initcalls as they are executed. Useful 1855 for working out where the kernel is dying during 1856 startup. 1857 1858 initcall_blacklist= [KNL] Do not execute a comma-separated list of 1859 initcall functions. Useful for debugging built-in 1860 modules and initcalls. 1861 1862 initrd= [BOOT] Specify the location of the initial ramdisk 1863 1864 initrdmem= [KNL] Specify a physical address and size from which to 1865 load the initrd. If an initrd is compiled in or 1866 specified in the bootparams, it takes priority over this 1867 setting. 1868 Format: ss[KMG],nn[KMG] 1869 Default is 0, 0 1870 1871 init_on_alloc= [MM] Fill newly allocated pages and heap objects with 1872 zeroes. 1873 Format: 0 | 1 1874 Default set by CONFIG_INIT_ON_ALLOC_DEFAULT_ON. 1875 1876 init_on_free= [MM] Fill freed pages and heap objects with zeroes. 1877 Format: 0 | 1 1878 Default set by CONFIG_INIT_ON_FREE_DEFAULT_ON. 1879 1880 init_pkru= [X86] Specify the default memory protection keys rights 1881 register contents for all processes. 0x55555554 by 1882 default (disallow access to all but pkey 0). Can 1883 override in debugfs after boot. 1884 1885 inport.irq= [HW] Inport (ATI XL and Microsoft) busmouse driver 1886 Format: <irq> 1887 1888 int_pln_enable [X86] Enable power limit notification interrupt 1889 1890 integrity_audit=[IMA] 1891 Format: { "0" | "1" } 1892 0 -- basic integrity auditing messages. (Default) 1893 1 -- additional integrity auditing messages. 1894 1895 intel_iommu= [DMAR] Intel IOMMU driver (DMAR) option 1896 on 1897 Enable intel iommu driver. 1898 off 1899 Disable intel iommu driver. 1900 igfx_off [Default Off] 1901 By default, gfx is mapped as normal device. If a gfx 1902 device has a dedicated DMAR unit, the DMAR unit is 1903 bypassed by not enabling DMAR with this option. In 1904 this case, gfx device will use physical address for 1905 DMA. 1906 forcedac [X86-64] 1907 With this option iommu will not optimize to look 1908 for io virtual address below 32-bit forcing dual 1909 address cycle on pci bus for cards supporting greater 1910 than 32-bit addressing. The default is to look 1911 for translation below 32-bit and if not available 1912 then look in the higher range. 1913 strict [Default Off] 1914 With this option on every unmap_single operation will 1915 result in a hardware IOTLB flush operation as opposed 1916 to batching them for performance. 1917 sp_off [Default Off] 1918 By default, super page will be supported if Intel IOMMU 1919 has the capability. With this option, super page will 1920 not be supported. 1921 sm_on [Default Off] 1922 By default, scalable mode will be disabled even if the 1923 hardware advertises that it has support for the scalable 1924 mode translation. With this option set, scalable mode 1925 will be used on hardware which claims to support it. 1926 tboot_noforce [Default Off] 1927 Do not force the Intel IOMMU enabled under tboot. 1928 By default, tboot will force Intel IOMMU on, which 1929 could harm performance of some high-throughput 1930 devices like 40GBit network cards, even if identity 1931 mapping is enabled. 1932 Note that using this option lowers the security 1933 provided by tboot because it makes the system 1934 vulnerable to DMA attacks. 1935 nobounce [Default off] 1936 Disable bounce buffer for untrusted devices such as 1937 the Thunderbolt devices. This will treat the untrusted 1938 devices as the trusted ones, hence might expose security 1939 risks of DMA attacks. 1940 1941 intel_idle.max_cstate= [KNL,HW,ACPI,X86] 1942 0 disables intel_idle and fall back on acpi_idle. 1943 1 to 9 specify maximum depth of C-state. 1944 1945 intel_pstate= [X86] 1946 disable 1947 Do not enable intel_pstate as the default 1948 scaling driver for the supported processors 1949 passive 1950 Use intel_pstate as a scaling driver, but configure it 1951 to work with generic cpufreq governors (instead of 1952 enabling its internal governor). This mode cannot be 1953 used along with the hardware-managed P-states (HWP) 1954 feature. 1955 force 1956 Enable intel_pstate on systems that prohibit it by default 1957 in favor of acpi-cpufreq. Forcing the intel_pstate driver 1958 instead of acpi-cpufreq may disable platform features, such 1959 as thermal controls and power capping, that rely on ACPI 1960 P-States information being indicated to OSPM and therefore 1961 should be used with caution. This option does not work with 1962 processors that aren't supported by the intel_pstate driver 1963 or on platforms that use pcc-cpufreq instead of acpi-cpufreq. 1964 no_hwp 1965 Do not enable hardware P state control (HWP) 1966 if available. 1967 hwp_only 1968 Only load intel_pstate on systems which support 1969 hardware P state control (HWP) if available. 1970 support_acpi_ppc 1971 Enforce ACPI _PPC performance limits. If the Fixed ACPI 1972 Description Table, specifies preferred power management 1973 profile as "Enterprise Server" or "Performance Server", 1974 then this feature is turned on by default. 1975 per_cpu_perf_limits 1976 Allow per-logical-CPU P-State performance control limits using 1977 cpufreq sysfs interface 1978 1979 intremap= [X86-64, Intel-IOMMU] 1980 on enable Interrupt Remapping (default) 1981 off disable Interrupt Remapping 1982 nosid disable Source ID checking 1983 no_x2apic_optout 1984 BIOS x2APIC opt-out request will be ignored 1985 nopost disable Interrupt Posting 1986 1987 iomem= Disable strict checking of access to MMIO memory 1988 strict regions from userspace. 1989 relaxed 1990 1991 iommu= [X86] 1992 off 1993 force 1994 noforce 1995 biomerge 1996 panic 1997 nopanic 1998 merge 1999 nomerge 2000 soft 2001 pt [X86] 2002 nopt [X86] 2003 nobypass [PPC/POWERNV] 2004 Disable IOMMU bypass, using IOMMU for PCI devices. 2005 2006 iommu.strict= [ARM64] Configure TLB invalidation behaviour 2007 Format: { "0" | "1" } 2008 0 - Lazy mode. 2009 Request that DMA unmap operations use deferred 2010 invalidation of hardware TLBs, for increased 2011 throughput at the cost of reduced device isolation. 2012 Will fall back to strict mode if not supported by 2013 the relevant IOMMU driver. 2014 1 - Strict mode (default). 2015 DMA unmap operations invalidate IOMMU hardware TLBs 2016 synchronously. 2017 2018 iommu.passthrough= 2019 [ARM64, X86] Configure DMA to bypass the IOMMU by default. 2020 Format: { "0" | "1" } 2021 0 - Use IOMMU translation for DMA. 2022 1 - Bypass the IOMMU for DMA. 2023 unset - Use value of CONFIG_IOMMU_DEFAULT_PASSTHROUGH. 2024 2025 io7= [HW] IO7 for Marvel-based Alpha systems 2026 See comment before marvel_specify_io7 in 2027 arch/alpha/kernel/core_marvel.c. 2028 2029 io_delay= [X86] I/O delay method 2030 0x80 2031 Standard port 0x80 based delay 2032 0xed 2033 Alternate port 0xed based delay (needed on some systems) 2034 udelay 2035 Simple two microseconds delay 2036 none 2037 No delay 2038 2039 ip= [IP_PNP] 2040 See Documentation/admin-guide/nfs/nfsroot.rst. 2041 2042 ipcmni_extend [KNL] Extend the maximum number of unique System V 2043 IPC identifiers from 32,768 to 16,777,216. 2044 2045 irqaffinity= [SMP] Set the default irq affinity mask 2046 The argument is a cpu list, as described above. 2047 2048 irqchip.gicv2_force_probe= 2049 [ARM, ARM64] 2050 Format: <bool> 2051 Force the kernel to look for the second 4kB page 2052 of a GICv2 controller even if the memory range 2053 exposed by the device tree is too small. 2054 2055 irqchip.gicv3_nolpi= 2056 [ARM, ARM64] 2057 Force the kernel to ignore the availability of 2058 LPIs (and by consequence ITSs). Intended for system 2059 that use the kernel as a bootloader, and thus want 2060 to let secondary kernels in charge of setting up 2061 LPIs. 2062 2063 irqchip.gicv3_pseudo_nmi= [ARM64] 2064 Enables support for pseudo-NMIs in the kernel. This 2065 requires the kernel to be built with 2066 CONFIG_ARM64_PSEUDO_NMI. 2067 2068 irqfixup [HW] 2069 When an interrupt is not handled search all handlers 2070 for it. Intended to get systems with badly broken 2071 firmware running. 2072 2073 irqpoll [HW] 2074 When an interrupt is not handled search all handlers 2075 for it. Also check all handlers each timer 2076 interrupt. Intended to get systems with badly broken 2077 firmware running. 2078 2079 isapnp= [ISAPNP] 2080 Format: <RDP>,<reset>,<pci_scan>,<verbosity> 2081 2082 isolcpus= [KNL,SMP,ISOL] Isolate a given set of CPUs from disturbance. 2083 [Deprecated - use cpusets instead] 2084 Format: [flag-list,]<cpu-list> 2085 2086 Specify one or more CPUs to isolate from disturbances 2087 specified in the flag list (default: domain): 2088 2089 nohz 2090 Disable the tick when a single task runs. 2091 2092 A residual 1Hz tick is offloaded to workqueues, which you 2093 need to affine to housekeeping through the global 2094 workqueue's affinity configured via the 2095 /sys/devices/virtual/workqueue/cpumask sysfs file, or 2096 by using the 'domain' flag described below. 2097 2098 NOTE: by default the global workqueue runs on all CPUs, 2099 so to protect individual CPUs the 'cpumask' file has to 2100 be configured manually after bootup. 2101 2102 domain 2103 Isolate from the general SMP balancing and scheduling 2104 algorithms. Note that performing domain isolation this way 2105 is irreversible: it's not possible to bring back a CPU to 2106 the domains once isolated through isolcpus. It's strongly 2107 advised to use cpusets instead to disable scheduler load 2108 balancing through the "cpuset.sched_load_balance" file. 2109 It offers a much more flexible interface where CPUs can 2110 move in and out of an isolated set anytime. 2111 2112 You can move a process onto or off an "isolated" CPU via 2113 the CPU affinity syscalls or cpuset. 2114 <cpu number> begins at 0 and the maximum value is 2115 "number of CPUs in system - 1". 2116 2117 managed_irq 2118 2119 Isolate from being targeted by managed interrupts 2120 which have an interrupt mask containing isolated 2121 CPUs. The affinity of managed interrupts is 2122 handled by the kernel and cannot be changed via 2123 the /proc/irq/* interfaces. 2124 2125 This isolation is best effort and only effective 2126 if the automatically assigned interrupt mask of a 2127 device queue contains isolated and housekeeping 2128 CPUs. If housekeeping CPUs are online then such 2129 interrupts are directed to the housekeeping CPU 2130 so that IO submitted on the housekeeping CPU 2131 cannot disturb the isolated CPU. 2132 2133 If a queue's affinity mask contains only isolated 2134 CPUs then this parameter has no effect on the 2135 interrupt routing decision, though interrupts are 2136 only delivered when tasks running on those 2137 isolated CPUs submit IO. IO submitted on 2138 housekeeping CPUs has no influence on those 2139 queues. 2140 2141 The format of <cpu-list> is described above. 2142 2143 iucv= [HW,NET] 2144 2145 ivrs_ioapic [HW,X86-64] 2146 Provide an override to the IOAPIC-ID<->DEVICE-ID 2147 mapping provided in the IVRS ACPI table. For 2148 example, to map IOAPIC-ID decimal 10 to 2149 PCI device 00:14.0 write the parameter as: 2150 ivrs_ioapic[10]=00:14.0 2151 2152 ivrs_hpet [HW,X86-64] 2153 Provide an override to the HPET-ID<->DEVICE-ID 2154 mapping provided in the IVRS ACPI table. For 2155 example, to map HPET-ID decimal 0 to 2156 PCI device 00:14.0 write the parameter as: 2157 ivrs_hpet[0]=00:14.0 2158 2159 ivrs_acpihid [HW,X86-64] 2160 Provide an override to the ACPI-HID:UID<->DEVICE-ID 2161 mapping provided in the IVRS ACPI table. For 2162 example, to map UART-HID:UID AMD0020:0 to 2163 PCI device 00:14.5 write the parameter as: 2164 ivrs_acpihid[00:14.5]=AMD0020:0 2165 2166 js= [HW,JOY] Analog joystick 2167 See Documentation/input/joydev/joystick.rst. 2168 2169 nokaslr [KNL] 2170 When CONFIG_RANDOMIZE_BASE is set, this disables 2171 kernel and module base offset ASLR (Address Space 2172 Layout Randomization). 2173 2174 kasan_multi_shot 2175 [KNL] Enforce KASAN (Kernel Address Sanitizer) to print 2176 report on every invalid memory access. Without this 2177 parameter KASAN will print report only for the first 2178 invalid access. 2179 2180 keepinitrd [HW,ARM] 2181 2182 kernelcore= [KNL,X86,IA-64,PPC] 2183 Format: nn[KMGTPE] | nn% | "mirror" 2184 This parameter specifies the amount of memory usable by 2185 the kernel for non-movable allocations. The requested 2186 amount is spread evenly throughout all nodes in the 2187 system as ZONE_NORMAL. The remaining memory is used for 2188 movable memory in its own zone, ZONE_MOVABLE. In the 2189 event, a node is too small to have both ZONE_NORMAL and 2190 ZONE_MOVABLE, kernelcore memory will take priority and 2191 other nodes will have a larger ZONE_MOVABLE. 2192 2193 ZONE_MOVABLE is used for the allocation of pages that 2194 may be reclaimed or moved by the page migration 2195 subsystem. Note that allocations like PTEs-from-HighMem 2196 still use the HighMem zone if it exists, and the Normal 2197 zone if it does not. 2198 2199 It is possible to specify the exact amount of memory in 2200 the form of "nn[KMGTPE]", a percentage of total system 2201 memory in the form of "nn%", or "mirror". If "mirror" 2202 option is specified, mirrored (reliable) memory is used 2203 for non-movable allocations and remaining memory is used 2204 for Movable pages. "nn[KMGTPE]", "nn%", and "mirror" 2205 are exclusive, so you cannot specify multiple forms. 2206 2207 kgdbdbgp= [KGDB,HW] kgdb over EHCI usb debug port. 2208 Format: <Controller#>[,poll interval] 2209 The controller # is the number of the ehci usb debug 2210 port as it is probed via PCI. The poll interval is 2211 optional and is the number seconds in between 2212 each poll cycle to the debug port in case you need 2213 the functionality for interrupting the kernel with 2214 gdb or control-c on the dbgp connection. When 2215 not using this parameter you use sysrq-g to break into 2216 the kernel debugger. 2217 2218 kgdboc= [KGDB,HW] kgdb over consoles. 2219 Requires a tty driver that supports console polling, 2220 or a supported polling keyboard driver (non-usb). 2221 Serial only format: <serial_device>[,baud] 2222 keyboard only format: kbd 2223 keyboard and serial format: kbd,<serial_device>[,baud] 2224 Optional Kernel mode setting: 2225 kms, kbd format: kms,kbd 2226 kms, kbd and serial format: kms,kbd,<ser_dev>[,baud] 2227 2228 kgdboc_earlycon= [KGDB,HW] 2229 If the boot console provides the ability to read 2230 characters and can work in polling mode, you can use 2231 this parameter to tell kgdb to use it as a backend 2232 until the normal console is registered. Intended to 2233 be used together with the kgdboc parameter which 2234 specifies the normal console to transition to. 2235 2236 The name of the early console should be specified 2237 as the value of this parameter. Note that the name of 2238 the early console might be different than the tty 2239 name passed to kgdboc. It's OK to leave the value 2240 blank and the first boot console that implements 2241 read() will be picked. 2242 2243 kgdbwait [KGDB] Stop kernel execution and enter the 2244 kernel debugger at the earliest opportunity. 2245 2246 kmac= [MIPS] Korina ethernet MAC address. 2247 Configure the RouterBoard 532 series on-chip 2248 Ethernet adapter MAC address. 2249 2250 kmemleak= [KNL] Boot-time kmemleak enable/disable 2251 Valid arguments: on, off 2252 Default: on 2253 Built with CONFIG_DEBUG_KMEMLEAK_DEFAULT_OFF=y, 2254 the default is off. 2255 2256 kprobe_event=[probe-list] 2257 [FTRACE] Add kprobe events and enable at boot time. 2258 The probe-list is a semicolon delimited list of probe 2259 definitions. Each definition is same as kprobe_events 2260 interface, but the parameters are comma delimited. 2261 For example, to add a kprobe event on vfs_read with 2262 arg1 and arg2, add to the command line; 2263 2264 kprobe_event=p,vfs_read,$arg1,$arg2 2265 2266 See also Documentation/trace/kprobetrace.rst "Kernel 2267 Boot Parameter" section. 2268 2269 kpti= [ARM64] Control page table isolation of user 2270 and kernel address spaces. 2271 Default: enabled on cores which need mitigation. 2272 0: force disabled 2273 1: force enabled 2274 2275 kvm.ignore_msrs=[KVM] Ignore guest accesses to unhandled MSRs. 2276 Default is 0 (don't ignore, but inject #GP) 2277 2278 kvm.enable_vmware_backdoor=[KVM] Support VMware backdoor PV interface. 2279 Default is false (don't support). 2280 2281 kvm.mmu_audit= [KVM] This is a R/W parameter which allows audit 2282 KVM MMU at runtime. 2283 Default is 0 (off) 2284 2285 kvm.nx_huge_pages= 2286 [KVM] Controls the software workaround for the 2287 X86_BUG_ITLB_MULTIHIT bug. 2288 force : Always deploy workaround. 2289 off : Never deploy workaround. 2290 auto : Deploy workaround based on the presence of 2291 X86_BUG_ITLB_MULTIHIT. 2292 2293 Default is 'auto'. 2294 2295 If the software workaround is enabled for the host, 2296 guests do need not to enable it for nested guests. 2297 2298 kvm.nx_huge_pages_recovery_ratio= 2299 [KVM] Controls how many 4KiB pages are periodically zapped 2300 back to huge pages. 0 disables the recovery, otherwise if 2301 the value is N KVM will zap 1/Nth of the 4KiB pages every 2302 minute. The default is 60. 2303 2304 kvm-amd.nested= [KVM,AMD] Allow nested virtualization in KVM/SVM. 2305 Default is 1 (enabled) 2306 2307 kvm-amd.npt= [KVM,AMD] Disable nested paging (virtualized MMU) 2308 for all guests. 2309 Default is 1 (enabled) if in 64-bit or 32-bit PAE mode. 2310 2311 kvm-arm.mode= 2312 [KVM,ARM] Select one of KVM/arm64's modes of operation. 2313 2314 nvhe: Standard nVHE-based mode, without support for 2315 protected guests. 2316 2317 protected: nVHE-based mode with support for guests whose 2318 state is kept private from the host. 2319 Not valid if the kernel is running in EL2. 2320 2321 Defaults to VHE/nVHE based on hardware support and 2322 the value of CONFIG_ARM64_VHE. 2323 2324 kvm-arm.vgic_v3_group0_trap= 2325 [KVM,ARM] Trap guest accesses to GICv3 group-0 2326 system registers 2327 2328 kvm-arm.vgic_v3_group1_trap= 2329 [KVM,ARM] Trap guest accesses to GICv3 group-1 2330 system registers 2331 2332 kvm-arm.vgic_v3_common_trap= 2333 [KVM,ARM] Trap guest accesses to GICv3 common 2334 system registers 2335 2336 kvm-arm.vgic_v4_enable= 2337 [KVM,ARM] Allow use of GICv4 for direct injection of 2338 LPIs. 2339 2340 kvm_cma_resv_ratio=n [PPC] 2341 Reserves given percentage from system memory area for 2342 contiguous memory allocation for KVM hash pagetable 2343 allocation. 2344 By default it reserves 5% of total system memory. 2345 Format: <integer> 2346 Default: 5 2347 2348 kvm-intel.ept= [KVM,Intel] Disable extended page tables 2349 (virtualized MMU) support on capable Intel chips. 2350 Default is 1 (enabled) 2351 2352 kvm-intel.emulate_invalid_guest_state= 2353 [KVM,Intel] Disable emulation of invalid guest state. 2354 Ignored if kvm-intel.enable_unrestricted_guest=1, as 2355 guest state is never invalid for unrestricted guests. 2356 This param doesn't apply to nested guests (L2), as KVM 2357 never emulates invalid L2 guest state. 2358 Default is 1 (enabled) 2359 2360 kvm-intel.flexpriority= 2361 [KVM,Intel] Disable FlexPriority feature (TPR shadow). 2362 Default is 1 (enabled) 2363 2364 kvm-intel.nested= 2365 [KVM,Intel] Enable VMX nesting (nVMX). 2366 Default is 0 (disabled) 2367 2368 kvm-intel.unrestricted_guest= 2369 [KVM,Intel] Disable unrestricted guest feature 2370 (virtualized real and unpaged mode) on capable 2371 Intel chips. Default is 1 (enabled) 2372 2373 kvm-intel.vmentry_l1d_flush=[KVM,Intel] Mitigation for L1 Terminal Fault 2374 CVE-2018-3620. 2375 2376 Valid arguments: never, cond, always 2377 2378 always: L1D cache flush on every VMENTER. 2379 cond: Flush L1D on VMENTER only when the code between 2380 VMEXIT and VMENTER can leak host memory. 2381 never: Disables the mitigation 2382 2383 Default is cond (do L1 cache flush in specific instances) 2384 2385 kvm-intel.vpid= [KVM,Intel] Disable Virtual Processor Identification 2386 feature (tagged TLBs) on capable Intel chips. 2387 Default is 1 (enabled) 2388 2389 l1tf= [X86] Control mitigation of the L1TF vulnerability on 2390 affected CPUs 2391 2392 The kernel PTE inversion protection is unconditionally 2393 enabled and cannot be disabled. 2394 2395 full 2396 Provides all available mitigations for the 2397 L1TF vulnerability. Disables SMT and 2398 enables all mitigations in the 2399 hypervisors, i.e. unconditional L1D flush. 2400 2401 SMT control and L1D flush control via the 2402 sysfs interface is still possible after 2403 boot. Hypervisors will issue a warning 2404 when the first VM is started in a 2405 potentially insecure configuration, 2406 i.e. SMT enabled or L1D flush disabled. 2407 2408 full,force 2409 Same as 'full', but disables SMT and L1D 2410 flush runtime control. Implies the 2411 'nosmt=force' command line option. 2412 (i.e. sysfs control of SMT is disabled.) 