1 AMD64 specific boot options
3 There are many others (usually documented in driver documentation), but
4 only the AMD64 specific ones are listed here.
8 Please see Documentation/x86/x86_64/machinecheck for sysfs runtime tunables.
13 Disable CMCI(Corrected Machine Check Interrupt) that
14 Intel processor supports. Usually this disablement is
15 not recommended, but it might be handy if your hardware
17 Note that you'll get more problems without CMCI than with
18 due to the shared banks, i.e. you might get duplicated
21 Don't make logs for corrected errors. All events reported
22 as corrected are silently cleared by OS.
23 This option will be useful if you have no interest in any
26 Disable features for corrected errors, e.g. polling timer
27 and CMCI. All events reported as corrected are not cleared
28 by OS and remained in its error banks.
29 Usually this disablement is not recommended, however if
30 there is an agent checking/clearing corrected errors
31 (e.g. BIOS or hardware monitoring applications), conflicting
32 with OS's error handling, and you cannot deactivate the agent,
33 then this option will be a help.
35 Enable logging of machine checks left over from booting.
36 Disabled by default on AMD because some BIOS leave bogus ones.
37 If your BIOS doesn't do that it's a good idea to enable though
38 to make sure you log even machine check events that result
39 in a reboot. On Intel systems it is enabled by default.
41 Disable boot machine check logging.
42 mce=tolerancelevel[,monarchtimeout] (number,number)
44 0: always panic on uncorrected errors, log corrected errors
45 1: panic or SIGBUS on uncorrected errors, log corrected errors
46 2: SIGBUS or log uncorrected errors, log corrected errors
47 3: never panic or SIGBUS, log all errors (for testing only)
49 Can be also set using sysfs which is preferable.
51 Sets the time in us to wait for other CPUs on machine checks. 0
53 mce=bios_cmci_threshold
54 Don't overwrite the bios-set CMCI threshold. This boot option
55 prevents Linux from overwriting the CMCI threshold set by the
56 bios. Without this option, Linux always sets the CMCI
57 threshold to 1. Enabling this may make memory predictive failure
58 analysis less effective if the bios sets thresholds for memory
59 errors since we will not see details for all errors.
61 nomce (for compatibility with i386): same as mce=off
63 Everything else is in sysfs now.
67 apic Use IO-APIC. Default
69 noapic Don't use the IO-APIC.
71 disableapic Don't use the local APIC
73 nolapic Don't use the local APIC (alias for i386 compatibility)
75 pirq=... See Documentation/x86/i386/IO-APIC.txt
77 noapictimer Don't set up the APIC timer
79 no_timer_check Don't check the IO-APIC timer. This can work around
80 problems with incorrect timer initialization on some boards.
82 apicmaintimer Run time keeping from the local APIC timer instead
83 of using the PIT/HPET interrupt for this. This is useful
84 when the PIT/HPET interrupts are unreliable.
86 noapicmaintimer Don't do time keeping using the APIC timer.
87 Useful when this option was auto selected, but doesn't work.
90 Do APIC timer calibration using the pmtimer. Implies
91 apicmaintimer. Useful when your PIT timer is totally
96 syntax: earlyprintk=vga
97 earlyprintk=serial[,ttySn[,baudrate]]
99 The early console is useful when the kernel crashes before the
100 normal console is initialized. It is not enabled by
101 default because it has some cosmetic problems.
102 Append ,keep to not disable it when the real console takes over.
103 Only vga or serial at a time, not both.
104 Currently only ttyS0 and ttyS1 are supported.
105 Interaction with the standard serial driver is not very good.
106 The VGA output is eventually overwritten by the real console.
111 Don't use the CPU time stamp counter to read the wall time.
112 This can be used to work around timing problems on multiprocessor systems
113 with not properly synchronized CPUs.
116 Report when timer interrupts are lost because some code turned off
117 interrupts for too long.
