Commit | Line | Data |
---|---|---|
1da177e4 LT |
1 | AMD64 specific boot options |
2 | ||
3 | There are many others (usually documented in driver documentation), but | |
4 | only the AMD64 specific ones are listed here. | |
5 | ||
6 | Machine check | |
7 | ||
8 | mce=off disable machine check | |
d5172f26 AK |
9 | mce=bootlog Enable logging of machine checks left over from booting. |
10 | Disabled by default 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. | |
8c566ef5 AK |
14 | mce=tolerancelevel (number) |
15 | 0: always panic, 1: panic if deadlock possible, | |
16 | 2: try to avoid panic, 3: never panic or exit (for testing) | |
17 | default is 1 | |
18 | Can be also set using sysfs which is preferable. | |
1da177e4 LT |
19 | |
20 | nomce (for compatibility with i386): same as mce=off | |
21 | ||
22 | Everything else is in sysfs now. | |
23 | ||
24 | APICs | |
25 | ||
26 | apic Use IO-APIC. Default | |
27 | ||
28 | noapic Don't use the IO-APIC. | |
29 | ||
30 | disableapic Don't use the local APIC | |
31 | ||
32 | nolapic Don't use the local APIC (alias for i386 compatibility) | |
33 | ||
34 | pirq=... See Documentation/i386/IO-APIC.txt | |
35 | ||
36 | noapictimer Don't set up the APIC timer | |
37 | ||
14d98cad AK |
38 | no_timer_check Don't check the IO-APIC timer. This can work around |
39 | problems with incorrect timer initialization on some boards. | |
40 | ||
1da177e4 LT |
41 | Early Console |
42 | ||
43 | syntax: earlyprintk=vga | |
44 | earlyprintk=serial[,ttySn[,baudrate]] | |
45 | ||
46 | The early console is useful when the kernel crashes before the | |
47 | normal console is initialized. It is not enabled by | |
48 | default because it has some cosmetic problems. | |
49 | Append ,keep to not disable it when the real console takes over. | |
50 | Only vga or serial at a time, not both. | |
51 | Currently only ttyS0 and ttyS1 are supported. | |
52 | Interaction with the standard serial driver is not very good. | |
53 | The VGA output is eventually overwritten by the real console. | |
54 | ||
55 | Timing | |
56 | ||
57 | notsc | |
58 | Don't use the CPU time stamp counter to read the wall time. | |
59 | This can be used to work around timing problems on multiprocessor systems | |
ef4d7cbe | 60 | with not properly synchronized CPUs. |
1da177e4 LT |
61 | |
62 | report_lost_ticks | |
63 | Report when timer interrupts are lost because some code turned off | |
64 | interrupts for too long. | |
65 | ||
66 | nmi_watchdog=NUMBER[,panic] | |
67 | NUMBER can be: | |
68 | 0 don't use an NMI watchdog | |
69 | 1 use the IO-APIC timer for the NMI watchdog | |
70 | 2 use the local APIC for the NMI watchdog using a performance counter. Note | |
71 | This will use one performance counter and the local APIC's performance | |
72 | vector. | |
73 | When panic is specified panic when an NMI watchdog timeout occurs. | |
74 | This is useful when you use a panic=... timeout and need the box | |
75 | quickly up again. | |
76 | ||
77 | nohpet | |
78 | Don't use the HPET timer. | |
79 | ||
80 | Idle loop | |
81 | ||
82 | idle=poll | |
83 | Don't do power saving in the idle loop using HLT, but poll for rescheduling | |
84 | event. This will make the CPUs eat a lot more power, but may be useful | |
85 | to get slightly better performance in multiprocessor benchmarks. It also | |
86 | makes some profiling using performance counters more accurate. | |
ef4d7cbe AK |
87 | Please note that on systems with MONITOR/MWAIT support (like Intel EM64T |
88 | CPUs) this option has no performance advantage over the normal idle loop. | |
89 | It may also interact badly with hyperthreading. | |
1da177e4 LT |
90 | |
91 | Rebooting | |
92 | ||
93 | reboot=b[ios] | t[riple] | k[bd] [, [w]arm | [c]old] | |
94 | bios Use the CPU reboto vector for warm reset | |
95 | warm Don't set the cold reboot flag | |
96 | cold Set the cold reboot flag | |