2413 2414 flush 2415 Leaves SMT enabled and enables the default 2416 hypervisor mitigation, i.e. conditional 2417 L1D flush. 2418 2419 SMT control and L1D flush control via the 2420 sysfs interface is still possible after 2421 boot. Hypervisors will issue a warning 2422 when the first VM is started in a 2423 potentially insecure configuration, 2424 i.e. SMT enabled or L1D flush disabled. 2425 2426 flush,nosmt 2427 2428 Disables SMT and enables the default 2429 hypervisor mitigation. 2430 2431 SMT control and L1D flush control via the 2432 sysfs interface is still possible after 2433 boot. Hypervisors will issue a warning 2434 when the first VM is started in a 2435 potentially insecure configuration, 2436 i.e. SMT enabled or L1D flush disabled. 2437 2438 flush,nowarn 2439 Same as 'flush', but hypervisors will not 2440 warn when a VM is started in a potentially 2441 insecure configuration. 2442 2443 off 2444 Disables hypervisor mitigations and doesn't 2445 emit any warnings. 2446 It also drops the swap size and available 2447 RAM limit restriction on both hypervisor and 2448 bare metal. 2449 2450 Default is 'flush'. 2451 2452 For details see: Documentation/admin-guide/hw-vuln/l1tf.rst 2453 2454 l2cr= [PPC] 2455 2456 l3cr= [PPC] 2457 2458 lapic [X86-32,APIC] Enable the local APIC even if BIOS 2459 disabled it. 2460 2461 lapic= [X86,APIC] Do not use TSC deadline 2462 value for LAPIC timer one-shot implementation. Default 2463 back to the programmable timer unit in the LAPIC. 2464 Format: notscdeadline 2465 2466 lapic_timer_c2_ok [X86,APIC] trust the local apic timer 2467 in C2 power state. 2468 2469 libata.dma= [LIBATA] DMA control 2470 libata.dma=0 Disable all PATA and SATA DMA 2471 libata.dma=1 PATA and SATA Disk DMA only 2472 libata.dma=2 ATAPI (CDROM) DMA only 2473 libata.dma=4 Compact Flash DMA only 2474 Combinations also work, so libata.dma=3 enables DMA 2475 for disks and CDROMs, but not CFs. 2476 2477 libata.ignore_hpa= [LIBATA] Ignore HPA limit 2478 libata.ignore_hpa=0 keep BIOS limits (default) 2479 libata.ignore_hpa=1 ignore limits, using full disk 2480 2481 libata.noacpi [LIBATA] Disables use of ACPI in libata suspend/resume 2482 when set. 2483 Format: <int> 2484 2485 libata.force= [LIBATA] Force configurations. The format is comma 2486 separated list of "[ID:]VAL" where ID is 2487 PORT[.DEVICE]. PORT and DEVICE are decimal numbers 2488 matching port, link or device. Basically, it matches 2489 the ATA ID string printed on console by libata. If 2490 the whole ID part is omitted, the last PORT and DEVICE 2491 values are used. If ID hasn't been specified yet, the 2492 configuration applies to all ports, links and devices. 2493 2494 If only DEVICE is omitted, the parameter applies to 2495 the port and all links and devices behind it. DEVICE 2496 number of 0 either selects the first device or the 2497 first fan-out link behind PMP device. It does not 2498 select the host link. DEVICE number of 15 selects the 2499 host link and device attached to it. 2500 2501 The VAL specifies the configuration to force. As long 2502 as there's no ambiguity shortcut notation is allowed. 2503 For example, both 1.5 and 1.5G would work for 1.5Gbps. 2504 The following configurations can be forced. 2505 2506 * Cable type: 40c, 80c, short40c, unk, ign or sata. 2507 Any ID with matching PORT is used. 2508 2509 * SATA link speed limit: 1.5Gbps or 3.0Gbps. 2510 2511 * Transfer mode: pio[0-7], mwdma[0-4] and udma[0-7]. 2512 udma[/][16,25,33,44,66,100,133] notation is also 2513 allowed. 2514 2515 * [no]ncq: Turn on or off NCQ. 2516 2517 * [no]ncqtrim: Turn off queued DSM TRIM. 2518 2519 * nohrst, nosrst, norst: suppress hard, soft 2520 and both resets. 2521 2522 * rstonce: only attempt one reset during 2523 hot-unplug link recovery 2524 2525 * dump_id: dump IDENTIFY data. 2526 2527 * atapi_dmadir: Enable ATAPI DMADIR bridge support 2528 2529 * disable: Disable this device. 2530 2531 If there are multiple matching configurations changing 2532 the same attribute, the last one is used. 2533 2534 memblock=debug [KNL] Enable memblock debug messages. 2535 2536 android12_only.will_be_removed_soon.memblock_nomap_remove= [KNL] 2537 Setting this to true through kernel command line will 2538 call memblock_remove on the regions marked with no-map 2539 property thereby saving memory by removing page structs 2540 for those regions. By default this is set to false. 2541 2542 load_ramdisk= [RAM] [Deprecated] 2543 2544 lockd.nlm_grace_period=P [NFS] Assign grace period. 2545 Format: <integer> 2546 2547 lockd.nlm_tcpport=N [NFS] Assign TCP port. 2548 Format: <integer> 2549 2550 lockd.nlm_timeout=T [NFS] Assign timeout value. 2551 Format: <integer> 2552 2553 lockd.nlm_udpport=M [NFS] Assign UDP port. 2554 Format: <integer> 2555 2556 lockdown= [SECURITY] 2557 { integrity | confidentiality } 2558 Enable the kernel lockdown feature. If set to 2559 integrity, kernel features that allow userland to 2560 modify the running kernel are disabled. If set to 2561 confidentiality, kernel features that allow userland 2562 to extract confidential information from the kernel 2563 are also disabled. 2564 2565 locktorture.nreaders_stress= [KNL] 2566 Set the number of locking read-acquisition kthreads. 2567 Defaults to being automatically set based on the 2568 number of online CPUs. 2569 2570 locktorture.nwriters_stress= [KNL] 2571 Set the number of locking write-acquisition kthreads. 2572 2573 locktorture.onoff_holdoff= [KNL] 2574 Set time (s) after boot for CPU-hotplug testing. 2575 2576 locktorture.onoff_interval= [KNL] 2577 Set time (s) between CPU-hotplug operations, or 2578 zero to disable CPU-hotplug testing. 2579 2580 locktorture.shuffle_interval= [KNL] 2581 Set task-shuffle interval (jiffies). Shuffling 2582 tasks allows some CPUs to go into dyntick-idle 2583 mode during the locktorture test. 2584 2585 locktorture.shutdown_secs= [KNL] 2586 Set time (s) after boot system shutdown. This 2587 is useful for hands-off automated testing. 2588 2589 locktorture.stat_interval= [KNL] 2590 Time (s) between statistics printk()s. 2591 2592 locktorture.stutter= [KNL] 2593 Time (s) to stutter testing, for example, 2594 specifying five seconds causes the test to run for 2595 five seconds, wait for five seconds, and so on. 2596 This tests the locking primitive's ability to 2597 transition abruptly to and from idle. 2598 2599 locktorture.torture_type= [KNL] 2600 Specify the locking implementation to test. 2601 2602 locktorture.verbose= [KNL] 2603 Enable additional printk() statements. 2604 2605 logibm.irq= [HW,MOUSE] Logitech Bus Mouse Driver 2606 Format: <irq> 2607 2608 loglevel= All Kernel Messages with a loglevel smaller than the 2609 console loglevel will be printed to the console. It can 2610 also be changed with klogd or other programs. The 2611 loglevels are defined as follows: 2612 2613 0 (KERN_EMERG) system is unusable 2614 1 (KERN_ALERT) action must be taken immediately 2615 2 (KERN_CRIT) critical conditions 2616 3 (KERN_ERR) error conditions 2617 4 (KERN_WARNING) warning conditions 2618 5 (KERN_NOTICE) normal but significant condition 2619 6 (KERN_INFO) informational 2620 7 (KERN_DEBUG) debug-level messages 2621 2622 log_buf_len=n[KMG] Sets the size of the printk ring buffer, 2623 in bytes. n must be a power of two and greater 2624 than the minimal size. The minimal size is defined 2625 by LOG_BUF_SHIFT kernel config parameter. There is 2626 also CONFIG_LOG_CPU_MAX_BUF_SHIFT config parameter 2627 that allows to increase the default size depending on 2628 the number of CPUs. See init/Kconfig for more details. 2629 2630 logo.nologo [FB] Disables display of the built-in Linux logo. 2631 This may be used to provide more screen space for 2632 kernel log messages and is useful when debugging 2633 kernel boot problems. 2634 2635 lp=0 [LP] Specify parallel ports to use, e.g, 2636 lp=port[,port...] lp=none,parport0 (lp0 not configured, lp1 uses 2637 lp=reset first parallel port). 'lp=0' disables the 2638 lp=auto printer driver. 'lp=reset' (which can be 2639 specified in addition to the ports) causes 2640 attached printers to be reset. Using 2641 lp=port1,port2,... specifies the parallel ports 2642 to associate lp devices with, starting with 2643 lp0. A port specification may be 'none' to skip 2644 that lp device, or a parport name such as 2645 'parport0'. Specifying 'lp=auto' instead of a 2646 port specification list means that device IDs 2647 from each port should be examined, to see if 2648 an IEEE 1284-compliant printer is attached; if 2649 so, the driver will manage that printer. 2650 See also header of drivers/char/lp.c. 2651 2652 lpj=n [KNL] 2653 Sets loops_per_jiffy to given constant, thus avoiding 2654 time-consuming boot-time autodetection (up to 250 ms per 2655 CPU). 0 enables autodetection (default). To determine 2656 the correct value for your kernel, boot with normal 2657 autodetection and see what value is printed. Note that 2658 on SMP systems the preset will be applied to all CPUs, 2659 which is likely to cause problems if your CPUs need 2660 significantly divergent settings. An incorrect value 2661 will cause delays in the kernel to be wrong, leading to 2662 unpredictable I/O errors and other breakage. Although 2663 unlikely, in the extreme case this might damage your 2664 hardware. 2665 2666 ltpc= [NET] 2667 Format: <io>,<irq>,<dma> 2668 2669 lsm.debug [SECURITY] Enable LSM initialization debugging output. 2670 2671 lsm=lsm1,...,lsmN 2672 [SECURITY] Choose order of LSM initialization. This 2673 overrides CONFIG_LSM, and the "security=" parameter. 2674 2675 machvec= [IA-64] Force the use of a particular machine-vector 2676 (machvec) in a generic kernel. 2677 Example: machvec=hpzx1 2678 2679 machtype= [Loongson] Share the same kernel image file between 2680 different yeeloong laptops. 2681 Example: machtype=lemote-yeeloong-2f-7inch 2682 2683 max_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory greater 2684 than or equal to this physical address is ignored. 2685 2686 maxcpus= [SMP] Maximum number of processors that an SMP kernel 2687 will bring up during bootup. maxcpus=n : n >= 0 limits 2688 the kernel to bring up 'n' processors. Surely after 2689 bootup you can bring up the other plugged cpu by executing 2690 "echo 1 > /sys/devices/system/cpu/cpuX/online". So maxcpus 2691 only takes effect during system bootup. 2692 While n=0 is a special case, it is equivalent to "nosmp", 2693 which also disables the IO APIC. 2694 2695 max_loop= [LOOP] The number of loop block devices that get 2696 (loop.max_loop) unconditionally pre-created at init time. The default 2697 number is configured by BLK_DEV_LOOP_MIN_COUNT. Instead 2698 of statically allocating a predefined number, loop 2699 devices can be requested on-demand with the 2700 /dev/loop-control interface. 2701 2702 mce [X86-32] Machine Check Exception 2703 2704 mce=option [X86-64] See Documentation/x86/x86_64/boot-options.rst 2705 2706 md= [HW] RAID subsystems devices and level 2707 See Documentation/admin-guide/md.rst. 2708 2709 mdacon= [MDA] 2710 Format: <first>,<last> 2711 Specifies range of consoles to be captured by the MDA. 2712 2713 mds= [X86,INTEL] 2714 Control mitigation for the Micro-architectural Data 2715 Sampling (MDS) vulnerability. 2716 2717 Certain CPUs are vulnerable to an exploit against CPU 2718 internal buffers which can forward information to a 2719 disclosure gadget under certain conditions. 2720 2721 In vulnerable processors, the speculatively 2722 forwarded data can be used in a cache side channel 2723 attack, to access data to which the attacker does 2724 not have direct access. 2725 2726 This parameter controls the MDS mitigation. The 2727 options are: 2728 2729 full - Enable MDS mitigation on vulnerable CPUs 2730 full,nosmt - Enable MDS mitigation and disable 2731 SMT on vulnerable CPUs 2732 off - Unconditionally disable MDS mitigation 2733 2734 On TAA-affected machines, mds=off can be prevented by 2735 an active TAA mitigation as both vulnerabilities are 2736 mitigated with the same mechanism so in order to disable 2737 this mitigation, you need to specify tsx_async_abort=off 2738 too. 2739 2740 Not specifying this option is equivalent to 2741 mds=full. 2742 2743 For details see: Documentation/admin-guide/hw-vuln/mds.rst 2744 2745 mem=nn[KMG] [KNL,BOOT] Force usage of a specific amount of memory 2746 Amount of memory to be used in cases as follows: 2747 2748 1 for test; 2749 2 when the kernel is not able to see the whole system memory; 2750 3 memory that lies after 'mem=' boundary is excluded from 2751 the hypervisor, then assigned to KVM guests. 2752 2753 [X86] Work as limiting max address. Use together 2754 with memmap= to avoid physical address space collisions. 2755 Without memmap= PCI devices could be placed at addresses 2756 belonging to unused RAM. 2757 2758 Note that this only takes effects during boot time since 2759 in above case 3, memory may need be hot added after boot 2760 if system memory of hypervisor is not sufficient. 2761 2762 mem=nopentium [BUGS=X86-32] Disable usage of 4MB pages for kernel 2763 memory. 2764 2765 memchunk=nn[KMG] 2766 [KNL,SH] Allow user to override the default size for 2767 per-device physically contiguous DMA buffers. 2768 2769 memhp_default_state=online/offline 2770 [KNL] Set the initial state for the memory hotplug 2771 onlining policy. If not specified, the default value is 2772 set according to the 2773 CONFIG_MEMORY_HOTPLUG_DEFAULT_ONLINE kernel config 2774 option. 2775 See Documentation/admin-guide/mm/memory-hotplug.rst. 2776 2777 memmap=exactmap [KNL,X86] Enable setting of an exact 2778 E820 memory map, as specified by the user. 2779 Such memmap=exactmap lines can be constructed based on 2780 BIOS output or other requirements. See the memmap=nn@ss 2781 option description. 2782 2783 memmap=nn[KMG]@ss[KMG] 2784 [KNL] Force usage of a specific region of memory. 2785 Region of memory to be used is from ss to ss+nn. 2786 If @ss[KMG] is omitted, it is equivalent to mem=nn[KMG], 2787 which limits max address to nn[KMG]. 2788 Multiple different regions can be specified, 2789 comma delimited. 2790 Example: 2791 memmap=100M@2G,100M#3G,1G!1024G 2792 2793 memmap=nn[KMG]#ss[KMG] 2794 [KNL,ACPI] Mark specific memory as ACPI data. 2795 Region of memory to be marked is from ss to ss+nn. 2796 2797 memmap=nn[KMG]$ss[KMG] 2798 [KNL,ACPI] Mark specific memory as reserved. 2799 Region of memory to be reserved is from ss to ss+nn. 2800 Example: Exclude memory from 0x18690000-0x1869ffff 2801 memmap=64K$0x18690000 2802 or 2803 memmap=0x10000$0x18690000 2804 Some bootloaders may need an escape character before '$', 2805 like Grub2, otherwise '$' and the following number 2806 will be eaten. 2807 2808 memmap=nn[KMG]!ss[KMG] 2809 [KNL,X86] Mark specific memory as protected. 2810 Region of memory to be used, from ss to ss+nn. 2811 The memory region may be marked as e820 type 12 (0xc) 2812 and is NVDIMM or ADR memory. 2813 2814 memmap=<size>%<offset>-<oldtype>+<newtype> 2815 [KNL,ACPI] Convert memory within the specified region 2816 from <oldtype> to <newtype>. If "-<oldtype>" is left 2817 out, the whole region will be marked as <newtype>, 2818 even if previously unavailable. If "+<newtype>" is left 2819 out, matching memory will be removed. Types are 2820 specified as e820 types, e.g., 1 = RAM, 2 = reserved, 2821 3 = ACPI, 12 = PRAM. 2822 2823 memory_corruption_check=0/1 [X86] 2824 Some BIOSes seem to corrupt the first 64k of 2825 memory when doing things like suspend/resume. 2826 Setting this option will scan the memory 2827 looking for corruption. Enabling this will 2828 both detect corruption and prevent the kernel 2829 from using the memory being corrupted. 2830 However, its intended as a diagnostic tool; if 2831 repeatable BIOS-originated corruption always 2832 affects the same memory, you can use memmap= 2833 to prevent the kernel from using that memory. 2834 2835 memory_corruption_check_size=size [X86] 2836 By default it checks for corruption in the low 2837 64k, making this memory unavailable for normal 2838 use. Use this parameter to scan for 2839 corruption in more or less memory. 2840 2841 memory_corruption_check_period=seconds [X86] 2842 By default it checks for corruption every 60 2843 seconds. Use this parameter to check at some 2844 other rate. 0 disables periodic checking. 2845 2846 memtest= [KNL,X86,ARM,PPC] Enable memtest 2847 Format: <integer> 2848 default : 0 <disable> 2849 Specifies the number of memtest passes to be 2850 performed. Each pass selects another test 2851 pattern from a given set of patterns. Memtest 2852 fills the memory with this pattern, validates 2853 memory contents and reserves bad memory 2854 regions that are detected. 2855 2856 mem_encrypt= [X86-64] AMD Secure Memory Encryption (SME) control 2857 Valid arguments: on, off 2858 Default (depends on kernel configuration option): 2859 on (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=y) 2860 off (CONFIG_AMD_MEM_ENCRYPT_ACTIVE_BY_DEFAULT=n) 2861 mem_encrypt=on: Activate SME 2862 mem_encrypt=off: Do not activate SME 2863 2864 Refer to Documentation/virt/kvm/amd-memory-encryption.rst 2865 for details on when memory encryption can be activated. 2866 2867 mem_sleep_default= [SUSPEND] Default system suspend mode: 2868 s2idle - Suspend-To-Idle 2869 shallow - Power-On Suspend or equivalent (if supported) 2870 deep - Suspend-To-RAM or equivalent (if supported) 2871 See Documentation/admin-guide/pm/sleep-states.rst. 2872 2873 meye.*= [HW] Set MotionEye Camera parameters 2874 See Documentation/admin-guide/media/meye.rst. 2875 2876 mfgpt_irq= [IA-32] Specify the IRQ to use for the 2877 Multi-Function General Purpose Timers on AMD Geode 2878 platforms. 2879 2880 mfgptfix [X86-32] Fix MFGPT timers on AMD Geode platforms when 2881 the BIOS has incorrectly applied a workaround. TinyBIOS 2882 version 0.98 is known to be affected, 0.99 fixes the 2883 problem by letting the user disable the workaround. 2884 2885 mga= [HW,DRM] 2886 2887 min_addr=nn[KMG] [KNL,BOOT,ia64] All physical memory below this 2888 physical address is ignored. 2889 2890 mini2440= [ARM,HW,KNL] 2891 Format:[0..2][b][c][t] 2892 Default: "0tb" 2893 MINI2440 configuration specification: 2894 0 - The attached screen is the 3.5" TFT 2895 1 - The attached screen is the 7" TFT 2896 2 - The VGA Shield is attached (1024x768) 2897 Leaving out the screen size parameter will not load 2898 the TFT driver, and the framebuffer will be left 2899 unconfigured. 2900 b - Enable backlight. The TFT backlight pin will be 2901 linked to the kernel VESA blanking code and a GPIO 2902 LED. This parameter is not necessary when using the 2903 VGA shield. 2904 c - Enable the s3c camera interface. 2905 t - Reserved for enabling touchscreen support. The 2906 touchscreen support is not enabled in the mainstream 2907 kernel as of 2.6.30, a preliminary port can be found 2908 in the "bleeding edge" mini2440 support kernel at 2909 https://repo.or.cz/w/linux-2.6/mini2440.git 2910 2911 mitigations= 2912 [X86,PPC,S390,ARM64] Control optional mitigations for 2913 CPU vulnerabilities. This is a set of curated, 2914 arch-independent options, each of which is an 2915 aggregation of existing arch-specific options. 2916 2917 off 2918 Disable all optional CPU mitigations. This 2919 improves system performance, but it may also 2920 expose users to several CPU vulnerabilities. 2921 Equivalent to: nopti [X86,PPC] 2922 kpti=0 [ARM64] 2923 nospectre_v1 [X86,PPC] 2924 nobp=0 [S390] 2925 nospectre_v2 [X86,PPC,S390,ARM64] 2926 spectre_v2_user=off [X86] 2927 spec_store_bypass_disable=off [X86,PPC] 2928 ssbd=force-off [ARM64] 2929 l1tf=off [X86] 2930 mds=off [X86] 2931 tsx_async_abort=off [X86] 2932 kvm.nx_huge_pages=off [X86] 2933 no_entry_flush [PPC] 2934 no_uaccess_flush [PPC] 2935 mmio_stale_data=off [X86] 2936 retbleed=off [X86] 2937 2938 Exceptions: 2939 This does not have any effect on 2940 kvm.nx_huge_pages when 2941 kvm.nx_huge_pages=force. 2942 2943 auto (default) 2944 Mitigate all CPU vulnerabilities, but leave SMT 2945 enabled, even if it's vulnerable. This is for 2946 users who don't want to be surprised by SMT 2947 getting disabled across kernel upgrades, or who 2948 have other ways of avoiding SMT-based attacks. 2949 Equivalent to: (default behavior) 2950 2951 auto,nosmt 2952 Mitigate all CPU vulnerabilities, disabling SMT 2953 if needed. This is for users who always want to 2954 be fully mitigated, even if it means losing SMT. 2955 Equivalent to: l1tf=flush,nosmt [X86] 2956 mds=full,nosmt [X86] 2957 tsx_async_abort=full,nosmt [X86] 2958 mmio_stale_data=full,nosmt [X86] 2959 retbleed=auto,nosmt [X86] 2960 2961 mminit_loglevel= 2962 [KNL] When CONFIG_DEBUG_MEMORY_INIT is set, this 2963 parameter allows control of the logging verbosity for 2964 the additional memory initialisation checks. A value 2965 of 0 disables mminit logging and a level of 4 will 2966 log everything. Information is printed at KERN_DEBUG 2967 so loglevel=8 may also need to be specified. 