120 Don't use the HPET timer.
125 Don't do power saving in the idle loop using HLT, but poll for rescheduling
126 event. This will make the CPUs eat a lot more power, but may be useful
127 to get slightly better performance in multiprocessor benchmarks. It also
128 makes some profiling using performance counters more accurate.
129 Please note that on systems with MONITOR/MWAIT support (like Intel EM64T
130 CPUs) this option has no performance advantage over the normal idle loop.
131 It may also interact badly with hyperthreading.
135 reboot=b[ios] | t[riple] | k[bd] | a[cpi] | e[fi] [, [w]arm | [c]old]
136 bios Use the CPU reboot vector for warm reset
137 warm Don't set the cold reboot flag
138 cold Set the cold reboot flag
139 triple Force a triple fault (init)
140 kbd Use the keyboard controller. cold reset (default)
141 acpi Use the ACPI RESET_REG in the FADT. If ACPI is not configured or the
142 ACPI reset does not work, the reboot path attempts the reset using
143 the keyboard controller.
144 efi Use efi reset_system runtime service. If EFI is not configured or the
145 EFI reset does not work, the reboot path attempts the reset using
146 the keyboard controller.
148 Using warm reset will be much faster especially on big memory
149 systems because the BIOS will not go through the memory check.
150 Disadvantage is that not all hardware will be completely reinitialized
151 on reboot so there may be boot problems on some systems.
155 Don't stop other CPUs on reboot. This can make reboot more reliable
158 Non Executable Mappings
167 additional_cpus=NUM Allow NUM more CPUs for hotplug
168 (defaults are specified by the BIOS, see Documentation/x86/x86_64/cpu-hotplug-spec)
172 numa=off Only set up a single NUMA node spanning all memory.
174 numa=noacpi Don't parse the SRAT table for NUMA setup
177 If given as a memory unit, fills all system RAM with nodes of
178 size interleaved over physical nodes.
181 If given as an integer, fills all system RAM with N fake nodes
182 interleaved over physical nodes.
186 acpi=off Don't enable ACPI
187 acpi=ht Use ACPI boot table parsing, but don't enable ACPI
189 acpi=force Force ACPI on (currently not needed)
191 acpi=strict Disable out of spec ACPI workarounds.
193 acpi_sci={edge,level,high,low} Set up ACPI SCI interrupt.
195 acpi=noirq Don't route interrupts
199 pci=off Don't use PCI
200 pci=conf1 Use conf1 access.
201 pci=conf2 Use conf2 access.
203 pci=assign-busses Assign busses
204 pci=irqmask=MASK Set PCI interrupt mask to MASK
205 pci=lastbus=NUMBER Scan up to NUMBER busses, no matter what the mptable says.
206 pci=noacpi Don't use ACPI to set up PCI interrupt routing.
208 IOMMU (input/output memory management unit)
210 Currently four x86-64 PCI-DMA mapping implementations exist:
212 1. <arch/x86_64/kernel/pci-nommu.c>: use no hardware/software IOMMU at all
213 (e.g. because you have < 3 GB memory).
214 Kernel boot message: "PCI-DMA: Disabling IOMMU"
216 2. <arch/x86/kernel/amd_gart_64.c>: AMD GART based hardware IOMMU.
217 Kernel boot message: "PCI-DMA: using GART IOMMU"
219 3. <arch/x86_64/kernel/pci-swiotlb.c> : Software IOMMU implementation. Used
220 e.g. if there is no hardware IOMMU in the system and it is need because
221 you have >3GB memory or told the kernel to us it (iommu=soft))
222 Kernel boot message: "PCI-DMA: Using software bounce buffering
225 4. <arch/x86_64/pci-calgary.c> : IBM Calgary hardware IOMMU. Used in IBM
226 pSeries and xSeries servers. This hardware IOMMU supports DMA address
227 mapping with memory protection, etc.
