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 mce=off disable machine check
9 mce=bootlog Enable logging of machine checks left over from booting.
10 Disabled by default on AMD because some BIOS leave bogus ones.
11 If your BIOS doesn't do that it's a good idea to enable though
12 to make sure you log even machine check events that result
13 in a reboot. On Intel systems it is enabled by default.
15 Disable boot machine check logging.
16 mce=tolerancelevel (number)
17 0: always panic, 1: panic if deadlock possible,
18 2: try to avoid panic, 3: never panic or exit (for testing)
20 Can be also set using sysfs which is preferable.
22 nomce (for compatibility with i386): same as mce=off
24 Everything else is in sysfs now.
28 apic Use IO-APIC. Default
30 noapic Don't use the IO-APIC.
32 disableapic Don't use the local APIC
34 nolapic Don't use the local APIC (alias for i386 compatibility)
36 pirq=... See Documentation/i386/IO-APIC.txt
38 noapictimer Don't set up the APIC timer
40 no_timer_check Don't check the IO-APIC timer. This can work around
41 problems with incorrect timer initialization on some boards.
43 apicmaintimer Run time keeping from the local APIC timer instead
44 of using the PIT/HPET interrupt for this. This is useful
45 when the PIT/HPET interrupts are unreliable.
47 noapicmaintimer Don't do time keeping using the APIC timer.
48 Useful when this option was auto selected, but doesn't work.
51 Do APIC timer calibration using the pmtimer. Implies
52 apicmaintimer. Useful when your PIT timer is totally
57 syntax: earlyprintk=vga
58 earlyprintk=serial[,ttySn[,baudrate]]
60 The early console is useful when the kernel crashes before the
61 normal console is initialized. It is not enabled by
62 default because it has some cosmetic problems.
63 Append ,keep to not disable it when the real console takes over.
64 Only vga or serial at a time, not both.
65 Currently only ttyS0 and ttyS1 are supported.
66 Interaction with the standard serial driver is not very good.
67 The VGA output is eventually overwritten by the real console.
72 Don't use the CPU time stamp counter to read the wall time.
73 This can be used to work around timing problems on multiprocessor systems
74 with not properly synchronized CPUs.
77 Report when timer interrupts are lost because some code turned off
78 interrupts for too long.
80 nmi_watchdog=NUMBER[,panic]
82 0 don't use an NMI watchdog
83 1 use the IO-APIC timer for the NMI watchdog
84 2 use the local APIC for the NMI watchdog using a performance counter. Note
85 This will use one performance counter and the local APIC's performance
87 When panic is specified panic when an NMI watchdog timeout occurs.
88 This is useful when you use a panic=... timeout and need the box
92 Don't use the HPET timer.
97 Don't do power saving in the idle loop using HLT, but poll for rescheduling
98 event. This will make the CPUs eat a lot more power, but may be useful
99 to get slightly better performance in multiprocessor benchmarks. It also
100 makes some profiling using performance counters more accurate.
101 Please note that on systems with MONITOR/MWAIT support (like Intel EM64T
102 CPUs) this option has no performance advantage over the normal idle loop.
103 It may also interact badly with hyperthreading.
107 reboot=b[ios] | t[riple] | k[bd] [, [w]arm | [c]old]
108 bios Use the CPU reboot vector for warm reset
109 warm Don't set the cold reboot flag
110 cold Set the cold reboot flag
111 triple Force a triple fault (init)
112 kbd Use the keyboard controller. cold reset (default)
114 Using warm reset will be much faster especially on big memory
115 systems because the BIOS will not go through the memory check.
116 Disadvantage is that not all hardware will be completely reinitialized
117 on reboot so there may be boot problems on some systems.
121 Don't stop other CPUs on reboot. This can make reboot more reliable
124 Non Executable Mappings
133 nosmp Only use a single CPU
135 maxcpus=NUMBER only use upto NUMBER CPUs
137 cpumask=MASK only use cpus with bits set in mask
139 additional_cpus=NUM Allow NUM more CPUs for hotplug
140 (defaults are specified by the BIOS, see Documentation/x86_64/cpu-hotplug-spec)
144 numa=off Only set up a single NUMA node spanning all memory.
146 numa=noacpi Don't parse the SRAT table for NUMA setup
148 numa=fake=X Fake X nodes and ignore NUMA setup of the actual machine.