2968 2969 mmio_stale_data= 2970 [X86,INTEL] Control mitigation for the Processor 2971 MMIO Stale Data vulnerabilities. 2972 2973 Processor MMIO Stale Data is a class of 2974 vulnerabilities that may expose data after an MMIO 2975 operation. Exposed data could originate or end in 2976 the same CPU buffers as affected by MDS and TAA. 2977 Therefore, similar to MDS and TAA, the mitigation 2978 is to clear the affected CPU buffers. 2979 2980 This parameter controls the mitigation. The 2981 options are: 2982 2983 full - Enable mitigation on vulnerable CPUs 2984 2985 full,nosmt - Enable mitigation and disable SMT on 2986 vulnerable CPUs. 2987 2988 off - Unconditionally disable mitigation 2989 2990 On MDS or TAA affected machines, 2991 mmio_stale_data=off can be prevented by an active 2992 MDS or TAA mitigation as these vulnerabilities are 2993 mitigated with the same mechanism so in order to 2994 disable this mitigation, you need to specify 2995 mds=off and tsx_async_abort=off too. 2996 2997 Not specifying this option is equivalent to 2998 mmio_stale_data=full. 2999 3000 For details see: 3001 Documentation/admin-guide/hw-vuln/processor_mmio_stale_data.rst 3002 3003 module.sig_enforce 3004 [KNL] When CONFIG_MODULE_SIG is set, this means that 3005 modules without (valid) signatures will fail to load. 3006 Note that if CONFIG_MODULE_SIG_FORCE is set, that 3007 is always true, so this option does nothing. 3008 3009 module_blacklist= [KNL] Do not load a comma-separated list of 3010 modules. Useful for debugging problem modules. 3011 3012 mousedev.tap_time= 3013 [MOUSE] Maximum time between finger touching and 3014 leaving touchpad surface for touch to be considered 3015 a tap and be reported as a left button click (for 3016 touchpads working in absolute mode only). 3017 Format: <msecs> 3018 mousedev.xres= [MOUSE] Horizontal screen resolution, used for devices 3019 reporting absolute coordinates, such as tablets 3020 mousedev.yres= [MOUSE] Vertical screen resolution, used for devices 3021 reporting absolute coordinates, such as tablets 3022 3023 movablecore= [KNL,X86,IA-64,PPC] 3024 Format: nn[KMGTPE] | nn% 3025 This parameter is the complement to kernelcore=, it 3026 specifies the amount of memory used for migratable 3027 allocations. If both kernelcore and movablecore is 3028 specified, then kernelcore will be at *least* the 3029 specified value but may be more. If movablecore on its 3030 own is specified, the administrator must be careful 3031 that the amount of memory usable for all allocations 3032 is not too small. 3033 3034 movable_node [KNL] Boot-time switch to make hotplugable memory 3035 NUMA nodes to be movable. This means that the memory 3036 of such nodes will be usable only for movable 3037 allocations which rules out almost all kernel 3038 allocations. Use with caution! 3039 3040 MTD_Partition= [MTD] 3041 Format: <name>,<region-number>,<size>,<offset> 3042 3043 MTD_Region= [MTD] Format: 3044 <name>,<region-number>[,<base>,<size>,<buswidth>,<altbuswidth>] 3045 3046 mtdparts= [MTD] 3047 See drivers/mtd/parsers/cmdlinepart.c 3048 3049 multitce=off [PPC] This parameter disables the use of the pSeries 3050 firmware feature for updating multiple TCE entries 3051 at a time. 3052 3053 kswapd_per_node= 3054 kswapd_per_node allows you to control the number of kswapd threads 3055 running on the system. This provides the ability to devote additional 3056 CPU resources toward proactive page replacement with the goal of 3057 reducing direct reclaims. When direct reclaims are prevented, the CPU 3058 consumed by them is prevented as well. Depending on the workload, the 3059 result can cause aggregate CPU usage on the system to go up, down or 3060 stay the same. 3061 3062 More aggressive page replacement can reduce direct reclaims which 3063 cause latency for tasks and decrease throughput when doing filesystem 3064 IO through the pagecache. Direct reclaims are recorded using the 3065 allocstall counter in /proc/vmstat. 3066 3067 The range of acceptible values are 1-16. Always start with lower 3068 values in the 2-6 range. Higher values should be justified with 3069 testing. If direct reclaims occur in spite of high values, the cost 3070 of direct reclaims (in latency) that occur can be higher due to 3071 increased lock contention. 3072 3073 onenand.bdry= [HW,MTD] Flex-OneNAND Boundary Configuration 3074 3075 Format: [die0_boundary][,die0_lock][,die1_boundary][,die1_lock] 3076 3077 boundary - index of last SLC block on Flex-OneNAND. 3078 The remaining blocks are configured as MLC blocks. 3079 lock - Configure if Flex-OneNAND boundary should be locked. 3080 Once locked, the boundary cannot be changed. 3081 1 indicates lock status, 0 indicates unlock status. 3082 3083 mtdset= [ARM] 3084 ARM/S3C2412 JIVE boot control 3085 3086 See arch/arm/mach-s3c2412/mach-jive.c 3087 3088 mtouchusb.raw_coordinates= 3089 [HW] Make the MicroTouch USB driver use raw coordinates 3090 ('y', default) or cooked coordinates ('n') 3091 3092 mtrr_chunk_size=nn[KMG] [X86] 3093 used for mtrr cleanup. It is largest continuous chunk 3094 that could hold holes aka. UC entries. 3095 3096 mtrr_gran_size=nn[KMG] [X86] 3097 Used for mtrr cleanup. It is granularity of mtrr block. 3098 Default is 1. 3099 Large value could prevent small alignment from 3100 using up MTRRs. 3101 3102 mtrr_spare_reg_nr=n [X86] 3103 Format: <integer> 3104 Range: 0,7 : spare reg number 3105 Default : 1 3106 Used for mtrr cleanup. It is spare mtrr entries number. 3107 Set to 2 or more if your graphical card needs more. 3108 3109 n2= [NET] SDL Inc. RISCom/N2 synchronous serial card 3110 3111 netdev= [NET] Network devices parameters 3112 Format: <irq>,<io>,<mem_start>,<mem_end>,<name> 3113 Note that mem_start is often overloaded to mean 3114 something different and driver-specific. 3115 This usage is only documented in each driver source 3116 file if at all. 3117 3118 nf_conntrack.acct= 3119 [NETFILTER] Enable connection tracking flow accounting 3120 0 to disable accounting 3121 1 to enable accounting 3122 Default value is 0. 3123 3124 nfsaddrs= [NFS] Deprecated. Use ip= instead. 3125 See Documentation/admin-guide/nfs/nfsroot.rst. 3126 3127 nfsroot= [NFS] nfs root filesystem for disk-less boxes. 3128 See Documentation/admin-guide/nfs/nfsroot.rst. 3129 3130 nfsrootdebug [NFS] enable nfsroot debugging messages. 3131 See Documentation/admin-guide/nfs/nfsroot.rst. 3132 3133 nfs.callback_nr_threads= 3134 [NFSv4] set the total number of threads that the 3135 NFS client will assign to service NFSv4 callback 3136 requests. 3137 3138 nfs.callback_tcpport= 3139 [NFS] set the TCP port on which the NFSv4 callback 3140 channel should listen. 3141 3142 nfs.cache_getent= 3143 [NFS] sets the pathname to the program which is used 3144 to update the NFS client cache entries. 3145 3146 nfs.cache_getent_timeout= 3147 [NFS] sets the timeout after which an attempt to 3148 update a cache entry is deemed to have failed. 3149 3150 nfs.idmap_cache_timeout= 3151 [NFS] set the maximum lifetime for idmapper cache 3152 entries. 3153 3154 nfs.enable_ino64= 3155 [NFS] enable 64-bit inode numbers. 3156 If zero, the NFS client will fake up a 32-bit inode 3157 number for the readdir() and stat() syscalls instead 3158 of returning the full 64-bit number. 3159 The default is to return 64-bit inode numbers. 3160 3161 nfs.max_session_cb_slots= 3162 [NFSv4.1] Sets the maximum number of session 3163 slots the client will assign to the callback 3164 channel. This determines the maximum number of 3165 callbacks the client will process in parallel for 3166 a particular server. 3167 3168 nfs.max_session_slots= 3169 [NFSv4.1] Sets the maximum number of session slots 3170 the client will attempt to negotiate with the server. 3171 This limits the number of simultaneous RPC requests 3172 that the client can send to the NFSv4.1 server. 3173 Note that there is little point in setting this 3174 value higher than the max_tcp_slot_table_limit. 3175 3176 nfs.nfs4_disable_idmapping= 3177 [NFSv4] When set to the default of '1', this option 3178 ensures that both the RPC level authentication 3179 scheme and the NFS level operations agree to use 3180 numeric uids/gids if the mount is using the 3181 'sec=sys' security flavour. In effect it is 3182 disabling idmapping, which can make migration from 3183 legacy NFSv2/v3 systems to NFSv4 easier. 3184 Servers that do not support this mode of operation 3185 will be autodetected by the client, and it will fall 3186 back to using the idmapper. 3187 To turn off this behaviour, set the value to '0'. 3188 nfs.nfs4_unique_id= 3189 [NFS4] Specify an additional fixed unique ident- 3190 ification string that NFSv4 clients can insert into 3191 their nfs_client_id4 string. This is typically a 3192 UUID that is generated at system install time. 3193 3194 nfs.send_implementation_id = 3195 [NFSv4.1] Send client implementation identification 3196 information in exchange_id requests. 3197 If zero, no implementation identification information 3198 will be sent. 3199 The default is to send the implementation identification 3200 information. 3201 3202 nfs.recover_lost_locks = 3203 [NFSv4] Attempt to recover locks that were lost due 3204 to a lease timeout on the server. Please note that 3205 doing this risks data corruption, since there are 3206 no guarantees that the file will remain unchanged 3207 after the locks are lost. 3208 If you want to enable the kernel legacy behaviour of 3209 attempting to recover these locks, then set this 3210 parameter to '1'. 3211 The default parameter value of '0' causes the kernel 3212 not to attempt recovery of lost locks. 3213 3214 nfs4.layoutstats_timer = 3215 [NFSv4.2] Change the rate at which the kernel sends 3216 layoutstats to the pNFS metadata server. 3217 3218 Setting this to value to 0 causes the kernel to use 3219 whatever value is the default set by the layout 3220 driver. A non-zero value sets the minimum interval 3221 in seconds between layoutstats transmissions. 3222 3223 nfsd.nfs4_disable_idmapping= 3224 [NFSv4] When set to the default of '1', the NFSv4 3225 server will return only numeric uids and gids to 3226 clients using auth_sys, and will accept numeric uids 3227 and gids from such clients. This is intended to ease 3228 migration from NFSv2/v3. 3229 3230 nmi_backtrace.backtrace_idle [KNL] 3231 Dump stacks even of idle CPUs in response to an 3232 NMI stack-backtrace request. 3233 3234 nmi_debug= [KNL,SH] Specify one or more actions to take 3235 when a NMI is triggered. 3236 Format: [state][,regs][,debounce][,die] 3237 3238 nmi_watchdog= [KNL,BUGS=X86] Debugging features for SMP kernels 3239 Format: [panic,][nopanic,][num] 3240 Valid num: 0 or 1 3241 0 - turn hardlockup detector in nmi_watchdog off 3242 1 - turn hardlockup detector in nmi_watchdog on 3243 When panic is specified, panic when an NMI watchdog 3244 timeout occurs (or 'nopanic' to not panic on an NMI 3245 watchdog, if CONFIG_BOOTPARAM_HARDLOCKUP_PANIC is set) 3246 To disable both hard and soft lockup detectors, 3247 please see 'nowatchdog'. 3248 This is useful when you use a panic=... timeout and 3249 need the box quickly up again. 3250 3251 These settings can be accessed at runtime via 3252 the nmi_watchdog and hardlockup_panic sysctls. 3253 3254 netpoll.carrier_timeout= 3255 [NET] Specifies amount of time (in seconds) that 3256 netpoll should wait for a carrier. By default netpoll 3257 waits 4 seconds. 3258 3259 no387 [BUGS=X86-32] Tells the kernel to use the 387 maths 3260 emulation library even if a 387 maths coprocessor 3261 is present. 3262 3263 no5lvl [X86-64] Disable 5-level paging mode. Forces 3264 kernel to use 4-level paging instead. 3265 3266 nofsgsbase [X86] Disables FSGSBASE instructions. 3267 3268 no_console_suspend 3269 [HW] Never suspend the console 3270 Disable suspending of consoles during suspend and 3271 hibernate operations. Once disabled, debugging 3272 messages can reach various consoles while the rest 3273 of the system is being put to sleep (ie, while 3274 debugging driver suspend/resume hooks). This may 3275 not work reliably with all consoles, but is known 3276 to work with serial and VGA consoles. 3277 To facilitate more flexible debugging, we also add 3278 console_suspend, a printk module parameter to control 3279 it. Users could use console_suspend (usually 3280 /sys/module/printk/parameters/console_suspend) to 3281 turn on/off it dynamically. 3282 3283 novmcoredd [KNL,KDUMP] 3284 Disable device dump. Device dump allows drivers to 3285 append dump data to vmcore so you can collect driver 3286 specified debug info. Drivers can append the data 3287 without any limit and this data is stored in memory, 3288 so this may cause significant memory stress. Disabling 3289 device dump can help save memory but the driver debug 3290 data will be no longer available. This parameter 3291 is only available when CONFIG_PROC_VMCORE_DEVICE_DUMP 3292 is set. 3293 3294 noaliencache [MM, NUMA, SLAB] Disables the allocation of alien 3295 caches in the slab allocator. Saves per-node memory, 3296 but will impact performance. 3297 3298 noalign [KNL,ARM] 3299 3300 noaltinstr [S390] Disables alternative instructions patching 3301 (CPU alternatives feature). 3302 3303 noapic [SMP,APIC] Tells the kernel to not make use of any 3304 IOAPICs that may be present in the system. 3305 3306 noautogroup Disable scheduler automatic task group creation. 3307 3308 nobats [PPC] Do not use BATs for mapping kernel lowmem 3309 on "Classic" PPC cores. 3310 3311 nocache [ARM] 3312 3313 noclflush [BUGS=X86] Don't use the CLFLUSH instruction 3314 3315 nodelayacct [KNL] Disable per-task delay accounting 3316 3317 nodsp [SH] Disable hardware DSP at boot time. 3318 3319 noefi Disable EFI runtime services support. 3320 3321 no_entry_flush [PPC] Don't flush the L1-D cache when entering the kernel. 3322 3323 noexec [IA-64] 3324 3325 noexec [X86] 3326 On X86-32 available only on PAE configured kernels. 3327 noexec=on: enable non-executable mappings (default) 3328 noexec=off: disable non-executable mappings 3329 3330 nosmap [X86,PPC] 3331 Disable SMAP (Supervisor Mode Access Prevention) 3332 even if it is supported by processor. 3333 3334 nosmep [X86,PPC] 3335 Disable SMEP (Supervisor Mode Execution Prevention) 3336 even if it is supported by processor. 3337 3338 noexec32 [X86-64] 3339 This affects only 32-bit executables. 3340 noexec32=on: enable non-executable mappings (default) 3341 read doesn't imply executable mappings 3342 noexec32=off: disable non-executable mappings 3343 read implies executable mappings 3344 3345 nofpu [MIPS,SH] Disable hardware FPU at boot time. 3346 3347 nofxsr [BUGS=X86-32] Disables x86 floating point extended 3348 register save and restore. The kernel will only save 3349 legacy floating-point registers on task switch. 3350 3351 nohugeiomap [KNL,X86,PPC,ARM64] Disable kernel huge I/O mappings. 3352 3353 nosmt [KNL,S390] Disable symmetric multithreading (SMT). 3354 Equivalent to smt=1. 3355 3356 [KNL,X86] Disable symmetric multithreading (SMT). 3357 nosmt=force: Force disable SMT, cannot be undone 3358 via the sysfs control file. 3359 3360 nospectre_v1 [X86,PPC] Disable mitigations for Spectre Variant 1 3361 (bounds check bypass). With this option data leaks are 3362 possible in the system. 3363 3364 nospectre_v2 [X86,PPC_FSL_BOOK3E,ARM64] Disable all mitigations for 3365 the Spectre variant 2 (indirect branch prediction) 3366 vulnerability. System may allow data leaks with this 3367 option. 3368 3369 nospec_store_bypass_disable 3370 [HW] Disable all mitigations for the Speculative Store Bypass vulnerability 3371 3372 no_uaccess_flush 3373 [PPC] Don't flush the L1-D cache after accessing user data. 3374 3375 noxsave [BUGS=X86] Disables x86 extended register state save 3376 and restore using xsave. The kernel will fallback to 3377 enabling legacy floating-point and sse state. 3378 3379 noxsaveopt [X86] Disables xsaveopt used in saving x86 extended 3380 register states. The kernel will fall back to use 3381 xsave to save the states. By using this parameter, 3382 performance of saving the states is degraded because 3383 xsave doesn't support modified optimization while 3384 xsaveopt supports it on xsaveopt enabled systems. 3385 3386 noxsaves [X86] Disables xsaves and xrstors used in saving and 3387 restoring x86 extended register state in compacted 3388 form of xsave area. The kernel will fall back to use 3389 xsaveopt and xrstor to save and restore the states 3390 in standard form of xsave area. By using this 3391 parameter, xsave area per process might occupy more 3392 memory on xsaves enabled systems. 3393 3394 nohlt [BUGS=ARM,SH] Tells the kernel that the sleep(SH) or 3395 wfi(ARM) instruction doesn't work correctly and not to 3396 use it. This is also useful when using JTAG debugger. 3397 3398 no_file_caps Tells the kernel not to honor file capabilities. The 3399 only way then for a file to be executed with privilege 3400 is to be setuid root or executed by root. 3401 3402 nohalt [IA-64] Tells the kernel not to use the power saving 3403 function PAL_HALT_LIGHT when idle. This increases 3404 power-consumption. On the positive side, it reduces 3405 interrupt wake-up latency, which may improve performance 3406 in certain environments such as networked servers or 3407 real-time systems. 3408 3409 no_hash_pointers 3410 Force pointers printed to the console or buffers to be 3411 unhashed. By default, when a pointer is printed via %p 3412 format string, that pointer is "hashed", i.e. obscured 3413 by hashing the pointer value. This is a security feature 3414 that hides actual kernel addresses from unprivileged 3415 users, but it also makes debugging the kernel more 3416 difficult since unequal pointers can no longer be 3417 compared. However, if this command-line option is 3418 specified, then all normal pointers will have their true 3419 value printed. This option should only be specified when 3420 debugging the kernel. Please do not use on production 3421 kernels. 3422 3423 nohibernate [HIBERNATION] Disable hibernation and resume. 3424 3425 nohz= [KNL] Boottime enable/disable dynamic ticks 3426 Valid arguments: on, off 3427 Default: on 3428 3429 nohz_full= [KNL,BOOT,SMP,ISOL] 3430 The argument is a cpu list, as described above. 3431 In kernels built with CONFIG_NO_HZ_FULL=y, set 3432 the specified list of CPUs whose tick will be stopped 3433 whenever possible. The boot CPU will be forced outside 3434 the range to maintain the timekeeping. Any CPUs 3435 in this list will have their RCU callbacks offloaded, 3436 just as if they had also been called out in the 3437 rcu_nocbs= boot parameter. 3438 3439 noiotrap [SH] Disables trapped I/O port accesses. 3440 3441 noirqdebug [X86-32] Disables the code which attempts to detect and 3442 disable unhandled interrupt sources. 3443 3444 no_timer_check [X86,APIC] Disables the code which tests for 3445 broken timer IRQ sources. 3446 3447 noisapnp [ISAPNP] Disables ISA PnP code. 3448 3449 noinitrd [RAM] Tells the kernel not to load any configured 3450 initial RAM disk. 3451 3452 nointremap [X86-64, Intel-IOMMU] Do not enable interrupt 3453 remapping. 3454 [Deprecated - use intremap=off] 3455 3456 nointroute [IA-64] 3457 3458 noinvpcid [X86] Disable the INVPCID cpu feature. 3459 3460 nojitter [IA-64] Disables jitter checking for ITC timers. 3461 3462 no-kvmclock [X86,KVM] Disable paravirtualized KVM clock driver 3463 3464 no-kvmapf [X86,KVM] Disable paravirtualized asynchronous page 3465 fault handling. 3466 3467 no-vmw-sched-clock 3468 [X86,PV_OPS] Disable paravirtualized VMware scheduler 3469 clock and use the default one. 3470 3471 no-steal-acc [X86,PV_OPS,ARM64] Disable paravirtualized steal time 3472 accounting. steal time is computed, but won't 3473 influence scheduler behaviour 3474 3475 nolapic [X86-32,APIC] Do not enable or use the local APIC. 3476 3477 nolapic_timer [X86-32,APIC] Do not use the local APIC timer. 3478 3479 noltlbs [PPC] Do not use large page/tlb entries for kernel 3480 lowmem mapping on PPC40x and PPC8xx 3481 3482 nomca [IA-64] Disable machine check abort handling 3483 3484 nomce [X86-32] Disable Machine Check Exception 3485 3486 nomfgpt [X86-32] Disable Multi-Function General Purpose 3487 Timer usage (for AMD Geode machines). 3488 3489 nonmi_ipi [X86] Disable using NMI IPIs during panic/reboot to 3490 shutdown the other cpus. Instead use the REBOOT_VECTOR 3491 irq. 3492 3493 nomodule Disable module load 3494 3495 nopat [X86] Disable PAT (page attribute table extension of 3496 pagetables) support. 3497 3498 nopcid [X86-64] Disable the PCID cpu feature. 