228 Kernel boot message: "PCI-DMA: Using Calgary IOMMU"
230 iommu=[<size>][,noagp][,off][,force][,noforce][,leak[=<nr_of_leak_pages>]
231 [,memaper[=<order>]][,merge][,forcesac][,fullflush][,nomerge]
232 [,noaperture][,calgary]
234 General iommu options:
235 off Don't initialize and use any kind of IOMMU.
236 noforce Don't force hardware IOMMU usage when it is not needed.
238 force Force the use of the hardware IOMMU even when it is
239 not actually needed (e.g. because < 3 GB memory).
240 soft Use software bounce buffering (SWIOTLB) (default for
241 Intel machines). This can be used to prevent the usage
242 of an available hardware IOMMU.
244 iommu options only relevant to the AMD GART hardware IOMMU:
245 <size> Set the size of the remapping area in bytes.
246 allowed Overwrite iommu off workarounds for specific chipsets.
247 fullflush Flush IOMMU on each allocation (default).
248 nofullflush Don't use IOMMU fullflush.
249 leak Turn on simple iommu leak tracing (only when
250 CONFIG_IOMMU_LEAK is on). Default number of leak pages
252 memaper[=<order>] Allocate an own aperture over RAM with size 32MB<<order.
253 (default: order=1, i.e. 64MB)
254 merge Do scatter-gather (SG) merging. Implies "force"
256 nomerge Don't do scatter-gather (SG) merging.
257 noaperture Ask the IOMMU not to touch the aperture for AGP.
258 forcesac Force single-address cycle (SAC) mode for masks <40bits
260 noagp Don't initialize the AGP driver and use full aperture.
261 allowdac Allow double-address cycle (DAC) mode, i.e. DMA >4GB.
262 DAC is used with 32-bit PCI to push a 64-bit address in
263 two cycles. When off all DMA over >4GB is forced through
264 an IOMMU or software bounce buffering.
265 nodac Forbid DAC mode, i.e. DMA >4GB.
266 panic Always panic when IOMMU overflows.
267 calgary Use the Calgary IOMMU if it is available
269 iommu options only relevant to the software bounce buffering (SWIOTLB) IOMMU
271 swiotlb=<pages>[,force]
272 <pages> Prereserve that many 128K pages for the software IO
274 force Force all IO through the software TLB.
276 Settings for the IBM Calgary hardware IOMMU currently found in IBM
277 pSeries and xSeries machines:
279 calgary=[64k,128k,256k,512k,1M,2M,4M,8M]
280 calgary=[translate_empty_slots]
281 calgary=[disable=<PCI bus number>]
282 panic Always panic when IOMMU overflows
284 64k,...,8M - Set the size of each PCI slot's translation table
285 when using the Calgary IOMMU. This is the size of the translation
286 table itself in main memory. The smallest table, 64k, covers an IO
287 space of 32MB; the largest, 8MB table, can cover an IO space of
288 4GB. Normally the kernel will make the right choice by itself.
290 translate_empty_slots - Enable translation even on slots that have
291 no devices attached to them, in case a device will be hotplugged
294 disable=<PCI bus number> - Disable translation on a given PHB. For
295 example, the built-in graphics adapter resides on the first bridge
296 (PCI bus number 0); if translation (isolation) is enabled on this
297 bridge, X servers that access the hardware directly from user
298 space might stop working. Use this option if you have devices that
299 are accessed from userspace directly on some PCI host bridge.
303 kstack=N Print N words from the kernel stack in oops dumps.
305 pagefaulttrace Dump all page faults. Only useful for extreme debugging
306 and will create a lot of output.
308 call_trace=[old|both|newfallback|new]
309 old: use old inexact backtracer
310 new: use new exact dwarf2 unwinder
311 both: print entries from both
312 newfallback: use new unwinder but fall back to old if it gets
318 Do not use GB pages for kernel direct mappings.
320 Use GB pages for kernel direct mappings.