151 Only allow hotadd memory to preallocate page structures upto
152 percent of already available memory.
153 numa=hotadd=0 will disable hotadd memory.
157 acpi=off Don't enable ACPI
158 acpi=ht Use ACPI boot table parsing, but don't enable ACPI
160 acpi=force Force ACPI on (currently not needed)
162 acpi=strict Disable out of spec ACPI workarounds.
164 acpi_sci={edge,level,high,low} Set up ACPI SCI interrupt.
166 acpi=noirq Don't route interrupts
170 pci=off Don't use PCI
171 pci=conf1 Use conf1 access.
172 pci=conf2 Use conf2 access.
174 pci=assign-busses Assign busses
175 pci=irqmask=MASK Set PCI interrupt mask to MASK
176 pci=lastbus=NUMBER Scan upto NUMBER busses, no matter what the mptable says.
177 pci=noacpi Don't use ACPI to set up PCI interrupt routing.
181 iommu=[size][,noagp][,off][,force][,noforce][,leak][,memaper[=order]][,merge]
182 [,forcesac][,fullflush][,nomerge][,noaperture][,calgary]
183 size set size of iommu (in bytes)
184 noagp don't initialize the AGP driver and use full aperture.
185 off don't use the IOMMU
186 leak turn on simple iommu leak tracing (only when CONFIG_IOMMU_LEAK is on)
187 memaper[=order] allocate an own aperture over RAM with size 32MB^order.
188 noforce don't force IOMMU usage. Default.
190 merge Do SG merging. Implies force (experimental)
191 nomerge Don't do SG merging.
192 forcesac For SAC mode for masks <40bits (experimental)
193 fullflush Flush IOMMU on each allocation (default)
194 nofullflush Don't use IOMMU fullflush
195 allowed overwrite iommu off workarounds for specific chipsets.
196 soft Use software bounce buffering (default for Intel machines)
197 noaperture Don't touch the aperture for AGP.
198 allowdac Allow DMA >4GB
199 When off all DMA over >4GB is forced through an IOMMU or bounce
201 nodac Forbid DMA >4GB
202 panic Always panic when IOMMU overflows
203 calgary Use the Calgary IOMMU if it is available
205 swiotlb=pages[,force]
207 pages Prereserve that many 128K pages for the software IO bounce buffering.
208 force Force all IO through the software TLB.
210 calgary=[64k,128k,256k,512k,1M,2M,4M,8M]
211 calgary=[translate_empty_slots]
212 calgary=[disable=<PCI bus number>]
214 64k,...,8M - Set the size of each PCI slot's translation table
215 when using the Calgary IOMMU. This is the size of the translation
216 table itself in main memory. The smallest table, 64k, covers an IO
217 space of 32MB; the largest, 8MB table, can cover an IO space of
218 4GB. Normally the kernel will make the right choice by itself.
220 translate_empty_slots - Enable translation even on slots that have
221 no devices attached to them, in case a device will be hotplugged
224 disable=<PCI bus number> - Disable translation on a given PHB. For
225 example, the built-in graphics adapter resides on the first bridge
226 (PCI bus number 0); if translation (isolation) is enabled on this
227 bridge, X servers that access the hardware directly from user
228 space might stop working. Use this option if you have devices that
229 are accessed from userspace directly on some PCI host bridge.
233 oops=panic Always panic on oopses. Default is to just kill the process,
234 but there is a small probability of deadlocking the machine.
235 This will also cause panics on machine check exceptions.
236 Useful together with panic=30 to trigger a reboot.
238 kstack=N Print that many words from the kernel stack in oops dumps.
240 pagefaulttrace Dump all page faults. Only useful for extreme debugging
241 and will create a lot of output.
243 call_trace=[old|both|newfallback|new]
244 old: use old inexact backtracer
245 new: use new exact dwarf2 unwinder
246 both: print entries from both
247 newfallback: use new unwinder but fall back to old if it gets
250 call_trace=[old|both|newfallback|new]
251 old: use old inexact backtracer
252 new: use new exact dwarf2 unwinder
253 both: print entries from both
254 newfallback: use new unwinder but fall back to old if it gets
259 noreplacement Don't replace instructions with more appropriate ones
260 for the CPU. This may be useful on asymmetric MP systems
261 where some CPU have less capabilities than the others.