3499 3500 norandmaps Don't use address space randomization. Equivalent to 3501 echo 0 > /proc/sys/kernel/randomize_va_space 3502 3503 noreplace-smp [X86-32,SMP] Don't replace SMP instructions 3504 with UP alternatives 3505 3506 nordrand [X86] Disable kernel use of the RDRAND and 3507 RDSEED instructions even if they are supported 3508 by the processor. RDRAND and RDSEED are still 3509 available to user space applications. 3510 3511 noresume [SWSUSP] Disables resume and restores original swap 3512 space. 3513 3514 no-scroll [VGA] Disables scrollback. 3515 This is required for the Braillex ib80-piezo Braille 3516 reader made by F.H. Papenmeier (Germany). 3517 3518 nosbagart [IA-64] 3519 3520 nosep [BUGS=X86-32] Disables x86 SYSENTER/SYSEXIT support. 3521 3522 nosmp [SMP] Tells an SMP kernel to act as a UP kernel, 3523 and disable the IO APIC. legacy for "maxcpus=0". 3524 3525 nosoftlockup [KNL] Disable the soft-lockup detector. 3526 3527 nosync [HW,M68K] Disables sync negotiation for all devices. 3528 3529 nowatchdog [KNL] Disable both lockup detectors, i.e. 3530 soft-lockup and NMI watchdog (hard-lockup). 3531 3532 nowb [ARM] 3533 3534 nox2apic [X86-64,APIC] Do not enable x2APIC mode. 3535 3536 cpu0_hotplug [X86] Turn on CPU0 hotplug feature when 3537 CONFIG_BOOTPARAM_HOTPLUG_CPU0 is off. 3538 Some features depend on CPU0. Known dependencies are: 3539 1. Resume from suspend/hibernate depends on CPU0. 3540 Suspend/hibernate will fail if CPU0 is offline and you 3541 need to online CPU0 before suspend/hibernate. 3542 2. PIC interrupts also depend on CPU0. CPU0 can't be 3543 removed if a PIC interrupt is detected. 3544 It's said poweroff/reboot may depend on CPU0 on some 3545 machines although I haven't seen such issues so far 3546 after CPU0 is offline on a few tested machines. 3547 If the dependencies are under your control, you can 3548 turn on cpu0_hotplug. 3549 3550 nps_mtm_hs_ctr= [KNL,ARC] 3551 This parameter sets the maximum duration, in 3552 cycles, each HW thread of the CTOP can run 3553 without interruptions, before HW switches it. 3554 The actual maximum duration is 16 times this 3555 parameter's value. 3556 Format: integer between 1 and 255 3557 Default: 255 3558 3559 nptcg= [IA-64] Override max number of concurrent global TLB 3560 purges which is reported from either PAL_VM_SUMMARY or 3561 SAL PALO. 3562 3563 nr_cpus= [SMP] Maximum number of processors that an SMP kernel 3564 could support. nr_cpus=n : n >= 1 limits the kernel to 3565 support 'n' processors. It could be larger than the 3566 number of already plugged CPU during bootup, later in 3567 runtime you can physically add extra cpu until it reaches 3568 n. So during boot up some boot time memory for per-cpu 3569 variables need be pre-allocated for later physical cpu 3570 hot plugging. 3571 3572 nr_uarts= [SERIAL] maximum number of UARTs to be registered. 3573 3574 numa_balancing= [KNL,X86] Enable or disable automatic NUMA balancing. 3575 Allowed values are enable and disable 3576 3577 numa_zonelist_order= [KNL, BOOT] Select zonelist order for NUMA. 3578 'node', 'default' can be specified 3579 This can be set from sysctl after boot. 3580 See Documentation/admin-guide/sysctl/vm.rst for details. 3581 3582 ohci1394_dma=early [HW] enable debugging via the ohci1394 driver. 3583 See Documentation/core-api/debugging-via-ohci1394.rst for more 3584 info. 3585 3586 olpc_ec_timeout= [OLPC] ms delay when issuing EC commands 3587 Rather than timing out after 20 ms if an EC 3588 command is not properly ACKed, override the length 3589 of the timeout. We have interrupts disabled while 3590 waiting for the ACK, so if this is set too high 3591 interrupts *may* be lost! 3592 3593 omap_mux= [OMAP] Override bootloader pin multiplexing. 3594 Format: <mux_mode0.mode_name=value>... 3595 For example, to override I2C bus2: 3596 omap_mux=i2c2_scl.i2c2_scl=0x100,i2c2_sda.i2c2_sda=0x100 3597 3598 oprofile.timer= [HW] 3599 Use timer interrupt instead of performance counters 3600 3601 oprofile.cpu_type= Force an oprofile cpu type 3602 This might be useful if you have an older oprofile 3603 userland or if you want common events. 3604 Format: { arch_perfmon } 3605 arch_perfmon: [X86] Force use of architectural 3606 perfmon on Intel CPUs instead of the 3607 CPU specific event set. 3608 timer: [X86] Force use of architectural NMI 3609 timer mode (see also oprofile.timer 3610 for generic hr timer mode) 3611 3612 oops=panic Always panic on oopses. Default is to just kill the 3613 process, but there is a small probability of 3614 deadlocking the machine. 3615 This will also cause panics on machine check exceptions. 3616 Useful together with panic=30 to trigger a reboot. 3617 3618 page_alloc.shuffle= 3619 [KNL] Boolean flag to control whether the page allocator 3620 should randomize its free lists. The randomization may 3621 be automatically enabled if the kernel detects it is 3622 running on a platform with a direct-mapped memory-side 3623 cache, and this parameter can be used to 3624 override/disable that behavior. The state of the flag 3625 can be read from sysfs at: 3626 /sys/module/page_alloc/parameters/shuffle. 3627 3628 page_owner= [KNL] Boot-time page_owner enabling option. 3629 Storage of the information about who allocated 3630 each page is disabled in default. With this switch, 3631 we can turn it on. 3632 on: enable the feature 3633 3634 page_poison= [KNL] Boot-time parameter changing the state of 3635 poisoning on the buddy allocator, available with 3636 CONFIG_PAGE_POISONING=y. 3637 off: turn off poisoning (default) 3638 on: turn on poisoning 3639 3640 panic= [KNL] Kernel behaviour on panic: delay <timeout> 3641 timeout > 0: seconds before rebooting 3642 timeout = 0: wait forever 3643 timeout < 0: reboot immediately 3644 Format: <timeout> 3645 3646 panic_print= Bitmask for printing system info when panic happens. 3647 User can chose combination of the following bits: 3648 bit 0: print all tasks info 3649 bit 1: print system memory info 3650 bit 2: print timer info 3651 bit 3: print locks info if CONFIG_LOCKDEP is on 3652 bit 4: print ftrace buffer 3653 bit 5: print all printk messages in buffer 3654 3655 panic_on_taint= Bitmask for conditionally calling panic() in add_taint() 3656 Format: <hex>[,nousertaint] 3657 Hexadecimal bitmask representing the set of TAINT flags 3658 that will cause the kernel to panic when add_taint() is 3659 called with any of the flags in this set. 3660 The optional switch "nousertaint" can be utilized to 3661 prevent userspace forced crashes by writing to sysctl 3662 /proc/sys/kernel/tainted any flagset matching with the 3663 bitmask set on panic_on_taint. 3664 See Documentation/admin-guide/tainted-kernels.rst for 3665 extra details on the taint flags that users can pick 3666 to compose the bitmask to assign to panic_on_taint. 3667 3668 panic_on_warn panic() instead of WARN(). Useful to cause kdump 3669 on a WARN(). 3670 3671 pelt= [KNL] Boot-time parameter changing the PELT half life in ms 3672 Format: <int> 3673 32: Set the half life to 32ms 3674 8: Set the half life to 8ms 3675 default: 32 3676 3677 crash_kexec_post_notifiers 3678 Run kdump after running panic-notifiers and dumping 3679 kmsg. This only for the users who doubt kdump always 3680 succeeds in any situation. 3681 Note that this also increases risks of kdump failure, 3682 because some panic notifiers can make the crashed 3683 kernel more unstable. 3684 3685 parkbd.port= [HW] Parallel port number the keyboard adapter is 3686 connected to, default is 0. 3687 Format: <parport#> 3688 parkbd.mode= [HW] Parallel port keyboard adapter mode of operation, 3689 0 for XT, 1 for AT (default is AT). 3690 Format: <mode> 3691 3692 parport= [HW,PPT] Specify parallel ports. 0 disables. 3693 Format: { 0 | auto | 0xBBB[,IRQ[,DMA]] } 3694 Use 'auto' to force the driver to use any 3695 IRQ/DMA settings detected (the default is to 3696 ignore detected IRQ/DMA settings because of 3697 possible conflicts). You can specify the base 3698 address, IRQ, and DMA settings; IRQ and DMA 3699 should be numbers, or 'auto' (for using detected 3700 settings on that particular port), or 'nofifo' 3701 (to avoid using a FIFO even if it is detected). 3702 Parallel ports are assigned in the order they 3703 are specified on the command line, starting 3704 with parport0. 3705 3706 parport_init_mode= [HW,PPT] 3707 Configure VIA parallel port to operate in 3708 a specific mode. This is necessary on Pegasos 3709 computer where firmware has no options for setting 3710 up parallel port mode and sets it to spp. 3711 Currently this function knows 686a and 8231 chips. 3712 Format: [spp|ps2|epp|ecp|ecpepp] 3713 3714 pause_on_oops= 3715 Halt all CPUs after the first oops has been printed for 3716 the specified number of seconds. This is to be used if 3717 your oopses keep scrolling off the screen. 3718 3719 pcbit= [HW,ISDN] 3720 3721 pcd. [PARIDE] 3722 See header of drivers/block/paride/pcd.c. 3723 See also Documentation/admin-guide/blockdev/paride.rst. 3724 3725 pci=option[,option...] [PCI] various PCI subsystem options. 3726 3727 Some options herein operate on a specific device 3728 or a set of devices (<pci_dev>). These are 3729 specified in one of the following formats: 3730 3731 [<domain>:]<bus>:<dev>.<func>[/<dev>.<func>]* 3732 pci:<vendor>:<device>[:<subvendor>:<subdevice>] 3733 3734 Note: the first format specifies a PCI 3735 bus/device/function address which may change 3736 if new hardware is inserted, if motherboard 3737 firmware changes, or due to changes caused 3738 by other kernel parameters. If the 3739 domain is left unspecified, it is 3740 taken to be zero. Optionally, a path 3741 to a device through multiple device/function 3742 addresses can be specified after the base 3743 address (this is more robust against 3744 renumbering issues). The second format 3745 selects devices using IDs from the 3746 configuration space which may match multiple 3747 devices in the system. 3748 3749 earlydump dump PCI config space before the kernel 3750 changes anything 3751 off [X86] don't probe for the PCI bus 3752 bios [X86-32] force use of PCI BIOS, don't access 3753 the hardware directly. Use this if your machine 3754 has a non-standard PCI host bridge. 3755 nobios [X86-32] disallow use of PCI BIOS, only direct 3756 hardware access methods are allowed. Use this 3757 if you experience crashes upon bootup and you 3758 suspect they are caused by the BIOS. 3759 conf1 [X86] Force use of PCI Configuration Access 3760 Mechanism 1 (config address in IO port 0xCF8, 3761 data in IO port 0xCFC, both 32-bit). 3762 conf2 [X86] Force use of PCI Configuration Access 3763 Mechanism 2 (IO port 0xCF8 is an 8-bit port for 3764 the function, IO port 0xCFA, also 8-bit, sets 3765 bus number. The config space is then accessed 3766 through ports 0xC000-0xCFFF). 3767 See http://wiki.osdev.org/PCI for more info 3768 on the configuration access mechanisms. 3769 noaer [PCIE] If the PCIEAER kernel config parameter is 3770 enabled, this kernel boot option can be used to 3771 disable the use of PCIE advanced error reporting. 3772 nodomains [PCI] Disable support for multiple PCI 3773 root domains (aka PCI segments, in ACPI-speak). 3774 nommconf [X86] Disable use of MMCONFIG for PCI 3775 Configuration 3776 check_enable_amd_mmconf [X86] check for and enable 3777 properly configured MMIO access to PCI 3778 config space on AMD family 10h CPU 3779 nomsi [MSI] If the PCI_MSI kernel config parameter is 3780 enabled, this kernel boot option can be used to 3781 disable the use of MSI interrupts system-wide. 3782 noioapicquirk [APIC] Disable all boot interrupt quirks. 3783 Safety option to keep boot IRQs enabled. This 3784 should never be necessary. 3785 ioapicreroute [APIC] Enable rerouting of boot IRQs to the 3786 primary IO-APIC for bridges that cannot disable 3787 boot IRQs. This fixes a source of spurious IRQs 3788 when the system masks IRQs. 3789 noioapicreroute [APIC] Disable workaround that uses the 3790 boot IRQ equivalent of an IRQ that connects to 3791 a chipset where boot IRQs cannot be disabled. 3792 The opposite of ioapicreroute. 3793 biosirq [X86-32] Use PCI BIOS calls to get the interrupt 3794 routing table. These calls are known to be buggy 3795 on several machines and they hang the machine 3796 when used, but on other computers it's the only 3797 way to get the interrupt routing table. Try 3798 this option if the kernel is unable to allocate 3799 IRQs or discover secondary PCI buses on your 3800 motherboard. 3801 rom [X86] Assign address space to expansion ROMs. 3802 Use with caution as certain devices share 3803 address decoders between ROMs and other 3804 resources. 3805 norom [X86] Do not assign address space to 3806 expansion ROMs that do not already have 3807 BIOS assigned address ranges. 3808 nobar [X86] Do not assign address space to the 3809 BARs that weren't assigned by the BIOS. 3810 irqmask=0xMMMM [X86] Set a bit mask of IRQs allowed to be 3811 assigned automatically to PCI devices. You can 3812 make the kernel exclude IRQs of your ISA cards 3813 this way. 3814 pirqaddr=0xAAAAA [X86] Specify the physical address 3815 of the PIRQ table (normally generated 3816 by the BIOS) if it is outside the 3817 F0000h-100000h range. 3818 lastbus=N [X86] Scan all buses thru bus #N. Can be 3819 useful if the kernel is unable to find your 3820 secondary buses and you want to tell it 3821 explicitly which ones they are. 3822 assign-busses [X86] Always assign all PCI bus 3823 numbers ourselves, overriding 3824 whatever the firmware may have done. 3825 usepirqmask [X86] Honor the possible IRQ mask stored 3826 in the BIOS $PIR table. This is needed on 3827 some systems with broken BIOSes, notably 3828 some HP Pavilion N5400 and Omnibook XE3 3829 notebooks. This will have no effect if ACPI 3830 IRQ routing is enabled. 3831 noacpi [X86] Do not use ACPI for IRQ routing 3832 or for PCI scanning. 3833 use_crs [X86] Use PCI host bridge window information 3834 from ACPI. On BIOSes from 2008 or later, this 3835 is enabled by default. If you need to use this, 3836 please report a bug. 3837 nocrs [X86] Ignore PCI host bridge windows from ACPI. 3838 If you need to use this, please report a bug. 3839 routeirq Do IRQ routing for all PCI devices. 3840 This is normally done in pci_enable_device(), 3841 so this option is a temporary workaround 3842 for broken drivers that don't call it. 3843 skip_isa_align [X86] do not align io start addr, so can 3844 handle more pci cards 3845 noearly [X86] Don't do any early type 1 scanning. 3846 This might help on some broken boards which 3847 machine check when some devices' config space 3848 is read. But various workarounds are disabled 3849 and some IOMMU drivers will not work. 3850 bfsort Sort PCI devices into breadth-first order. 3851 This sorting is done to get a device 3852 order compatible with older (<= 2.4) kernels. 3853 nobfsort Don't sort PCI devices into breadth-first order. 3854 pcie_bus_tune_off Disable PCIe MPS (Max Payload Size) 3855 tuning and use the BIOS-configured MPS defaults. 3856 pcie_bus_safe Set every device's MPS to the largest value 3857 supported by all devices below the root complex. 3858 pcie_bus_perf Set device MPS to the largest allowable MPS 3859 based on its parent bus. Also set MRRS (Max 3860 Read Request Size) to the largest supported 3861 value (no larger than the MPS that the device 3862 or bus can support) for best performance. 3863 pcie_bus_peer2peer Set every device's MPS to 128B, which 3864 every device is guaranteed to support. This 3865 configuration allows peer-to-peer DMA between 3866 any pair of devices, possibly at the cost of 3867 reduced performance. This also guarantees 3868 that hot-added devices will work. 3869 cbiosize=nn[KMG] The fixed amount of bus space which is 3870 reserved for the CardBus bridge's IO window. 3871 The default value is 256 bytes. 3872 cbmemsize=nn[KMG] The fixed amount of bus space which is 3873 reserved for the CardBus bridge's memory 3874 window. The default value is 64 megabytes. 3875 resource_alignment= 3876 Format: 3877 [<order of align>@]<pci_dev>[; ...] 3878 Specifies alignment and device to reassign 3879 aligned memory resources. How to 3880 specify the device is described above. 3881 If <order of align> is not specified, 3882 PAGE_SIZE is used as alignment. 3883 A PCI-PCI bridge can be specified if resource 3884 windows need to be expanded. 3885 To specify the alignment for several 3886 instances of a device, the PCI vendor, 3887 device, subvendor, and subdevice may be 3888 specified, e.g., 12@pci:8086:9c22:103c:198f 3889 for 4096-byte alignment. 3890 ecrc= Enable/disable PCIe ECRC (transaction layer 3891 end-to-end CRC checking). 3892 bios: Use BIOS/firmware settings. This is the 3893 the default. 3894 off: Turn ECRC off 3895 on: Turn ECRC on. 3896 hpiosize=nn[KMG] The fixed amount of bus space which is 3897 reserved for hotplug bridge's IO window. 3898 Default size is 256 bytes. 3899 hpmmiosize=nn[KMG] The fixed amount of bus space which is 3900 reserved for hotplug bridge's MMIO window. 3901 Default size is 2 megabytes. 3902 hpmmioprefsize=nn[KMG] The fixed amount of bus space which is 3903 reserved for hotplug bridge's MMIO_PREF window. 3904 Default size is 2 megabytes. 3905 hpmemsize=nn[KMG] The fixed amount of bus space which is 3906 reserved for hotplug bridge's MMIO and 3907 MMIO_PREF window. 3908 Default size is 2 megabytes. 3909 hpbussize=nn The minimum amount of additional bus numbers 3910 reserved for buses below a hotplug bridge. 3911 Default is 1. 3912 realloc= Enable/disable reallocating PCI bridge resources 3913 if allocations done by BIOS are too small to 3914 accommodate resources required by all child 3915 devices. 3916 off: Turn realloc off 3917 on: Turn realloc on 3918 realloc same as realloc=on 3919 noari do not use PCIe ARI. 3920 noats [PCIE, Intel-IOMMU, AMD-IOMMU] 3921 do not use PCIe ATS (and IOMMU device IOTLB). 3922 pcie_scan_all Scan all possible PCIe devices. Otherwise we 3923 only look for one device below a PCIe downstream 3924 port. 3925 big_root_window Try to add a big 64bit memory window to the PCIe 3926 root complex on AMD CPUs. Some GFX hardware 3927 can resize a BAR to allow access to all VRAM. 3928 Adding the window is slightly risky (it may 3929 conflict with unreported devices), so this 3930 taints the kernel. 3931 disable_acs_redir=<pci_dev>[; ...] 3932 Specify one or more PCI devices (in the format 3933 specified above) separated by semicolons. 3934 Each device specified will have the PCI ACS 3935 redirect capabilities forced off which will 3936 allow P2P traffic between devices through 3937 bridges without forcing it upstream. Note: 3938 this removes isolation between devices and 3939 may put more devices in an IOMMU group. 3940 force_floating [S390] Force usage of floating interrupts. 3941 nomio [S390] Do not use MIO instructions. 3942 norid [S390] ignore the RID field and force use of 3943 one PCI domain per PCI function 3944 3945 pcie_aspm= [PCIE] Forcibly enable or disable PCIe Active State Power 3946 Management. 3947 off Disable ASPM. 3948 force Enable ASPM even on devices that claim not to support it. 3949 WARNING: Forcing ASPM on may cause system lockups. 3950 3951 pcie_ports= [PCIE] PCIe port services handling: 3952 native Use native PCIe services (PME, AER, DPC, PCIe hotplug) 3953 even if the platform doesn't give the OS permission to 3954 use them. This may cause conflicts if the platform 3955 also tries to use these services. 3956 dpc-native Use native PCIe service for DPC only. May 3957 cause conflicts if firmware uses AER or DPC. 3958 compat Disable native PCIe services (PME, AER, DPC, PCIe 3959 hotplug). 3960 3961 pcie_port_pm= [PCIE] PCIe port power management handling: 3962 off Disable power management of all PCIe ports 3963 force Forcibly enable power management of all PCIe ports 3964 3965 pcie_pme= [PCIE,PM] Native PCIe PME signaling options: 3966 nomsi Do not use MSI for native PCIe PME signaling (this makes 3967 all PCIe root ports use INTx for all services). 3968 3969 pcmv= [HW,PCMCIA] BadgePAD 4 3970 3971 pd_ignore_unused 3972 [PM] 3973 Keep all power-domains already enabled by bootloader on, 3974 even if no driver has claimed them. This is useful 3975 for debug and development, but should not be 3976 needed on a platform with proper driver support. 3977 3978 pd. [PARIDE] 3979 See Documentation/admin-guide/blockdev/paride.rst. 3980 3981 pdcchassis= [PARISC,HW] Disable/Enable PDC Chassis Status codes at 3982 boot time. 3983 Format: { 0 | 1 } 3984 See arch/parisc/kernel/pdc_chassis.c 3985 3986 percpu_alloc= Select which percpu first chunk allocator to use. 3987 Currently supported values are "embed" and "page". 3988 Archs may support subset or none of the selections. 3989 See comments in mm/percpu.c for details on each 3990 allocator. This parameter is primarily for debugging 3991 and performance comparison. 