97 | triple Force a triple fault (init) | |
98 | kbd Use the keyboard controller. cold reset (default) | |
99 | ||
100 | Using warm reset will be much faster especially on big memory | |
101 | systems because the BIOS will not go through the memory check. | |
102 | Disadvantage is that not all hardware will be completely reinitialized | |
103 | on reboot so there may be boot problems on some systems. | |
104 | ||
105 | reboot=force | |
106 | ||
107 | Don't stop other CPUs on reboot. This can make reboot more reliable | |
108 | in some cases. | |
109 | ||
110 | Non Executable Mappings | |
111 | ||
112 | noexec=on|off | |
113 | ||
114 | on Enable(default) | |
115 | off Disable | |
116 | ||
117 | SMP | |
118 | ||
119 | nosmp Only use a single CPU | |
120 | ||
121 | maxcpus=NUMBER only use upto NUMBER CPUs | |
122 | ||
123 | cpumask=MASK only use cpus with bits set in mask | |
124 | ||
125 | NUMA | |
126 | ||
127 | numa=off Only set up a single NUMA node spanning all memory. | |
128 | ||
129 | numa=noacpi Don't parse the SRAT table for NUMA setup | |
130 | ||
131 | numa=fake=X Fake X nodes and ignore NUMA setup of the actual machine. | |
132 | ||
133 | ACPI | |
134 | ||
135 | acpi=off Don't enable ACPI | |
136 | acpi=ht Use ACPI boot table parsing, but don't enable ACPI | |
137 | interpreter | |
138 | acpi=force Force ACPI on (currently not needed) | |
139 | ||
140 | acpi=strict Disable out of spec ACPI workarounds. | |
141 | ||
142 | acpi_sci={edge,level,high,low} Set up ACPI SCI interrupt. | |
143 | ||
144 | acpi=noirq Don't route interrupts | |
145 | ||
146 | PCI | |
147 | ||
148 | pci=off Don't use PCI | |
149 | pci=conf1 Use conf1 access. | |
150 | pci=conf2 Use conf2 access. | |
151 | pci=rom Assign ROMs. | |
152 | pci=assign-busses Assign busses | |
153 | pci=irqmask=MASK Set PCI interrupt mask to MASK | |
154 | pci=lastbus=NUMBER Scan upto NUMBER busses, no matter what the mptable says. | |
155 | pci=noacpi Don't use ACPI to set up PCI interrupt routing. | |
156 | ||
157 | IOMMU | |
158 | ||
159 | iommu=[size][,noagp][,off][,force][,noforce][,leak][,memaper[=order]][,merge] | |
160 | [,forcesac][,fullflush][,nomerge][,noaperture] | |
161 | size set size of iommu (in bytes) | |
162 | noagp don't initialize the AGP driver and use full aperture. | |
163 | off don't use the IOMMU | |
164 | leak turn on simple iommu leak tracing (only when CONFIG_IOMMU_LEAK is on) | |
165 | memaper[=order] allocate an own aperture over RAM with size 32MB^order. | |
166 | noforce don't force IOMMU usage. Default. | |
167 | force Force IOMMU. | |
168 | merge Do SG merging. Implies force (experimental) | |
169 | nomerge Don't do SG merging. | |
170 | forcesac For SAC mode for masks <40bits (experimental) | |
171 | fullflush Flush IOMMU on each allocation (default) | |
172 | nofullflush Don't use IOMMU fullflush | |
173 | allowed overwrite iommu off workarounds for specific chipsets. | |
174 | soft Use software bounce buffering (default for Intel machines) | |
175 | noaperture Don't touch the aperture for AGP. | |
176 | ||
177 | swiotlb=pages[,force] | |
178 | ||
179 | pages Prereserve that many 128K pages for the software IO bounce buffering. | |
180 | force Force all IO through the software TLB. | |
181 | ||
182 | Debugging | |
183 | ||
184 | oops=panic Always panic on oopses. Default is to just kill the process, | |
185 | but there is a small probability of deadlocking the machine. | |
186 | This will also cause panics on machine check exceptions. | |
187 | Useful together with panic=30 to trigger a reboot. | |
188 | ||
189 | kstack=N Print that many words from the kernel stack in oops dumps. | |
190 | ||
191 | Misc | |
192 | ||
193 | noreplacement Don't replace instructions with more appropiate ones | |
ef4d7cbe AK |
194 | for the CPU. This may be useful on asymmetric MP systems |
195 | where some CPU have less capabilities than the others. |