3992 3993 pf. [PARIDE] 3994 See Documentation/admin-guide/blockdev/paride.rst. 3995 3996 pg. [PARIDE] 3997 See Documentation/admin-guide/blockdev/paride.rst. 3998 3999 pirq= [SMP,APIC] Manual mp-table setup 4000 See Documentation/x86/i386/IO-APIC.rst. 4001 4002 plip= [PPT,NET] Parallel port network link 4003 Format: { parport<nr> | timid | 0 } 4004 See also Documentation/admin-guide/parport.rst. 4005 4006 pmtmr= [X86] Manual setup of pmtmr I/O Port. 4007 Override pmtimer IOPort with a hex value. 4008 e.g. pmtmr=0x508 4009 4010 pm_debug_messages [SUSPEND,KNL] 4011 Enable suspend/resume debug messages during boot up. 4012 4013 pnp.debug=1 [PNP] 4014 Enable PNP debug messages (depends on the 4015 CONFIG_PNP_DEBUG_MESSAGES option). Change at run-time 4016 via /sys/module/pnp/parameters/debug. We always show 4017 current resource usage; turning this on also shows 4018 possible settings and some assignment information. 4019 4020 pnpacpi= [ACPI] 4021 { off } 4022 4023 pnpbios= [ISAPNP] 4024 { on | off | curr | res | no-curr | no-res } 4025 4026 pnp_reserve_irq= 4027 [ISAPNP] Exclude IRQs for the autoconfiguration 4028 4029 pnp_reserve_dma= 4030 [ISAPNP] Exclude DMAs for the autoconfiguration 4031 4032 pnp_reserve_io= [ISAPNP] Exclude I/O ports for the autoconfiguration 4033 Ranges are in pairs (I/O port base and size). 4034 4035 pnp_reserve_mem= 4036 [ISAPNP] Exclude memory regions for the 4037 autoconfiguration. 4038 Ranges are in pairs (memory base and size). 4039 4040 ports= [IP_VS_FTP] IPVS ftp helper module 4041 Default is 21. 4042 Up to 8 (IP_VS_APP_MAX_PORTS) ports 4043 may be specified. 4044 Format: <port>,<port>.... 4045 4046 powersave=off [PPC] This option disables power saving features. 4047 It specifically disables cpuidle and sets the 4048 platform machine description specific power_save 4049 function to NULL. On Idle the CPU just reduces 4050 execution priority. 4051 4052 ppc_strict_facility_enable 4053 [PPC] This option catches any kernel floating point, 4054 Altivec, VSX and SPE outside of regions specifically 4055 allowed (eg kernel_enable_fpu()/kernel_disable_fpu()). 4056 There is some performance impact when enabling this. 4057 4058 ppc_tm= [PPC] 4059 Format: {"off"} 4060 Disable Hardware Transactional Memory 4061 4062 print-fatal-signals= 4063 [KNL] debug: print fatal signals 4064 4065 If enabled, warn about various signal handling 4066 related application anomalies: too many signals, 4067 too many POSIX.1 timers, fatal signals causing a 4068 coredump - etc. 4069 4070 If you hit the warning due to signal overflow, 4071 you might want to try "ulimit -i unlimited". 4072 4073 default: off. 4074 4075 printk.always_kmsg_dump= 4076 Trigger kmsg_dump for cases other than kernel oops or 4077 panics 4078 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4079 default: disabled 4080 4081 printk.devkmsg={on,off,ratelimit} 4082 Control writing to /dev/kmsg. 4083 on - unlimited logging to /dev/kmsg from userspace 4084 off - logging to /dev/kmsg disabled 4085 ratelimit - ratelimit the logging 4086 Default: ratelimit 4087 4088 printk.time= Show timing data prefixed to each printk message line 4089 Format: <bool> (1/Y/y=enable, 0/N/n=disable) 4090 4091 processor.max_cstate= [HW,ACPI] 4092 Limit processor to maximum C-state 4093 max_cstate=9 overrides any DMI blacklist limit. 4094 4095 processor.nocst [HW,ACPI] 4096 Ignore the _CST method to determine C-states, 4097 instead using the legacy FADT method 4098 4099 profile= [KNL] Enable kernel profiling via /proc/profile 4100 Format: [<profiletype>,]<number> 4101 Param: <profiletype>: "schedule", "sleep", or "kvm" 4102 [defaults to kernel profiling] 4103 Param: "schedule" - profile schedule points. 4104 Param: "sleep" - profile D-state sleeping (millisecs). 4105 Requires CONFIG_SCHEDSTATS 4106 Param: "kvm" - profile VM exits. 4107 Param: <number> - step/bucket size as a power of 2 for 4108 statistical time based profiling. 4109 4110 prompt_ramdisk= [RAM] [Deprecated] 4111 4112 prot_virt= [S390] enable hosting protected virtual machines 4113 isolated from the hypervisor (if hardware supports 4114 that). 4115 Format: <bool> 4116 4117 psi= [KNL] Enable or disable pressure stall information 4118 tracking. 4119 Format: <bool> 4120 4121 psmouse.proto= [HW,MOUSE] Highest PS2 mouse protocol extension to 4122 probe for; one of (bare|imps|exps|lifebook|any). 4123 psmouse.rate= [HW,MOUSE] Set desired mouse report rate, in reports 4124 per second. 4125 psmouse.resetafter= [HW,MOUSE] 4126 Try to reset the device after so many bad packets 4127 (0 = never). 4128 psmouse.resolution= 4129 [HW,MOUSE] Set desired mouse resolution, in dpi. 4130 psmouse.smartscroll= 4131 [HW,MOUSE] Controls Logitech smartscroll autorepeat. 4132 0 = disabled, 1 = enabled (default). 4133 4134 pstore.backend= Specify the name of the pstore backend to use 4135 4136 pt. [PARIDE] 4137 See Documentation/admin-guide/blockdev/paride.rst. 4138 4139 pti= [X86-64] Control Page Table Isolation of user and 4140 kernel address spaces. Disabling this feature 4141 removes hardening, but improves performance of 4142 system calls and interrupts. 4143 4144 on - unconditionally enable 4145 off - unconditionally disable 4146 auto - kernel detects whether your CPU model is 4147 vulnerable to issues that PTI mitigates 4148 4149 Not specifying this option is equivalent to pti=auto. 4150 4151 nopti [X86-64] 4152 Equivalent to pti=off 4153 4154 pty.legacy_count= 4155 [KNL] Number of legacy pty's. Overwrites compiled-in 4156 default number. 4157 4158 quiet [KNL] Disable most log messages 4159 4160 r128= [HW,DRM] 4161 4162 raid= [HW,RAID] 4163 See Documentation/admin-guide/md.rst. 4164 4165 ramdisk_size= [RAM] Sizes of RAM disks in kilobytes 4166 See Documentation/admin-guide/blockdev/ramdisk.rst. 4167 4168 ramdisk_start= [RAM] RAM disk image start address 4169 4170 random.trust_cpu={on,off} 4171 [KNL] Enable or disable trusting the use of the 4172 CPU's random number generator (if available) to 4173 fully seed the kernel's CRNG. Default is controlled 4174 by CONFIG_RANDOM_TRUST_CPU. 4175 4176 random.trust_bootloader={on,off} 4177 [KNL] Enable or disable trusting the use of a 4178 seed passed by the bootloader (if available) to 4179 fully seed the kernel's CRNG. Default is controlled 4180 by CONFIG_RANDOM_TRUST_BOOTLOADER. 4181 4182 ras=option[,option,...] [KNL] RAS-specific options 4183 4184 cec_disable [X86] 4185 Disable the Correctable Errors Collector, 4186 see CONFIG_RAS_CEC help text. 4187 4188 rcu_nocbs= [KNL] 4189 The argument is a cpu list, as described above, 4190 except that the string "all" can be used to 4191 specify every CPU on the system. 4192 4193 In kernels built with CONFIG_RCU_NOCB_CPU=y, set 4194 the specified list of CPUs to be no-callback CPUs. 4195 Invocation of these CPUs' RCU callbacks will be 4196 offloaded to "rcuox/N" kthreads created for that 4197 purpose, where "x" is "p" for RCU-preempt, and 4198 "s" for RCU-sched, and "N" is the CPU number. 4199 This reduces OS jitter on the offloaded CPUs, 4200 which can be useful for HPC and real-time 4201 workloads. It can also improve energy efficiency 4202 for asymmetric multiprocessors. 4203 4204 rcu_nocb_poll [KNL] 4205 Rather than requiring that offloaded CPUs 4206 (specified by rcu_nocbs= above) explicitly 4207 awaken the corresponding "rcuoN" kthreads, 4208 make these kthreads poll for callbacks. 4209 This improves the real-time response for the 4210 offloaded CPUs by relieving them of the need to 4211 wake up the corresponding kthread, but degrades 4212 energy efficiency by requiring that the kthreads 4213 periodically wake up to do the polling. 4214 4215 rcutree.blimit= [KNL] 4216 Set maximum number of finished RCU callbacks to 4217 process in one batch. 4218 4219 rcutree.dump_tree= [KNL] 4220 Dump the structure of the rcu_node combining tree 4221 out at early boot. This is used for diagnostic 4222 purposes, to verify correct tree setup. 4223 4224 rcutree.gp_cleanup_delay= [KNL] 4225 Set the number of jiffies to delay each step of 4226 RCU grace-period cleanup. 4227 4228 rcutree.gp_init_delay= [KNL] 4229 Set the number of jiffies to delay each step of 4230 RCU grace-period initialization. 4231 4232 rcutree.gp_preinit_delay= [KNL] 4233 Set the number of jiffies to delay each step of 4234 RCU grace-period pre-initialization, that is, 4235 the propagation of recent CPU-hotplug changes up 4236 the rcu_node combining tree. 4237 4238 rcutree.use_softirq= [KNL] 4239 If set to zero, move all RCU_SOFTIRQ processing to 4240 per-CPU rcuc kthreads. Defaults to a non-zero 4241 value, meaning that RCU_SOFTIRQ is used by default. 4242 Specify rcutree.use_softirq=0 to use rcuc kthreads. 4243 4244 rcutree.rcu_fanout_exact= [KNL] 4245 Disable autobalancing of the rcu_node combining 4246 tree. This is used by rcutorture, and might 4247 possibly be useful for architectures having high 4248 cache-to-cache transfer latencies. 4249 4250 rcutree.rcu_fanout_leaf= [KNL] 4251 Change the number of CPUs assigned to each 4252 leaf rcu_node structure. Useful for very 4253 large systems, which will choose the value 64, 4254 and for NUMA systems with large remote-access 4255 latencies, which will choose a value aligned 4256 with the appropriate hardware boundaries. 4257 4258 rcutree.rcu_min_cached_objs= [KNL] 4259 Minimum number of objects which are cached and 4260 maintained per one CPU. Object size is equal 4261 to PAGE_SIZE. The cache allows to reduce the 4262 pressure to page allocator, also it makes the 4263 whole algorithm to behave better in low memory 4264 condition. 4265 4266 rcutree.jiffies_till_first_fqs= [KNL] 4267 Set delay from grace-period initialization to 4268 first attempt to force quiescent states. 4269 Units are jiffies, minimum value is zero, 4270 and maximum value is HZ. 4271 4272 rcutree.jiffies_till_next_fqs= [KNL] 4273 Set delay between subsequent attempts to force 4274 quiescent states. Units are jiffies, minimum 4275 value is one, and maximum value is HZ. 4276 4277 rcutree.jiffies_till_sched_qs= [KNL] 4278 Set required age in jiffies for a 4279 given grace period before RCU starts 4280 soliciting quiescent-state help from 4281 rcu_note_context_switch() and cond_resched(). 4282 If not specified, the kernel will calculate 4283 a value based on the most recent settings 4284 of rcutree.jiffies_till_first_fqs 4285 and rcutree.jiffies_till_next_fqs. 4286 This calculated value may be viewed in 4287 rcutree.jiffies_to_sched_qs. Any attempt to set 4288 rcutree.jiffies_to_sched_qs will be cheerfully 4289 overwritten. 4290 4291 rcutree.kthread_prio= [KNL,BOOT] 4292 Set the SCHED_FIFO priority of the RCU per-CPU 4293 kthreads (rcuc/N). This value is also used for 4294 the priority of the RCU boost threads (rcub/N) 4295 and for the RCU grace-period kthreads (rcu_bh, 4296 rcu_preempt, and rcu_sched). If RCU_BOOST is 4297 set, valid values are 1-99 and the default is 1 4298 (the least-favored priority). Otherwise, when 4299 RCU_BOOST is not set, valid values are 0-99 and 4300 the default is zero (non-realtime operation). 4301 4302 rcutree.rcu_nocb_gp_stride= [KNL] 4303 Set the number of NOCB callback kthreads in 4304 each group, which defaults to the square root 4305 of the number of CPUs. Larger numbers reduce 4306 the wakeup overhead on the global grace-period 4307 kthread, but increases that same overhead on 4308 each group's NOCB grace-period kthread. 4309 4310 rcutree.qhimark= [KNL] 4311 Set threshold of queued RCU callbacks beyond which 4312 batch limiting is disabled. 4313 4314 rcutree.qlowmark= [KNL] 4315 Set threshold of queued RCU callbacks below which 4316 batch limiting is re-enabled. 4317 4318 rcutree.qovld= [KNL] 4319 Set threshold of queued RCU callbacks beyond which 4320 RCU's force-quiescent-state scan will aggressively 4321 enlist help from cond_resched() and sched IPIs to 4322 help CPUs more quickly reach quiescent states. 4323 Set to less than zero to make this be set based 4324 on rcutree.qhimark at boot time and to zero to 4325 disable more aggressive help enlistment. 4326 4327 rcutree.rcu_idle_gp_delay= [KNL] 4328 Set wakeup interval for idle CPUs that have 4329 RCU callbacks (RCU_FAST_NO_HZ=y). 4330 4331 rcutree.rcu_idle_lazy_gp_delay= [KNL] 4332 Set wakeup interval for idle CPUs that have 4333 only "lazy" RCU callbacks (RCU_FAST_NO_HZ=y). 4334 Lazy RCU callbacks are those which RCU can 4335 prove do nothing more than free memory. 4336 4337 rcutree.rcu_kick_kthreads= [KNL] 4338 Cause the grace-period kthread to get an extra 4339 wake_up() if it sleeps three times longer than 4340 it should at force-quiescent-state time. 4341 This wake_up() will be accompanied by a 4342 WARN_ONCE() splat and an ftrace_dump(). 4343 4344 rcutree.rcu_unlock_delay= [KNL] 4345 In CONFIG_RCU_STRICT_GRACE_PERIOD=y kernels, 4346 this specifies an rcu_read_unlock()-time delay 4347 in microseconds. This defaults to zero. 4348 Larger delays increase the probability of 4349 catching RCU pointer leaks, that is, buggy use 4350 of RCU-protected pointers after the relevant 4351 rcu_read_unlock() has completed. 4352 4353 rcutree.sysrq_rcu= [KNL] 4354 Commandeer a sysrq key to dump out Tree RCU's 4355 rcu_node tree with an eye towards determining 4356 why a new grace period has not yet started. 4357 4358 rcuscale.gp_async= [KNL] 4359 Measure performance of asynchronous 4360 grace-period primitives such as call_rcu(). 4361 4362 rcuscale.gp_async_max= [KNL] 4363 Specify the maximum number of outstanding 4364 callbacks per writer thread. When a writer 4365 thread exceeds this limit, it invokes the 4366 corresponding flavor of rcu_barrier() to allow 4367 previously posted callbacks to drain. 4368 4369 rcuscale.gp_exp= [KNL] 4370 Measure performance of expedited synchronous 4371 grace-period primitives. 4372 4373 rcuscale.holdoff= [KNL] 4374 Set test-start holdoff period. The purpose of 4375 this parameter is to delay the start of the 4376 test until boot completes in order to avoid 4377 interference. 4378 4379 rcuscale.kfree_rcu_test= [KNL] 4380 Set to measure performance of kfree_rcu() flooding. 4381 4382 rcuscale.kfree_nthreads= [KNL] 4383 The number of threads running loops of kfree_rcu(). 4384 4385 rcuscale.kfree_alloc_num= [KNL] 4386 Number of allocations and frees done in an iteration. 4387 4388 rcuscale.kfree_loops= [KNL] 4389 Number of loops doing rcuscale.kfree_alloc_num number 4390 of allocations and frees. 4391 4392 rcuscale.nreaders= [KNL] 4393 Set number of RCU readers. The value -1 selects 4394 N, where N is the number of CPUs. A value 4395 "n" less than -1 selects N-n+1, where N is again 4396 the number of CPUs. For example, -2 selects N 4397 (the number of CPUs), -3 selects N+1, and so on. 4398 A value of "n" less than or equal to -N selects 4399 a single reader. 4400 4401 rcuscale.nwriters= [KNL] 4402 Set number of RCU writers. The values operate 4403 the same as for rcuscale.nreaders. 4404 N, where N is the number of CPUs 4405 4406 rcuscale.perf_type= [KNL] 4407 Specify the RCU implementation to test. 4408 4409 rcuscale.shutdown= [KNL] 4410 Shut the system down after performance tests 4411 complete. This is useful for hands-off automated 4412 testing. 4413 4414 rcuscale.verbose= [KNL] 4415 Enable additional printk() statements. 4416 4417 rcuscale.writer_holdoff= [KNL] 4418 Write-side holdoff between grace periods, 4419 in microseconds. The default of zero says 4420 no holdoff. 4421 4422 rcutorture.fqs_duration= [KNL] 4423 Set duration of force_quiescent_state bursts 4424 in microseconds. 4425 4426 rcutorture.fqs_holdoff= [KNL] 4427 Set holdoff time within force_quiescent_state bursts 4428 in microseconds. 4429 4430 rcutorture.fqs_stutter= [KNL] 4431 Set wait time between force_quiescent_state bursts 4432 in seconds. 4433 4434 rcutorture.fwd_progress= [KNL] 4435 Enable RCU grace-period forward-progress testing 4436 for the types of RCU supporting this notion. 4437 4438 rcutorture.fwd_progress_div= [KNL] 4439 Specify the fraction of a CPU-stall-warning 4440 period to do tight-loop forward-progress testing. 4441 4442 rcutorture.fwd_progress_holdoff= [KNL] 4443 Number of seconds to wait between successive 4444 forward-progress tests. 4445 4446 rcutorture.fwd_progress_need_resched= [KNL] 4447 Enclose cond_resched() calls within checks for 4448 need_resched() during tight-loop forward-progress 4449 testing. 4450 4451 rcutorture.gp_cond= [KNL] 4452 Use conditional/asynchronous update-side 4453 primitives, if available. 4454 4455 rcutorture.gp_exp= [KNL] 4456 Use expedited update-side primitives, if available. 4457 4458 rcutorture.gp_normal= [KNL] 4459 Use normal (non-expedited) asynchronous 4460 update-side primitives, if available. 4461 4462 rcutorture.gp_sync= [KNL] 4463 Use normal (non-expedited) synchronous 4464 update-side primitives, if available. If all 4465 of rcutorture.gp_cond=, rcutorture.gp_exp=, 4466 rcutorture.gp_normal=, and rcutorture.gp_sync= 4467 are zero, rcutorture acts as if is interpreted 4468 they are all non-zero. 4469 4470 rcutorture.irqreader= [KNL] 4471 Run RCU readers from irq handlers, or, more 4472 accurately, from a timer handler. Not all RCU 4473 flavors take kindly to this sort of thing. 4474 4475 rcutorture.leakpointer= [KNL] 4476 Leak an RCU-protected pointer out of the reader. 4477 This can of course result in splats, and is 4478 intended to test the ability of things like 4479 CONFIG_RCU_STRICT_GRACE_PERIOD=y to detect 4480 such leaks. 4481 4482 rcutorture.n_barrier_cbs= [KNL] 4483 Set callbacks/threads for rcu_barrier() testing. 4484 4485 rcutorture.nfakewriters= [KNL] 4486 Set number of concurrent RCU writers. These just 4487 stress RCU, they don't participate in the actual 4488 test, hence the "fake". 4489 4490 rcutorture.nreaders= [KNL] 4491 Set number of RCU readers. The value -1 selects 4492 N-1, where N is the number of CPUs. A value 4493 "n" less than -1 selects N-n-2, where N is again 4494 the number of CPUs. For example, -2 selects N 4495 (the number of CPUs), -3 selects N+1, and so on. 4496 4497 rcutorture.object_debug= [KNL] 4498 Enable debug-object double-call_rcu() testing. 4499 4500 rcutorture.onoff_holdoff= [KNL] 4501 Set time (s) after boot for CPU-hotplug testing. 4502 4503 rcutorture.onoff_interval= [KNL] 4504 Set time (jiffies) between CPU-hotplug operations, 4505 or zero to disable CPU-hotplug testing. 4506 4507 rcutorture.read_exit= [KNL] 4508 Set the number of read-then-exit kthreads used 4509 to test the interaction of RCU updaters and 4510 task-exit processing. 4511 4512 rcutorture.read_exit_burst= [KNL] 4513 The number of times in a given read-then-exit 4514 episode that a set of read-then-exit kthreads 4515 is spawned. 4516 4517 rcutorture.read_exit_delay= [KNL] 4518 The delay, in seconds, between successive 4519 read-then-exit testing episodes. 4520 4521 rcutorture.shuffle_interval= [KNL] 4522 Set task-shuffle interval (s). Shuffling tasks 4523 allows some CPUs to go into dyntick-idle mode 4524 during the rcutorture test. 4525 4526 rcutorture.shutdown_secs= [KNL] 4527 Set time (s) after boot system shutdown. This 4528 is useful for hands-off automated testing. 4529 4530 rcutorture.stall_cpu= [KNL] 4531 Duration of CPU stall (s) to test RCU CPU stall 4532 warnings, zero to disable. 4533 4534 rcutorture.stall_cpu_block= [KNL] 4535 Sleep while stalling if set. This will result 4536 in warnings from preemptible RCU in addition 4537 to any other stall-related activity. 4538 4539 rcutorture.stall_cpu_holdoff= [KNL] 4540 Time to wait (s) after boot before inducing stall. 4541 4542 rcutorture.stall_cpu_irqsoff= [KNL] 4543 Disable interrupts while stalling if set. 4544 4545 rcutorture.stall_gp_kthread= [KNL] 4546 Duration (s) of forced sleep within RCU 4547 grace-period kthread to test RCU CPU stall 4548 warnings, zero to disable. If both stall_cpu 4549 and stall_gp_kthread are specified, the 4550 kthread is starved first, then the CPU. 4551 4552 rcutorture.stat_interval= [KNL] 4553 Time (s) between statistics printk()s. 4554 4555 rcutorture.stutter= [KNL] 4556 Time (s) to stutter testing, for example, specifying 4557 five seconds causes the test to run for five seconds, 4558 wait for five seconds, and so on. This tests RCU's 4559 ability to transition abruptly to and from idle. 4560 4561 rcutorture.test_boost= [KNL] 4562 Test RCU priority boosting? 0=no, 1=maybe, 2=yes. 4563 "Maybe" means test if the RCU implementation 4564 under test support RCU priority boosting. 4565 4566 rcutorture.test_boost_duration= [KNL] 4567 Duration (s) of each individual boost test. 4568 4569 rcutorture.test_boost_interval= [KNL] 4570 Interval (s) between each boost test. 4571 4572 rcutorture.test_no_idle_hz= [KNL] 4573 Test RCU's dyntick-idle handling. See also the 4574 rcutorture.shuffle_interval parameter. 4575 4576 rcutorture.torture_type= [KNL] 4577 Specify the RCU implementation to test. 4578 4579 rcutorture.verbose= [KNL] 4580 Enable additional printk() statements. 4581 4582 rcupdate.rcu_cpu_stall_ftrace_dump= [KNL] 4583 Dump ftrace buffer after reporting RCU CPU 4584 stall warning. 4585 4586 rcupdate.rcu_cpu_stall_suppress= [KNL] 4587 Suppress RCU CPU stall warning messages. 4588 4589 rcupdate.rcu_cpu_stall_suppress_at_boot= [KNL] 4590 Suppress RCU CPU stall warning messages and 4591 rcutorture writer stall warnings that occur 4592 during early boot, that is, during the time 4593 before the init task is spawned. 4594 4595 rcupdate.rcu_cpu_stall_timeout= [KNL] 4596 Set timeout for RCU CPU stall warning messages. 4597 4598 rcupdate.rcu_expedited= [KNL] 4599 Use expedited grace-period primitives, for 4600 example, synchronize_rcu_expedited() instead 4601 of synchronize_rcu(). This reduces latency, 4602 but can increase CPU utilization, degrade 4603 real-time latency, and degrade energy efficiency. 4604 No effect on CONFIG_TINY_RCU kernels. 4605 4606 rcupdate.rcu_normal= [KNL] 4607 Use only normal grace-period primitives, 4608 for example, synchronize_rcu() instead of 4609 synchronize_rcu_expedited(). This improves 4610 real-time latency, CPU utilization, and 4611 energy efficiency, but can expose users to 4612 increased grace-period latency. This parameter 4613 overrides rcupdate.rcu_expedited. No effect on 4614 CONFIG_TINY_RCU kernels. 4615 4616 rcupdate.rcu_normal_after_boot= [KNL] 4617 Once boot has completed (that is, after 4618 rcu_end_inkernel_boot() has been invoked), use 4619 only normal grace-period primitives. No effect 4620 on CONFIG_TINY_RCU kernels. 4621 4622 rcupdate.rcu_task_ipi_delay= [KNL] 4623 Set time in jiffies during which RCU tasks will 4624 avoid sending IPIs, starting with the beginning 4625 of a given grace period. Setting a large 4626 number avoids disturbing real-time workloads, 4627 but lengthens grace periods. 4628 4629 rcupdate.rcu_task_stall_timeout= [KNL] 4630 Set timeout in jiffies for RCU task stall warning 4631 messages. Disable with a value less than or equal 4632 to zero. 4633 4634 rcupdate.rcu_self_test= [KNL] 4635 Run the RCU early boot self tests 4636 4637 rdinit= [KNL] 4638 Format: <full_path> 4639 Run specified binary instead of /init from the ramdisk, 4640 used for early userspace startup. See initrd. 4641 4642 rdrand= [X86] 4643 force - Override the decision by the kernel to hide the 4644 advertisement of RDRAND support (this affects 4645 certain AMD processors because of buggy BIOS 4646 support, specifically around the suspend/resume 4647 path). 4648 4649 rdt= [HW,X86,RDT] 4650 Turn on/off individual RDT features. List is: 4651 cmt, mbmtotal, mbmlocal, l3cat, l3cdp, l2cat, l2cdp, 4652 mba. 4653 E.g. to turn on cmt and turn off mba use: 4654 rdt=cmt,!mba 4655 4656 reboot= [KNL] 4657 Format (x86 or x86_64): 4658 [w[arm] | c[old] | h[ard] | s[oft] | g[pio]] \ 4659 [[,]s[mp]#### \ 4660 [[,]b[ios] | a[cpi] | k[bd] | t[riple] | e[fi] | p[ci]] \ 4661 [[,]f[orce] 4662 Where reboot_mode is one of warm (soft) or cold (hard) or gpio 4663 (prefix with 'panic_' to set mode for panic 4664 reboot only), 4665 reboot_type is one of bios, acpi, kbd, triple, efi, or pci, 4666 reboot_force is either force or not specified, 4667 reboot_cpu is s[mp]#### with #### being the processor 4668 to be used for rebooting. 4669 4670 refscale.holdoff= [KNL] 4671 Set test-start holdoff period. The purpose of 4672 this parameter is to delay the start of the 4673 test until boot completes in order to avoid 4674 interference. 4675 4676 refscale.loops= [KNL] 4677 Set the number of loops over the synchronization 4678 primitive under test. Increasing this number 4679 reduces noise due to loop start/end overhead, 4680 but the default has already reduced the per-pass 4681 noise to a handful of picoseconds on ca. 2020 4682 x86 laptops. 4683 4684 refscale.nreaders= [KNL] 4685 Set number of readers. The default value of -1 4686 selects N, where N is roughly 75% of the number 4687 of CPUs. A value of zero is an interesting choice. 4688 4689 refscale.nruns= [KNL] 4690 Set number of runs, each of which is dumped onto 4691 the console log. 4692 4693 refscale.readdelay= [KNL] 4694 Set the read-side critical-section duration, 4695 measured in microseconds. 4696 4697 refscale.scale_type= [KNL] 4698 Specify the read-protection implementation to test. 4699 4700 refscale.shutdown= [KNL] 4701 Shut down the system at the end of the performance 4702 test. This defaults to 1 (shut it down) when 4703 refscale is built into the kernel and to 0 (leave 4704 it running) when refscale is built as a module. 4705 4706 refscale.verbose= [KNL] 4707 Enable additional printk() statements. 4708 4709 relax_domain_level= 4710 [KNL, SMP] Set scheduler's default relax_domain_level. 4711 See Documentation/admin-guide/cgroup-v1/cpusets.rst. 4712 4713 reserve= [KNL,BUGS] Force kernel to ignore I/O ports or memory 4714 Format: <base1>,<size1>[,<base2>,<size2>,...] 4715 Reserve I/O ports or memory so the kernel won't use 4716 them. If <base> is less than 0x10000, the region 4717 is assumed to be I/O ports; otherwise it is memory. 4718 4719 reservetop= [X86-32] 4720 Format: nn[KMG] 4721 Reserves a hole at the top of the kernel virtual 4722 address space. 4723 4724 reservelow= [X86] 4725 Format: nn[K] 4726 Set the amount of memory to reserve for BIOS at 4727 the bottom of the address space. 4728 4729 reset_devices [KNL] Force drivers to reset the underlying device 4730 during initialization. 4731 4732 resume= [SWSUSP] 4733 Specify the partition device for software suspend 4734 Format: 4735 {/dev/<dev> | PARTUUID=<uuid> | <int>:<int> | <hex>} 4736 4737 resume_offset= [SWSUSP] 4738 Specify the offset from the beginning of the partition 4739 given by "resume=" at which the swap header is located, 4740 in <PAGE_SIZE> units (needed only for swap files). 4741 See Documentation/power/swsusp-and-swap-files.rst 4742 4743 resumedelay= [HIBERNATION] Delay (in seconds) to pause before attempting to 4744 read the resume files 4745 4746 resumewait [HIBERNATION] Wait (indefinitely) for resume device to show up. 4747 Useful for devices that are detected asynchronously 4748 (e.g. USB and MMC devices). 4749 4750 hibernate= [HIBERNATION] 4751 noresume Don't check if there's a hibernation image 4752 present during boot. 4753 nocompress Don't compress/decompress hibernation images. 4754 no Disable hibernation and resume. 4755 protect_image Turn on image protection during restoration 4756 (that will set all pages holding image data 4757 during restoration read-only). 4758 4759 retain_initrd [RAM] Keep initrd memory after extraction 4760 4761 retbleed= [X86] Control mitigation of RETBleed (Arbitrary 4762 Speculative Code Execution with Return Instructions) 4763 vulnerability. 4764 4765 AMD-based UNRET and IBPB mitigations alone do not stop 4766 sibling threads from influencing the predictions of other 4767 sibling threads. For that reason, STIBP is used on pro- 4768 cessors that support it, and mitigate SMT on processors 4769 that don't. 4770 4771 off - no mitigation 4772 auto - automatically select a migitation 4773 auto,nosmt - automatically select a mitigation, 4774 disabling SMT if necessary for 4775 the full mitigation (only on Zen1 4776 and older without STIBP). 4777 ibpb - On AMD, mitigate short speculation 4778 windows on basic block boundaries too. 4779 Safe, highest perf impact. It also 4780 enables STIBP if present. Not suitable 4781 on Intel. 4782 ibpb,nosmt - Like "ibpb" above but will disable SMT 4783 when STIBP is not available. This is 4784 the alternative for systems which do not 4785 have STIBP. 4786 unret - Force enable untrained return thunks, 4787 only effective on AMD f15h-f17h based 4788 systems. 4789 unret,nosmt - Like unret, but will disable SMT when STIBP 4790 is not available. This is the alternative for 4791 systems which do not have STIBP. 4792 4793 Selecting 'auto' will choose a mitigation method at run 4794 time according to the CPU. 4795 4796 Not specifying this option is equivalent to retbleed=auto. 4797 4798 rfkill.default_state= 4799 0 "airplane mode". All wifi, bluetooth, wimax, gps, fm, 4800 etc. communication is blocked by default. 4801 1 Unblocked. 4802 4803 rfkill.master_switch_mode= 4804 0 The "airplane mode" button does nothing. 4805 1 The "airplane mode" button toggles between everything 4806 blocked and the previous configuration. 4807 2 The "airplane mode" button toggles between everything 4808 blocked and everything unblocked. 4809 4810 rhash_entries= [KNL,NET] 4811 Set number of hash buckets for route cache 4812 4813 ring3mwait=disable 4814 [KNL] Disable ring 3 MONITOR/MWAIT feature on supported 4815 CPUs. 4816 4817 ro [KNL] Mount root device read-only on boot 4818 4819 rodata= [KNL] 4820 on Mark read-only kernel memory as read-only (default). 4821 off Leave read-only kernel memory writable for debugging. 4822 4823 rockchip.usb_uart 4824 Enable the uart passthrough on the designated usb port 4825 on Rockchip SoCs. When active, the signals of the 4826 debug-uart get routed to the D+ and D- pins of the usb 4827 port and the regular usb controller gets disabled. 4828 4829 root= [KNL] Root filesystem 4830 See name_to_dev_t comment in init/do_mounts.c. 4831 4832 rootdelay= [KNL] Delay (in seconds) to pause before attempting to 4833 mount the root filesystem 4834 4835 rootflags= [KNL] Set root filesystem mount option string 4836 4837 rootfstype= [KNL] Set root filesystem type 4838 4839 rootwait [KNL] Wait (indefinitely) for root device to show up. 4840 Useful for devices that are detected asynchronously 4841 (e.g. USB and MMC devices). 4842 4843 rproc_mem=nn[KMG][@address] 4844 [KNL,ARM,CMA] Remoteproc physical memory block. 4845 Memory area to be used by remote processor image, 4846 managed by CMA. 4847 4848 rw [KNL] Mount root device read-write on boot 4849 4850 S [KNL] Run init in single mode 4851 4852 s390_iommu= [HW,S390] 4853 Set s390 IOTLB flushing mode 4854 strict 4855 With strict flushing every unmap operation will result in 4856 an IOTLB flush. Default is lazy flushing before reuse, 4857 which is faster. 4858 4859 sa1100ir [NET] 4860 See drivers/net/irda/sa1100_ir.c. 4861 4862 sbni= [NET] Granch SBNI12 leased line adapter 4863 4864 sched_debug [KNL] Enables verbose scheduler debug messages. 4865 4866 schedstats= [KNL,X86] Enable or disable scheduled statistics. 4867 Allowed values are enable and disable. This feature 4868 incurs a small amount of overhead in the scheduler 4869 but is useful for debugging and performance tuning. 4870 4871 sched_thermal_decay_shift= 4872 [KNL, SMP] Set a decay shift for scheduler thermal 4873 pressure signal. Thermal pressure signal follows the 4874 default decay period of other scheduler pelt 4875 signals(usually 32 ms but configurable). Setting 4876 sched_thermal_decay_shift will left shift the decay 4877 period for the thermal pressure signal by the shift 4878 value. 4879 i.e. with the default pelt decay period of 32 ms 4880 sched_thermal_decay_shift thermal pressure decay pr 4881 1 64 ms 4882 2 128 ms 4883 and so on. 4884 Format: integer between 0 and 10 4885 Default is 0. 4886 4887 scftorture.holdoff= [KNL] 4888 Number of seconds to hold off before starting 4889 test. Defaults to zero for module insertion and 4890 to 10 seconds for built-in smp_call_function() 4891 tests. 4892 4893 scftorture.longwait= [KNL] 4894 Request ridiculously long waits randomly selected 4895 up to the chosen limit in seconds. Zero (the 4896 default) disables this feature. Please note 4897 that requesting even small non-zero numbers of 4898 seconds can result in RCU CPU stall warnings, 4899 softlockup complaints, and so on. 4900 4901 scftorture.nthreads= [KNL] 4902 Number of kthreads to spawn to invoke the 4903 smp_call_function() family of functions. 4904 The default of -1 specifies a number of kthreads 4905 equal to the number of CPUs. 4906 4907 scftorture.onoff_holdoff= [KNL] 4908 Number seconds to wait after the start of the 4909 test before initiating CPU-hotplug operations. 4910 4911 scftorture.onoff_interval= [KNL] 4912 Number seconds to wait between successive 4913 CPU-hotplug operations. Specifying zero (which 4914 is the default) disables CPU-hotplug operations. 4915 4916 scftorture.shutdown_secs= [KNL] 4917 The number of seconds following the start of the 4918 test after which to shut down the system. The 4919 default of zero avoids shutting down the system. 4920 Non-zero values are useful for automated tests. 4921 4922 scftorture.stat_interval= [KNL] 4923 The number of seconds between outputting the 4924 current test statistics to the console. A value 4925 of zero disables statistics output. 4926 4927 scftorture.stutter_cpus= [KNL] 4928 The number of jiffies to wait between each change 4929 to the set of CPUs under test. 4930 4931 scftorture.use_cpus_read_lock= [KNL] 4932 Use use_cpus_read_lock() instead of the default 4933 preempt_disable() to disable CPU hotplug 4934 while invoking one of the smp_call_function*() 4935 functions. 4936 4937 scftorture.verbose= [KNL] 4938 Enable additional printk() statements. 4939 4940 scftorture.weight_single= [KNL] 4941 The probability weighting to use for the 4942 smp_call_function_single() function with a zero 4943 "wait" parameter. A value of -1 selects the 4944 default if all other weights are -1. However, 4945 if at least one weight has some other value, a 4946 value of -1 will instead select a weight of zero. 4947 4948 scftorture.weight_single_wait= [KNL] 4949 The probability weighting to use for the 4950 smp_call_function_single() function with a 4951 non-zero "wait" parameter. See weight_single. 4952 4953 scftorture.weight_many= [KNL] 4954 The probability weighting to use for the 4955 smp_call_function_many() function with a zero 4956 "wait" parameter. See weight_single. 4957 Note well that setting a high probability for 4958 this weighting can place serious IPI load 4959 on the system. 4960 4961 scftorture.weight_many_wait= [KNL] 4962 The probability weighting to use for the 4963 smp_call_function_many() function with a 4964 non-zero "wait" parameter. See weight_single 4965 and weight_many. 4966 4967 scftorture.weight_all= [KNL] 4968 The probability weighting to use for the 4969 smp_call_function_all() function with a zero 4970 "wait" parameter. See weight_single and 4971 weight_many. 4972 4973 scftorture.weight_all_wait= [KNL] 4974 The probability weighting to use for the 4975 smp_call_function_all() function with a 4976 non-zero "wait" parameter. See weight_single 4977 and weight_many. 4978 4979 skew_tick= [KNL] Offset the periodic timer tick per cpu to mitigate 4980 xtime_lock contention on larger systems, and/or RCU lock 4981 contention on all systems with CONFIG_MAXSMP set. 4982 Format: { "0" | "1" } 4983 0 -- disable. (may be 1 via CONFIG_CMDLINE="skew_tick=1" 4984 1 -- enable. 4985 Note: increases power consumption, thus should only be 4986 enabled if running jitter sensitive (HPC/RT) workloads. 4987 4988 security= [SECURITY] Choose a legacy "major" security module to 4989 enable at boot. This has been deprecated by the 4990 "lsm=" parameter. 4991 4992 selinux= [SELINUX] Disable or enable SELinux at boot time. 4993 Format: { "0" | "1" } 4994 See security/selinux/Kconfig help text. 4995 0 -- disable. 4996 1 -- enable. 4997 Default value is 1. 4998 4999 apparmor= [APPARMOR] Disable or enable AppArmor at boot time 5000 Format: { "0" | "1" } 5001 See security/apparmor/Kconfig help text 5002 0 -- disable. 5003 1 -- enable. 5004 Default value is set via kernel config option. 5005 5006 serialnumber [BUGS=X86-32] 5007 5008 shapers= [NET] 5009 Maximal number of shapers. 5010 5011 simeth= [IA-64] 5012 simscsi= 5013 5014 slram= [HW,MTD] 5015 5016 slab_nomerge [MM] 5017 Disable merging of slabs with similar size. May be 5018 necessary if there is some reason to distinguish 5019 allocs to different slabs, especially in hardened 5020 environments where the risk of heap overflows and 5021 layout control by attackers can usually be 5022 frustrated by disabling merging. This will reduce 5023 most of the exposure of a heap attack to a single 5024 cache (risks via metadata attacks are mostly 5025 unchanged). Debug options disable merging on their 5026 own. 5027 For more information see Documentation/vm/slub.rst. 5028 5029 slab_max_order= [MM, SLAB] 5030 Determines the maximum allowed order for slabs. 5031 A high setting may cause OOMs due to memory 5032 fragmentation. Defaults to 1 for systems with 5033 more than 32MB of RAM, 0 otherwise. 5034 5035 slub_debug[=options[,slabs][;[options[,slabs]]...] [MM, SLUB] 5036 Enabling slub_debug allows one to determine the 5037 culprit if slab objects become corrupted. Enabling 5038 slub_debug can create guard zones around objects and 5039 may poison objects when not in use. Also tracks the 5040 last alloc / free. For more information see 5041 Documentation/vm/slub.rst. 5042 5043 slub_memcg_sysfs= [MM, SLUB] 5044 Determines whether to enable sysfs directories for 5045 memory cgroup sub-caches. 1 to enable, 0 to disable. 5046 The default is determined by CONFIG_SLUB_MEMCG_SYSFS_ON. 5047 Enabling this can lead to a very high number of debug 5048 directories and files being created under 5049 /sys/kernel/slub. 5050 5051 slub_max_order= [MM, SLUB] 5052 Determines the maximum allowed order for slabs. 5053 A high setting may cause OOMs due to memory 5054 fragmentation. For more information see 5055 Documentation/vm/slub.rst. 5056 5057 slub_min_objects= [MM, SLUB] 5058 The minimum number of objects per slab. SLUB will 5059 increase the slab order up to slub_max_order to 5060 generate a sufficiently large slab able to contain 5061 the number of objects indicated. The higher the number 5062 of objects the smaller the overhead of tracking slabs 5063 and the less frequently locks need to be acquired. 5064 For more information see Documentation/vm/slub.rst. 5065 5066 slub_min_order= [MM, SLUB] 5067 Determines the minimum page order for slabs. Must be 5068 lower than slub_max_order. 5069 For more information see Documentation/vm/slub.rst. 5070 5071 slub_nomerge [MM, SLUB] 5072 Same with slab_nomerge. This is supported for legacy. 5073 See slab_nomerge for more information. 5074 5075 smart2= [HW] 5076 Format: <io1>[,<io2>[,...,<io8>]] 5077 5078 smsc-ircc2.nopnp [HW] Don't use PNP to discover SMC devices 5079 smsc-ircc2.ircc_cfg= [HW] Device configuration I/O port 5080 smsc-ircc2.ircc_sir= [HW] SIR base I/O port 5081 smsc-ircc2.ircc_fir= [HW] FIR base I/O port 5082 smsc-ircc2.ircc_irq= [HW] IRQ line 5083 smsc-ircc2.ircc_dma= [HW] DMA channel 5084 smsc-ircc2.ircc_transceiver= [HW] Transceiver type: 5085 0: Toshiba Satellite 1800 (GP data pin select) 5086 1: Fast pin select (default) 5087 2: ATC IRMode 5088 5089 smt [KNL,S390] Set the maximum number of threads (logical 5090 CPUs) to use per physical CPU on systems capable of 5091 symmetric multithreading (SMT). Will be capped to the 5092 actual hardware limit. 5093 Format: <integer> 5094 Default: -1 (no limit) 5095 5096 softlockup_panic= 5097 [KNL] Should the soft-lockup detector generate panics. 5098 Format: 0 | 1 5099 5100 A value of 1 instructs the soft-lockup detector 5101 to panic the machine when a soft-lockup occurs. It is 5102 also controlled by the kernel.softlockup_panic sysctl 5103 and CONFIG_BOOTPARAM_SOFTLOCKUP_PANIC, which is the 5104 respective build-time switch to that functionality. 5105 5106 softlockup_all_cpu_backtrace= 5107 [KNL] Should the soft-lockup detector generate 5108 backtraces on all cpus. 5109 Format: 0 | 1 5110 5111 sonypi.*= [HW] Sony Programmable I/O Control Device driver 5112 See Documentation/admin-guide/laptops/sonypi.rst 5113 5114 spectre_v2= [X86] Control mitigation of Spectre variant 2 5115 (indirect branch speculation) vulnerability. 5116 The default operation protects the kernel from 5117 user space attacks. 5118 5119 on - unconditionally enable, implies 5120 spectre_v2_user=on 5121 off - unconditionally disable, implies 5122 spectre_v2_user=off 5123 auto - kernel detects whether your CPU model is 5124 vulnerable 5125 5126 Selecting 'on' will, and 'auto' may, choose a 5127 mitigation method at run time according to the 5128 CPU, the available microcode, the setting of the 5129 CONFIG_RETPOLINE configuration option, and the 5130 compiler with which the kernel was built. 5131 5132 Selecting 'on' will also enable the mitigation 5133 against user space to user space task attacks. 5134 5135 Selecting 'off' will disable both the kernel and 5136 the user space protections. 5137 5138 Specific mitigations can also be selected manually: 5139 5140 retpoline - replace indirect branches 5141 retpoline,generic - Retpolines 5142 retpoline,lfence - LFENCE; indirect branch 5143 retpoline,amd - alias for retpoline,lfence 5144 eibrs - enhanced IBRS 5145 eibrs,retpoline - enhanced IBRS + Retpolines 5146 eibrs,lfence - enhanced IBRS + LFENCE 5147 ibrs - use IBRS to protect kernel 5148 5149 Not specifying this option is equivalent to 5150 spectre_v2=auto. 5151 5152 spectre_v2_user= 5153 [X86] Control mitigation of Spectre variant 2 5154 (indirect branch speculation) vulnerability between 5155 user space tasks 5156 5157 on - Unconditionally enable mitigations. Is 5158 enforced by spectre_v2=on 5159 5160 off - Unconditionally disable mitigations. Is 5161 enforced by spectre_v2=off 5162 5163 prctl - Indirect branch speculation is enabled, 5164 but mitigation can be enabled via prctl 5165 per thread. The mitigation control state 5166 is inherited on fork. 5167 5168 prctl,ibpb 5169 - Like "prctl" above, but only STIBP is 5170 controlled per thread. IBPB is issued 5171 always when switching between different user 5172 space processes. 5173 5174 seccomp 5175 - Same as "prctl" above, but all seccomp 5176 threads will enable the mitigation unless 5177 they explicitly opt out. 5178 5179 seccomp,ibpb 5180 - Like "seccomp" above, but only STIBP is 5181 controlled per thread. IBPB is issued 5182 always when switching between different 5183 user space processes. 5184 5185 auto - Kernel selects the mitigation depending on 5186 the available CPU features and vulnerability. 5187 5188 Default mitigation: 5189 If CONFIG_SECCOMP=y then "seccomp", otherwise "prctl" 5190 5191 Not specifying this option is equivalent to 5192 spectre_v2_user=auto. 5193 5194 spec_store_bypass_disable= 5195 [HW] Control Speculative Store Bypass (SSB) Disable mitigation 5196 (Speculative Store Bypass vulnerability) 5197 5198 Certain CPUs are vulnerable to an exploit against a 5199 a common industry wide performance optimization known 5200 as "Speculative Store Bypass" in which recent stores 5201 to the same memory location may not be observed by 5202 later loads during speculative execution. The idea 5203 is that such stores are unlikely and that they can 5204 be detected prior to instruction retirement at the 5205 end of a particular speculation execution window. 5206 5207 In vulnerable processors, the speculatively forwarded 5208 store can be used in a cache side channel attack, for 5209 example to read memory to which the attacker does not 5210 directly have access (e.g. inside sandboxed code). 5211 5212 This parameter controls whether the Speculative Store 5213 Bypass optimization is used. 5214 5215 On x86 the options are: 5216 5217 on - Unconditionally disable Speculative Store Bypass 5218 off - Unconditionally enable Speculative Store Bypass 5219 auto - Kernel detects whether the CPU model contains an 5220 implementation of Speculative Store Bypass and 5221 picks the most appropriate mitigation. If the 5222 CPU is not vulnerable, "off" is selected. If the 5223 CPU is vulnerable the default mitigation is 5224 architecture and Kconfig dependent. See below. 5225 prctl - Control Speculative Store Bypass per thread 5226 via prctl. Speculative Store Bypass is enabled 5227 for a process by default. The state of the control 5228 is inherited on fork. 5229 seccomp - Same as "prctl" above, but all seccomp threads 5230 will disable SSB unless they explicitly opt out. 5231 5232 Default mitigations: 5233 X86: If CONFIG_SECCOMP=y "seccomp", otherwise "prctl" 5234 5235 On powerpc the options are: 5236 5237 on,auto - On Power8 and Power9 insert a store-forwarding 5238 barrier on kernel entry and exit. On Power7 5239 perform a software flush on kernel entry and 5240 exit. 5241 off - No action. 5242 5243 Not specifying this option is equivalent to 5244 spec_store_bypass_disable=auto. 5245 5246 spia_io_base= [HW,MTD] 5247 spia_fio_base= 5248 spia_pedr= 5249 spia_peddr= 5250 5251 split_lock_detect= 5252 [X86] Enable split lock detection 5253 5254 When enabled (and if hardware support is present), atomic 5255 instructions that access data across cache line 5256 boundaries will result in an alignment check exception. 5257 5258 off - not enabled 5259 5260 warn - the kernel will emit rate limited warnings 5261 about applications triggering the #AC 5262 exception. This mode is the default on CPUs 5263 that supports split lock detection. 5264 5265 fatal - the kernel will send SIGBUS to applications 5266 that trigger the #AC exception. 5267 5268 If an #AC exception is hit in the kernel or in 5269 firmware (i.e. not while executing in user mode) 5270 the kernel will oops in either "warn" or "fatal" 5271 mode. 5272 5273 srbds= [X86,INTEL] 5274 Control the Special Register Buffer Data Sampling 5275 (SRBDS) mitigation. 5276 5277 Certain CPUs are vulnerable to an MDS-like 5278 exploit which can leak bits from the random 5279 number generator. 5280 5281 By default, this issue is mitigated by 5282 microcode. However, the microcode fix can cause 5283 the RDRAND and RDSEED instructions to become 5284 much slower. Among other effects, this will 5285 result in reduced throughput from /dev/urandom. 5286 5287 The microcode mitigation can be disabled with 5288 the following option: 5289 5290 off: Disable mitigation and remove 5291 performance impact to RDRAND and RDSEED 5292 5293 srcutree.counter_wrap_check [KNL] 5294 Specifies how frequently to check for 5295 grace-period sequence counter wrap for the 5296 srcu_data structure's ->srcu_gp_seq_needed field. 5297 The greater the number of bits set in this kernel 5298 parameter, the less frequently counter wrap will 5299 be checked for. Note that the bottom two bits 5300 are ignored. 5301 5302 srcutree.exp_holdoff [KNL] 5303 Specifies how many nanoseconds must elapse 5304 since the end of the last SRCU grace period for 5305 a given srcu_struct until the next normal SRCU 5306 grace period will be considered for automatic 5307 expediting. Set to zero to disable automatic 5308 expediting. 5309 5310 ssbd= [ARM64,HW] 5311 Speculative Store Bypass Disable control 5312 5313 On CPUs that are vulnerable to the Speculative 5314 Store Bypass vulnerability and offer a 5315 firmware based mitigation, this parameter 5316 indicates how the mitigation should be used: 5317 5318 force-on: Unconditionally enable mitigation for 5319 for both kernel and userspace 5320 force-off: Unconditionally disable mitigation for 5321 for both kernel and userspace 5322 kernel: Always enable mitigation in the 5323 kernel, and offer a prctl interface 5324 to allow userspace to register its 5325 interest in being mitigated too. 5326 5327 stack_guard_gap= [MM] 5328 override the default stack gap protection. The value 5329 is in page units and it defines how many pages prior 5330 to (for stacks growing down) resp. after (for stacks 5331 growing up) the main stack are reserved for no other 5332 mapping. Default value is 256 pages. 5333 5334 stack_depot_disable= [KNL] 5335 Setting this to true through kernel command line will 5336 disable the stack depot thereby saving the static memory 5337 consumed by the stack hash table. By default this is set 5338 to false. 5339 5340 stacktrace [FTRACE] 5341 Enabled the stack tracer on boot up. 5342 5343 stacktrace_filter=[function-list] 5344 [FTRACE] Limit the functions that the stack tracer 5345 will trace at boot up. function-list is a comma separated 5346 list of functions. This list can be changed at run 5347 time by the stack_trace_filter file in the debugfs 5348 tracing directory. Note, this enables stack tracing 5349 and the stacktrace above is not needed. 5350 5351 sti= [PARISC,HW] 5352 Format: <num> 5353 Set the STI (builtin display/keyboard on the HP-PARISC 5354 machines) console (graphic card) which should be used 5355 as the initial boot-console. 5356 See also comment in drivers/video/console/sticore.c. 5357 5358 sti_font= [HW] 5359 See comment in drivers/video/console/sticore.c. 5360 5361 stifb= [HW] 5362 Format: bpp:<bpp1>[:<bpp2>[:<bpp3>...]] 5363 5364 sunrpc.min_resvport= 5365 sunrpc.max_resvport= 5366 [NFS,SUNRPC] 5367 SunRPC servers often require that client requests 5368 originate from a privileged port (i.e. a port in the 5369 range 0 < portnr < 1024). 5370 An administrator who wishes to reserve some of these 5371 ports for other uses may adjust the range that the 5372 kernel's sunrpc client considers to be privileged 5373 using these two parameters to set the minimum and 5374 maximum port values. 5375 5376 sunrpc.svc_rpc_per_connection_limit= 5377 [NFS,SUNRPC] 5378 Limit the number of requests that the server will 5379 process in parallel from a single connection. 5380 The default value is 0 (no limit). 5381 5382 sunrpc.pool_mode= 5383 [NFS] 5384 Control how the NFS server code allocates CPUs to 5385 service thread pools. Depending on how many NICs 5386 you have and where their interrupts are bound, this 5387 option will affect which CPUs will do NFS serving. 5388 Note: this parameter cannot be changed while the 5389 NFS server is running. 5390 5391 auto the server chooses an appropriate mode 5392 automatically using heuristics 5393 global a single global pool contains all CPUs 5394 percpu one pool for each CPU 5395 pernode one pool for each NUMA node (equivalent 5396 to global on non-NUMA machines) 5397 5398 sunrpc.tcp_slot_table_entries= 5399 sunrpc.udp_slot_table_entries= 5400 [NFS,SUNRPC] 5401 Sets the upper limit on the number of simultaneous 5402 RPC calls that can be sent from the client to a 5403 server. Increasing these values may allow you to 5404 improve throughput, but will also increase the 5405 amount of memory reserved for use by the client. 5406 5407 suspend.pm_test_delay= 5408 [SUSPEND] 5409 Sets the number of seconds to remain in a suspend test 5410 mode before resuming the system (see 5411 /sys/power/pm_test). Only available when CONFIG_PM_DEBUG 5412 is set. Default value is 5. 5413 5414 svm= [PPC] 5415 Format: { on | off | y | n | 1 | 0 } 5416 This parameter controls use of the Protected 5417 Execution Facility on pSeries. 5418 5419 swapaccount=[0|1] 5420 [KNL] Enable accounting of swap in memory resource 5421 controller if no parameter or 1 is given or disable 5422 it if 0 is given (See Documentation/admin-guide/cgroup-v1/memory.rst) 5423 5424 swiotlb= [ARM,IA-64,PPC,MIPS,X86] 5425 Format: { <int> | force | noforce } 5426 <int> -- Number of I/O TLB slabs 5427 force -- force using of bounce buffers even if they 5428 wouldn't be automatically used by the kernel 5429 noforce -- Never use bounce buffers (for debugging) 5430 5431 switches= [HW,M68k] 5432 5433 sysctl.*= [KNL] 5434 Set a sysctl parameter, right before loading the init 5435 process, as if the value was written to the respective 5436 /proc/sys/... file. Both '.' and '/' are recognized as 5437 separators. Unrecognized parameters and invalid values 5438 are reported in the kernel log. Sysctls registered 5439 later by a loaded module cannot be set this way. 5440 Example: sysctl.vm.swappiness=40 5441 5442 sysfs.deprecated=0|1 [KNL] 5443 Enable/disable old style sysfs layout for old udev 5444 on older distributions. When this option is enabled 5445 very new udev will not work anymore. When this option 5446 is disabled (or CONFIG_SYSFS_DEPRECATED not compiled) 5447 in older udev will not work anymore. 5448 Default depends on CONFIG_SYSFS_DEPRECATED_V2 set in 5449 the kernel configuration. 5450 5451 sysrq_always_enabled 5452 [KNL] 5453 Ignore sysrq setting - this boot parameter will 5454 neutralize any effect of /proc/sys/kernel/sysrq. 5455 Useful for debugging. 5456 5457 tcpmhash_entries= [KNL,NET] 5458 Set the number of tcp_metrics_hash slots. 5459 Default value is 8192 or 16384 depending on total 5460 ram pages. This is used to specify the TCP metrics 5461 cache size. See Documentation/networking/ip-sysctl.rst 5462 "tcp_no_metrics_save" section for more details. 5463 5464 tdfx= [HW,DRM] 5465 5466 test_suspend= [SUSPEND][,N] 5467 Specify "mem" (for Suspend-to-RAM) or "standby" (for 5468 standby suspend) or "freeze" (for suspend type freeze) 5469 as the system sleep state during system startup with 5470 the optional capability to repeat N number of times. 5471 The system is woken from this state using a 5472 wakeup-capable RTC alarm. 5473 5474 thash_entries= [KNL,NET] 5475 Set number of hash buckets for TCP connection 5476 5477 thermal.act= [HW,ACPI] 5478 -1: disable all active trip points in all thermal zones 5479 <degrees C>: override all lowest active trip points 5480 5481 thermal.crt= [HW,ACPI] 5482 -1: disable all critical trip points in all thermal zones 5483 <degrees C>: override all critical trip points 5484 5485 thermal.nocrt= [HW,ACPI] 5486 Set to disable actions on ACPI thermal zone 5487 critical and hot trip points. 5488 5489 thermal.off= [HW,ACPI] 5490 1: disable ACPI thermal control 5491 5492 thermal.psv= [HW,ACPI] 5493 -1: disable all passive trip points 5494 <degrees C>: override all passive trip points to this 5495 value 5496 5497 thermal.tzp= [HW,ACPI] 5498 Specify global default ACPI thermal zone polling rate 5499 <deci-seconds>: poll all this frequency 5500 0: no polling (default) 5501 5502 threadirqs [KNL] 5503 Force threading of all interrupt handlers except those 5504 marked explicitly IRQF_NO_THREAD. 5505 5506 topology= [S390] 5507 Format: {off | on} 5508 Specify if the kernel should make use of the cpu 5509 topology information if the hardware supports this. 5510 The scheduler will make use of this information and 5511 e.g. base its process migration decisions on it. 5512 Default is on. 5513 5514 topology_updates= [KNL, PPC, NUMA] 5515 Format: {off} 5516 Specify if the kernel should ignore (off) 5517 topology updates sent by the hypervisor to this 5518 LPAR. 5519 5520 torture.disable_onoff_at_boot= [KNL] 5521 Prevent the CPU-hotplug component of torturing 5522 until after init has spawned. 5523 5524 torture.ftrace_dump_at_shutdown= [KNL] 5525 Dump the ftrace buffer at torture-test shutdown, 5526 even if there were no errors. This can be a 5527 very costly operation when many torture tests 5528 are running concurrently, especially on systems 5529 with rotating-rust storage. 5530 5531 tp720= [HW,PS2] 5532 5533 tpm_suspend_pcr=[HW,TPM] 5534 Format: integer pcr id 5535 Specify that at suspend time, the tpm driver 5536 should extend the specified pcr with zeros, 5537 as a workaround for some chips which fail to 5538 flush the last written pcr on TPM_SaveState. 5539 This will guarantee that all the other pcrs 5540 are saved. 5541 5542 trace_buf_size=nn[KMG] 5543 [FTRACE] will set tracing buffer size on each cpu. 5544 5545 trace_event=[event-list] 5546 [FTRACE] Set and start specified trace events in order 5547 to facilitate early boot debugging. The event-list is a 5548 comma separated list of trace events to enable. See 5549 also Documentation/trace/events.rst 5550 5551 trace_options=[option-list] 5552 [FTRACE] Enable or disable tracer options at boot. 5553 The option-list is a comma delimited list of options 5554 that can be enabled or disabled just as if you were 5555 to echo the option name into 5556 5557 /sys/kernel/debug/tracing/trace_options 5558 5559 For example, to enable stacktrace option (to dump the 5560 stack trace of each event), add to the command line: 5561 5562 trace_options=stacktrace 5563 5564 See also Documentation/trace/ftrace.rst "trace options" 5565 section. 5566 5567 tp_printk[FTRACE] 5568 Have the tracepoints sent to printk as well as the 5569 tracing ring buffer. This is useful for early boot up 5570 where the system hangs or reboots and does not give the 5571 option for reading the tracing buffer or performing a 5572 ftrace_dump_on_oops. 5573 5574 To turn off having tracepoints sent to printk, 5575 echo 0 > /proc/sys/kernel/tracepoint_printk 5576 Note, echoing 1 into this file without the 5577 tracepoint_printk kernel cmdline option has no effect. 5578 5579 ** CAUTION ** 5580 5581 Having tracepoints sent to printk() and activating high 5582 frequency tracepoints such as irq or sched, can cause 5583 the system to live lock. 5584 5585 traceoff_on_warning 5586 [FTRACE] enable this option to disable tracing when a 5587 warning is hit. This turns off "tracing_on". Tracing can 5588 be enabled again by echoing '1' into the "tracing_on" 5589 file located in /sys/kernel/debug/tracing/ 5590 5591 This option is useful, as it disables the trace before 5592 the WARNING dump is called, which prevents the trace to 5593 be filled with content caused by the warning output. 5594 5595 This option can also be set at run time via the sysctl 5596 option: kernel/traceoff_on_warning 5597 5598 transparent_hugepage= 5599 [KNL] 5600 Format: [always|madvise|never] 5601 Can be used to control the default behavior of the system 5602 with respect to transparent hugepages. 5603 See Documentation/admin-guide/mm/transhuge.rst 5604 for more details. 5605 5606 tsc= Disable clocksource stability checks for TSC. 5607 Format: <string> 5608 [x86] reliable: mark tsc clocksource as reliable, this 5609 disables clocksource verification at runtime, as well 5610 as the stability checks done at bootup. Used to enable 5611 high-resolution timer mode on older hardware, and in 5612 virtualized environment. 5613 [x86] noirqtime: Do not use TSC to do irq accounting. 5614 Used to run time disable IRQ_TIME_ACCOUNTING on any 5615 platforms where RDTSC is slow and this accounting 5616 can add overhead. 5617 [x86] unstable: mark the TSC clocksource as unstable, this 5618 marks the TSC unconditionally unstable at bootup and 5619 avoids any further wobbles once the TSC watchdog notices. 5620 [x86] nowatchdog: disable clocksource watchdog. Used 5621 in situations with strict latency requirements (where 5622 interruptions from clocksource watchdog are not 5623 acceptable). 5624 5625 tsc_early_khz= [X86] Skip early TSC calibration and use the given 5626 value instead. Useful when the early TSC frequency discovery 5627 procedure is not reliable, such as on overclocked systems 5628 with CPUID.16h support and partial CPUID.15h support. 5629 Format: <unsigned int> 5630 5631 tsx= [X86] Control Transactional Synchronization 5632 Extensions (TSX) feature in Intel processors that 5633 support TSX control. 5634 5635 This parameter controls the TSX feature. The options are: 5636 5637 on - Enable TSX on the system. Although there are 5638 mitigations for all known security vulnerabilities, 5639 TSX has been known to be an accelerator for 5640 several previous speculation-related CVEs, and 5641 so there may be unknown security risks associated 5642 with leaving it enabled. 5643 5644 off - Disable TSX on the system. (Note that this 5645 option takes effect only on newer CPUs which are 5646 not vulnerable to MDS, i.e., have 5647 MSR_IA32_ARCH_CAPABILITIES.MDS_NO=1 and which get 5648 the new IA32_TSX_CTRL MSR through a microcode 5649 update. This new MSR allows for the reliable 5650 deactivation of the TSX functionality.) 5651 5652 auto - Disable TSX if X86_BUG_TAA is present, 5653 otherwise enable TSX on the system. 5654 5655 Not specifying this option is equivalent to tsx=off. 5656 5657 See Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5658 for more details. 5659 5660 tsx_async_abort= [X86,INTEL] Control mitigation for the TSX Async 5661 Abort (TAA) vulnerability. 5662 5663 Similar to Micro-architectural Data Sampling (MDS) 5664 certain CPUs that support Transactional 5665 Synchronization Extensions (TSX) are vulnerable to an 5666 exploit against CPU internal buffers which can forward 5667 information to a disclosure gadget under certain 5668 conditions. 5669 5670 In vulnerable processors, the speculatively forwarded 5671 data can be used in a cache side channel attack, to 5672 access data to which the attacker does not have direct 5673 access. 5674 5675 This parameter controls the TAA mitigation. The 5676 options are: 5677 5678 full - Enable TAA mitigation on vulnerable CPUs 5679 if TSX is enabled. 5680 5681 full,nosmt - Enable TAA mitigation and disable SMT on 5682 vulnerable CPUs. If TSX is disabled, SMT 5683 is not disabled because CPU is not 5684 vulnerable to cross-thread TAA attacks. 5685 off - Unconditionally disable TAA mitigation 5686 5687 On MDS-affected machines, tsx_async_abort=off can be 5688 prevented by an active MDS mitigation as both vulnerabilities 5689 are mitigated with the same mechanism so in order to disable 5690 this mitigation, you need to specify mds=off too. 5691 5692 Not specifying this option is equivalent to 5693 tsx_async_abort=full. On CPUs which are MDS affected 5694 and deploy MDS mitigation, TAA mitigation is not 5695 required and doesn't provide any additional 5696 mitigation. 5697 5698 For details see: 5699 Documentation/admin-guide/hw-vuln/tsx_async_abort.rst 5700 5701 turbografx.map[2|3]= [HW,JOY] 5702 TurboGraFX parallel port interface 5703 Format: 5704 <port#>,<js1>,<js2>,<js3>,<js4>,<js5>,<js6>,<js7> 5705 See also Documentation/input/devices/joystick-parport.rst 5706 5707 udbg-immortal [PPC] When debugging early kernel crashes that 5708 happen after console_init() and before a proper 5709 console driver takes over, this boot options might 5710 help "seeing" what's going on. 5711 5712 uhash_entries= [KNL,NET] 5713 Set number of hash buckets for UDP/UDP-Lite connections 5714 5715 uhci-hcd.ignore_oc= 5716 [USB] Ignore overcurrent events (default N). 5717 Some badly-designed motherboards generate lots of 5718 bogus events, for ports that aren't wired to 5719 anything. Set this parameter to avoid log spamming. 5720 Note that genuine overcurrent events won't be 5721 reported either. 5722 5723 unknown_nmi_panic 5724 [X86] Cause panic on unknown NMI. 5725 5726 usbcore.authorized_default= 5727 [USB] Default USB device authorization: 5728 (default -1 = authorized except for wireless USB, 5729 0 = not authorized, 1 = authorized, 2 = authorized 5730 if device connected to internal port) 5731 5732 usbcore.autosuspend= 5733 [USB] The autosuspend time delay (in seconds) used 5734 for newly-detected USB devices (default 2). This 5735 is the time required before an idle device will be 5736 autosuspended. Devices for which the delay is set 5737 to a negative value won't be autosuspended at all. 5738 5739 usbcore.usbfs_snoop= 5740 [USB] Set to log all usbfs traffic (default 0 = off). 5741 5742 usbcore.usbfs_snoop_max= 5743 [USB] Maximum number of bytes to snoop in each URB 5744 (default = 65536). 5745 5746 usbcore.blinkenlights= 5747 [USB] Set to cycle leds on hubs (default 0 = off). 5748 5749 usbcore.old_scheme_first= 5750 [USB] Start with the old device initialization 5751 scheme (default 0 = off). 5752 5753 usbcore.usbfs_memory_mb= 5754 [USB] Memory limit (in MB) for buffers allocated by 5755 usbfs (default = 16, 0 = max = 2047). 5756 5757 usbcore.use_both_schemes= 5758 [USB] Try the other device initialization scheme 5759 if the first one fails (default 1 = enabled). 5760 5761 usbcore.initial_descriptor_timeout= 5762 [USB] Specifies timeout for the initial 64-byte 5763 USB_REQ_GET_DESCRIPTOR request in milliseconds 5764 (default 5000 = 5.0 seconds). 5765 5766 usbcore.nousb [USB] Disable the USB subsystem 5767 5768 usbcore.quirks= 5769 [USB] A list of quirk entries to augment the built-in 5770 usb core quirk list. List entries are separated by 5771 commas. Each entry has the form 5772 VendorID:ProductID:Flags. The IDs are 4-digit hex 5773 numbers and Flags is a set of letters. Each letter 5774 will change the built-in quirk; setting it if it is 5775 clear and clearing it if it is set. The letters have 5776 the following meanings: 5777 a = USB_QUIRK_STRING_FETCH_255 (string 5778 descriptors must not be fetched using 5779 a 255-byte read); 5780 b = USB_QUIRK_RESET_RESUME (device can't resume 5781 correctly so reset it instead); 5782 c = USB_QUIRK_NO_SET_INTF (device can't handle 5783 Set-Interface requests); 5784 d = USB_QUIRK_CONFIG_INTF_STRINGS (device can't 5785 handle its Configuration or Interface 5786 strings); 5787 e = USB_QUIRK_RESET (device can't be reset 5788 (e.g morph devices), don't use reset); 5789 f = USB_QUIRK_HONOR_BNUMINTERFACES (device has 5790 more interface descriptions than the 5791 bNumInterfaces count, and can't handle 5792 talking to these interfaces); 5793 g = USB_QUIRK_DELAY_INIT (device needs a pause 5794 during initialization, after we read 5795 the device descriptor); 5796 h = USB_QUIRK_LINEAR_UFRAME_INTR_BINTERVAL (For 5797 high speed and super speed interrupt 5798 endpoints, the USB 2.0 and USB 3.0 spec 5799 require the interval in microframes (1 5800 microframe = 125 microseconds) to be 5801 calculated as interval = 2 ^ 5802 (bInterval-1). 5803 Devices with this quirk report their 5804 bInterval as the result of this 5805 calculation instead of the exponent 5806 variable used in the calculation); 5807 i = USB_QUIRK_DEVICE_QUALIFIER (device can't 5808 handle device_qualifier descriptor 5809 requests); 5810 j = USB_QUIRK_IGNORE_REMOTE_WAKEUP (device 5811 generates spurious wakeup, ignore 5812 remote wakeup capability); 5813 k = USB_QUIRK_NO_LPM (device can't handle Link 5814 Power Management); 5815 l = USB_QUIRK_LINEAR_FRAME_INTR_BINTERVAL 5816 (Device reports its bInterval as linear 5817 frames instead of the USB 2.0 5818 calculation); 5819 m = USB_QUIRK_DISCONNECT_SUSPEND (Device needs 5820 to be disconnected before suspend to 5821 prevent spurious wakeup); 5822 n = USB_QUIRK_DELAY_CTRL_MSG (Device needs a 5823 pause after every control message); 5824 o = USB_QUIRK_HUB_SLOW_RESET (Hub needs extra 5825 delay after resetting its port); 5826 Example: quirks=0781:5580:bk,0a5c:5834:gij 5827 5828 usbhid.mousepoll= 5829 [USBHID] The interval which mice are to be polled at. 5830 5831 usbhid.jspoll= 5832 [USBHID] The interval which joysticks are to be polled at. 5833 5834 usbhid.kbpoll= 5835 [USBHID] The interval which keyboards are to be polled at. 5836 5837 usb-storage.delay_use= 5838 [UMS] The delay in seconds before a new device is 5839 scanned for Logical Units (default 1). 5840 5841 usb-storage.quirks= 5842 [UMS] A list of quirks entries to supplement or 5843 override the built-in unusual_devs list. List 5844 entries are separated by commas. Each entry has 5845 the form VID:PID:Flags where VID and PID are Vendor 5846 and Product ID values (4-digit hex numbers) and 5847 Flags is a set of characters, each corresponding 5848 to a common usb-storage quirk flag as follows: 5849 a = SANE_SENSE (collect more than 18 bytes 5850 of sense data, not on uas); 5851 b = BAD_SENSE (don't collect more than 18 5852 bytes of sense data, not on uas); 5853 c = FIX_CAPACITY (decrease the reported 5854 device capacity by one sector); 5855 d = NO_READ_DISC_INFO (don't use 5856 READ_DISC_INFO command, not on uas); 5857 e = NO_READ_CAPACITY_16 (don't use 5858 READ_CAPACITY_16 command); 5859 f = NO_REPORT_OPCODES (don't use report opcodes 5860 command, uas only); 5861 g = MAX_SECTORS_240 (don't transfer more than 5862 240 sectors at a time, uas only); 5863 h = CAPACITY_HEURISTICS (decrease the 5864 reported device capacity by one 5865 sector if the number is odd); 5866 i = IGNORE_DEVICE (don't bind to this 5867 device); 5868 j = NO_REPORT_LUNS (don't use report luns 5869 command, uas only); 5870 k = NO_SAME (do not use WRITE_SAME, uas only) 5871 l = NOT_LOCKABLE (don't try to lock and 5872 unlock ejectable media, not on uas); 5873 m = MAX_SECTORS_64 (don't transfer more 5874 than 64 sectors = 32 KB at a time, 5875 not on uas); 5876 n = INITIAL_READ10 (force a retry of the 5877 initial READ(10) command, not on uas); 5878 o = CAPACITY_OK (accept the capacity 5879 reported by the device, not on uas); 5880 p = WRITE_CACHE (the device cache is ON 5881 by default, not on uas); 5882 r = IGNORE_RESIDUE (the device reports 5883 bogus residue values, not on uas); 5884 s = SINGLE_LUN (the device has only one 5885 Logical Unit); 5886 t = NO_ATA_1X (don't allow ATA(12) and ATA(16) 5887 commands, uas only); 5888 u = IGNORE_UAS (don't bind to the uas driver); 5889 w = NO_WP_DETECT (don't test whether the 5890 medium is write-protected). 5891 y = ALWAYS_SYNC (issue a SYNCHRONIZE_CACHE 5892 even if the device claims no cache, 5893 not on uas) 5894 Example: quirks=0419:aaf5:rl,0421:0433:rc 5895 5896 user_debug= [KNL,ARM] 5897 Format: <int> 5898 See arch/arm/Kconfig.debug help text. 5899 1 - undefined instruction events 5900 2 - system calls 5901 4 - invalid data aborts 5902 8 - SIGSEGV faults 5903 16 - SIGBUS faults 5904 Example: user_debug=31 5905 5906 userpte= 5907 [X86] Flags controlling user PTE allocations. 5908 5909 nohigh = do not allocate PTE pages in 5910 HIGHMEM regardless of setting 5911 of CONFIG_HIGHPTE. 5912 5913 vdso= [X86,SH] 5914 On X86_32, this is an alias for vdso32=. Otherwise: 5915 5916 vdso=1: enable VDSO (the default) 5917 vdso=0: disable VDSO mapping 5918 5919 vdso32= [X86] Control the 32-bit vDSO 5920 vdso32=1: enable 32-bit VDSO 5921 vdso32=0 or vdso32=2: disable 32-bit VDSO 5922 5923 See the help text for CONFIG_COMPAT_VDSO for more 5924 details. If CONFIG_COMPAT_VDSO is set, the default is 5925 vdso32=0; otherwise, the default is vdso32=1. 5926 5927 For compatibility with older kernels, vdso32=2 is an 5928 alias for vdso32=0. 5929 5930 Try vdso32=0 if you encounter an error that says: 5931 dl_main: Assertion `(void *) ph->p_vaddr == _rtld_local._dl_sysinfo_dso' failed! 5932 5933 vector= [IA-64,SMP] 5934 vector=percpu: enable percpu vector domain 5935 5936 video= [FB] Frame buffer configuration 5937 See Documentation/fb/modedb.rst. 5938 5939 video.brightness_switch_enabled= [0,1] 5940 If set to 1, on receiving an ACPI notify event 5941 generated by hotkey, video driver will adjust brightness 5942 level and then send out the event to user space through 5943 the allocated input device; If set to 0, video driver 5944 will only send out the event without touching backlight 5945 brightness level. 5946 default: 1 5947 5948 virtio_mmio.device= 5949 [VMMIO] Memory mapped virtio (platform) device. 5950 5951 <size>@<baseaddr>:<irq>[:<id>] 5952 where: 5953 <size> := size (can use standard suffixes 5954 like K, M and G) 5955 <baseaddr> := physical base address 5956 <irq> := interrupt number (as passed to 5957 request_irq()) 5958 <id> := (optional) platform device id 5959 example: 5960 virtio_mmio.device=1K@0x100b0000:48:7 5961 5962 Can be used multiple times for multiple devices. 5963 5964 vga= [BOOT,X86-32] Select a particular video mode 5965 See Documentation/x86/boot.rst and 5966 Documentation/admin-guide/svga.rst. 5967 Use vga=ask for menu. 5968 This is actually a boot loader parameter; the value is 5969 passed to the kernel using a special protocol. 5970 5971 vm_debug[=options] [KNL] Available with CONFIG_DEBUG_VM=y. 5972 May slow down system boot speed, especially when 5973 enabled on systems with a large amount of memory. 5974 All options are enabled by default, and this 5975 interface is meant to allow for selectively 5976 enabling or disabling specific virtual memory 5977 debugging features. 5978 5979 Available options are: 5980 P Enable page structure init time poisoning 5981 - Disable all of the above options 5982 5983 vmalloc=nn[KMG] [KNL,BOOT] Forces the vmalloc area to have an exact 5984 size of <nn>. This can be used to increase the 5985 minimum size (128MB on x86). It can also be used to 5986 decrease the size and leave more room for directly 5987 mapped kernel RAM. 5988 5989 vmcp_cma=nn[MG] [KNL,S390] 5990 Sets the memory size reserved for contiguous memory 5991 allocations for the vmcp device driver. 5992 5993 vmhalt= [KNL,S390] Perform z/VM CP command after system halt. 5994 Format: <command> 5995 5996 vmpanic= [KNL,S390] Perform z/VM CP command after kernel panic. 5997 Format: <command> 5998 5999 vmpoff= [KNL,S390] Perform z/VM CP command after power off. 6000 Format: <command> 6001 6002 vsyscall= [X86-64] 6003 Controls the behavior of vsyscalls (i.e. calls to 6004 fixed addresses of 0xffffffffff600x00 from legacy 6005 code). Most statically-linked binaries and older 6006 versions of glibc use these calls. Because these 6007 functions are at fixed addresses, they make nice 6008 targets for exploits that can control RIP. 6009 6010 emulate [default] Vsyscalls turn into traps and are 6011 emulated reasonably safely. The vsyscall 6012 page is readable. 6013 6014 xonly Vsyscalls turn into traps and are 6015 emulated reasonably safely. The vsyscall 6016 page is not readable. 6017 6018 none Vsyscalls don't work at all. This makes 6019 them quite hard to use for exploits but 6020 might break your system. 6021 6022 vt.color= [VT] Default text color. 6023 Format: 0xYX, X = foreground, Y = background. 6024 Default: 0x07 = light gray on black. 6025 6026 vt.cur_default= [VT] Default cursor shape. 6027 Format: 0xCCBBAA, where AA, BB, and CC are the same as 6028 the parameters of the <Esc>[?A;B;Cc escape sequence; 6029 see VGA-softcursor.txt. Default: 2 = underline. 6030 6031 vt.default_blu= [VT] 6032 Format: <blue0>,<blue1>,<blue2>,...,<blue15> 6033 Change the default blue palette of the console. 6034 This is a 16-member array composed of values 6035 ranging from 0-255. 6036 6037 vt.default_grn= [VT] 6038 Format: <green0>,<green1>,<green2>,...,<green15> 6039 Change the default green palette of the console. 6040 This is a 16-member array composed of values 6041 ranging from 0-255. 6042 6043 vt.default_red= [VT] 6044 Format: <red0>,<red1>,<red2>,...,<red15> 6045 Change the default red palette of the console. 6046 This is a 16-member array composed of values 6047 ranging from 0-255. 6048 6049 vt.default_utf8= 6050 [VT] 6051 Format=<0|1> 6052 Set system-wide default UTF-8 mode for all tty's. 6053 Default is 1, i.e. UTF-8 mode is enabled for all 6054 newly opened terminals. 6055 6056 vt.global_cursor_default= 6057 [VT] 6058 Format=<-1|0|1> 6059 Set system-wide default for whether a cursor 6060 is shown on new VTs. Default is -1, 6061 i.e. cursors will be created by default unless 6062 overridden by individual drivers. 0 will hide 6063 cursors, 1 will display them. 6064 6065 vt.italic= [VT] Default color for italic text; 0-15. 6066 Default: 2 = green. 6067 6068 vt.underline= [VT] Default color for underlined text; 0-15. 6069 Default: 3 = cyan. 6070 6071 watchdog timers [HW,WDT] For information on watchdog timers, 6072 see Documentation/watchdog/watchdog-parameters.rst 6073 or other driver-specific files in the 6074 Documentation/watchdog/ directory. 6075 6076 watchdog_thresh= 6077 [KNL] 6078 Set the hard lockup detector stall duration 6079 threshold in seconds. The soft lockup detector 6080 threshold is set to twice the value. A value of 0 6081 disables both lockup detectors. Default is 10 6082 seconds. 6083 6084 workqueue.watchdog_thresh= 6085 If CONFIG_WQ_WATCHDOG is configured, workqueue can 6086 warn stall conditions and dump internal state to 6087 help debugging. 0 disables workqueue stall 6088 detection; otherwise, it's the stall threshold 6089 duration in seconds. The default value is 30 and 6090 it can be updated at runtime by writing to the 6091 corresponding sysfs file. 6092 6093 workqueue.disable_numa 6094 By default, all work items queued to unbound 6095 workqueues are affine to the NUMA nodes they're 6096 issued on, which results in better behavior in 6097 general. If NUMA affinity needs to be disabled for 6098 whatever reason, this option can be used. Note 6099 that this also can be controlled per-workqueue for 6100 workqueues visible under /sys/bus/workqueue/. 6101 6102 workqueue.power_efficient 6103 Per-cpu workqueues are generally preferred because 6104 they show better performance thanks to cache 6105 locality; unfortunately, per-cpu workqueues tend to 6106 be more power hungry than unbound workqueues. 6107 6108 Enabling this makes the per-cpu workqueues which 6109 were observed to contribute significantly to power 6110 consumption unbound, leading to measurably lower 6111 power usage at the cost of small performance 6112 overhead. 6113 6114 The default value of this parameter is determined by 6115 the config option CONFIG_WQ_POWER_EFFICIENT_DEFAULT. 6116 6117 workqueue.debug_force_rr_cpu 6118 Workqueue used to implicitly guarantee that work 6119 items queued without explicit CPU specified are put 6120 on the local CPU. This guarantee is no longer true 6121 and while local CPU is still preferred work items 6122 may be put on foreign CPUs. This debug option 6123 forces round-robin CPU selection to flush out 6124 usages which depend on the now broken guarantee. 6125 When enabled, memory and cache locality will be 6126 impacted. 6127 6128 x2apic_phys [X86-64,APIC] Use x2apic physical mode instead of 6129 default x2apic cluster mode on platforms 6130 supporting x2apic. 6131 6132 x86_intel_mid_timer= [X86-32,APBT] 6133 Choose timer option for x86 Intel MID platform. 6134 Two valid options are apbt timer only and lapic timer 6135 plus one apbt timer for broadcast timer. 6136 x86_intel_mid_timer=apbt_only | lapic_and_apbt 6137 6138 xen_512gb_limit [KNL,X86-64,XEN] 6139 Restricts the kernel running paravirtualized under Xen 6140 to use only up to 512 GB of RAM. The reason to do so is 6141 crash analysis tools and Xen tools for doing domain 6142 save/restore/migration must be enabled to handle larger 6143 domains. 6144 6145 xen_emul_unplug= [HW,X86,XEN] 6146 Unplug Xen emulated devices 6147 Format: [unplug0,][unplug1] 6148 ide-disks -- unplug primary master IDE devices 6149 aux-ide-disks -- unplug non-primary-master IDE devices 6150 nics -- unplug network devices 6151 all -- unplug all emulated devices (NICs and IDE disks) 6152 unnecessary -- unplugging emulated devices is 6153 unnecessary even if the host did not respond to 6154 the unplug protocol 6155 never -- do not unplug even if version check succeeds 6156 6157 xen_legacy_crash [X86,XEN] 6158 Crash from Xen panic notifier, without executing late 6159 panic() code such as dumping handler. 6160 6161 xen_nopvspin [X86,XEN] 6162 Disables the qspinlock slowpath using Xen PV optimizations. 6163 This parameter is obsoleted by "nopvspin" parameter, which 6164 has equivalent effect for XEN platform. 6165 6166 xen_nopv [X86] 6167 Disables the PV optimizations forcing the HVM guest to 6168 run as generic HVM guest with no PV drivers. 6169 This option is obsoleted by the "nopv" option, which 6170 has equivalent effect for XEN platform. 6171 6172 xen_no_vector_callback 6173 [KNL,X86,XEN] Disable the vector callback for Xen 6174 event channel interrupts. 6175 6176 xen_scrub_pages= [XEN] 6177 Boolean option to control scrubbing pages before giving them back 6178 to Xen, for use by other domains. Can be also changed at runtime 6179 with /sys/devices/system/xen_memory/xen_memory0/scrub_pages. 6180 Default value controlled with CONFIG_XEN_SCRUB_PAGES_DEFAULT. 6181 6182 xen_timer_slop= [X86-64,XEN] 6183 Set the timer slop (in nanoseconds) for the virtual Xen 6184 timers (default is 100000). This adjusts the minimum 6185 delta of virtualized Xen timers, where lower values 6186 improve timer resolution at the expense of processing 6187 more timer interrupts. 6188 6189 xen.balloon_boot_timeout= [XEN] 6190 The time (in seconds) to wait before giving up to boot 6191 in case initial ballooning fails to free enough memory. 6192 Applies only when running as HVM or PVH guest and 6193 started with less memory configured than allowed at 6194 max. Default is 180. 6195 6196 xen.event_eoi_delay= [XEN] 6197 How long to delay EOI handling in case of event 6198 storms (jiffies). Default is 10. 6199 6200 xen.event_loop_timeout= [XEN] 6201 After which time (jiffies) the event handling loop 6202 should start to delay EOI handling. Default is 2. 6203 6204 xen.fifo_events= [XEN] 6205 Boolean parameter to disable using fifo event handling 6206 even if available. Normally fifo event handling is 6207 preferred over the 2-level event handling, as it is 6208 fairer and the number of possible event channels is 6209 much higher. Default is on (use fifo events). 6210 6211 nopv= [X86,XEN,KVM,HYPER_V,VMWARE] 6212 Disables the PV optimizations forcing the guest to run 6213 as generic guest with no PV drivers. Currently support 6214 XEN HVM, KVM, HYPER_V and VMWARE guest. 6215 6216 nopvspin [X86,XEN,KVM] 6217 Disables the qspinlock slow path using PV optimizations 6218 which allow the hypervisor to 'idle' the guest on lock 6219 contention. 6220 6221 xirc2ps_cs= [NET,PCMCIA] 6222 Format: 6223 <irq>,<irq_mask>,<io>,<full_duplex>,<do_sound>,<lockup_hack>[,<irq2>[,<irq3>[,<irq4>]]] 6224 6225 xive= [PPC] 6226 By default on POWER9 and above, the kernel will 6227 natively use the XIVE interrupt controller. This option 6228 allows the fallback firmware mode to be used: 6229 6230 off Fallback to firmware control of XIVE interrupt 6231 controller on both pseries and powernv 6232 platforms. Only useful on POWER9 and above. 6233 6234 xhci-hcd.quirks [USB,KNL] 6235 A hex value specifying bitmask with supplemental xhci 6236 host controller quirks. Meaning of each bit can be 6237 consulted in header drivers/usb/host/xhci.h. 6238 6239 xmon [PPC] 6240 Format: { early | on | rw | ro | off } 6241 Controls if xmon debugger is enabled. Default is off. 6242 Passing only "xmon" is equivalent to "xmon=early". 6243 early Call xmon as early as possible on boot; xmon 6244 debugger is called from setup_arch(). 6245 on xmon debugger hooks will be installed so xmon 6246 is only called on a kernel crash. Default mode, 6247 i.e. either "ro" or "rw" mode, is controlled 6248 with CONFIG_XMON_DEFAULT_RO_MODE. 6249 rw xmon debugger hooks will be installed so xmon 6250 is called only on a kernel crash, mode is write, 6251 meaning SPR registers, memory and, other data 6252 can be written using xmon commands. 6253 ro same as "rw" option above but SPR registers, 6254 memory, and other data can't be written using 6255 xmon commands. 6256 off xmon is disabled. 6257