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. | |
1da177e4 LT |
14 | |
15 | nomce (for compatibility with i386): same as mce=off | |
16 | ||
17 | Everything else is in sysfs now. | |
18 | ||
19 | APICs | |
20 | ||
21 | apic Use IO-APIC. Default | |
22 | ||
23 | noapic Don't use the IO-APIC. | |
24 | ||
25 | disableapic Don't use the local APIC | |
26 | ||
27 | nolapic Don't use the local APIC (alias for i386 compatibility) | |
28 | ||
29 | pirq=... See Documentation/i386/IO-APIC.txt | |
30 | ||
31 | noapictimer Don't set up the APIC timer | |
32 | ||
14d98cad AK |
33 | no_timer_check Don't check the IO-APIC timer. This can work around |
34 | problems with incorrect timer initialization on some boards. | |
35 | ||
1da177e4 LT |
36 | Early Console |
37 | ||
38 | syntax: earlyprintk=vga | |
39 | earlyprintk=serial[,ttySn[,baudrate]] | |
40 | ||
41 | The early console is useful when the kernel crashes before the | |
42 | normal console is initialized. It is not enabled by | |
43 | default because it has some cosmetic problems. | |
44 | Append ,keep to not disable it when the real console takes over. | |
45 | Only vga or serial at a time, not both. | |
46 | Currently only ttyS0 and ttyS1 are supported. | |
47 | Interaction with the standard serial driver is not very good. | |
48 | The VGA output is eventually overwritten by the real console. | |
49 | ||
50 | Timing | |
51 | ||
52 | notsc | |
53 | Don't use the CPU time stamp counter to read the wall time. | |
54 | This can be used to work around timing problems on multiprocessor systems | |
ef4d7cbe | 55 | with not properly synchronized CPUs. |
1da177e4 LT |
56 | |
57 | report_lost_ticks | |
58 | Report when timer interrupts are lost because some code turned off | |
59 | interrupts for too long. | |
60 | ||
61 | nmi_watchdog=NUMBER[,panic] | |
62 | NUMBER can be: | |
63 | 0 don't use an NMI watchdog | |
64 | 1 use the IO-APIC timer for the NMI watchdog | |
65 | 2 use the local APIC for the NMI watchdog using a performance counter. Note | |
66 | This will use one performance counter and the local APIC's performance | |
67 | vector. | |
68 | When panic is specified panic when an NMI watchdog timeout occurs. | |
69 | This is useful when you use a panic=... timeout and need the box | |
70 | quickly up again. | |
71 | ||
72 | nohpet | |
73 | Don't use the HPET timer. | |
74 | ||
75 | Idle loop | |
76 | ||
77 | idle=poll | |
78 | Don't do power saving in the idle loop using HLT, but poll for rescheduling | |
79 | event. This will make the CPUs eat a lot more power, but may be useful | |
80 | to get slightly better performance in multiprocessor benchmarks. It also | |
81 | makes some profiling using performance counters more accurate. | |
ef4d7cbe AK |
82 | Please note that on systems with MONITOR/MWAIT support (like Intel EM64T |
83 | CPUs) this option has no performance advantage over the normal idle loop. | |
84 | It may also interact badly with hyperthreading. | |
1da177e4 LT |
85 | |
86 | Rebooting | |
87 | ||
88 | reboot=b[ios] | t[riple] | k[bd] [, [w]arm | [c]old] | |
89 | bios Use the CPU reboto vector for warm reset | |
90 | warm Don't set the cold reboot flag | |
91 | cold Set the cold reboot flag | |
92 | triple Force a triple fault (init) | |
93 | kbd Use the keyboard controller. cold reset (default) | |
94 | ||
95 | Using warm reset will be much faster especially on big memory | |
96 | systems because the BIOS will not go through the memory check. | |
97 | Disadvantage is that not all hardware will be completely reinitialized | |
98 | on reboot so there may be boot problems on some systems. | |
99 | ||
100 | reboot=force | |
101 | ||
102 | Don't stop other CPUs on reboot. This can make reboot more reliable | |
103 | in some cases. | |
104 | ||
105 | Non Executable Mappings | |
106 | ||
107 | noexec=on|off | |
108 | ||
109 | on Enable(default) | |
110 | off Disable | |
111 | ||
112 | SMP | |
113 | ||
114 | nosmp Only use a single CPU | |
115 | ||
116 | maxcpus=NUMBER only use upto NUMBER CPUs | |
117 | ||
118 | cpumask=MASK only use cpus with bits set in mask | |
119 | ||
120 | NUMA | |
121 | ||
122 | numa=off Only set up a single NUMA node spanning all memory. | |
123 | ||
124 | numa=noacpi Don't parse the SRAT table for NUMA setup | |
125 | ||
126 | numa=fake=X Fake X nodes and ignore NUMA setup of the actual machine. | |
127 | ||
128 | ACPI | |
129 | ||
130 | acpi=off Don't enable ACPI | |
131 | acpi=ht Use ACPI boot table parsing, but don't enable ACPI | |
132 | interpreter | |
133 | acpi=force Force ACPI on (currently not needed) | |
134 | ||
135 | acpi=strict Disable out of spec ACPI workarounds. | |
136 | ||
137 | acpi_sci={edge,level,high,low} Set up ACPI SCI interrupt. | |
138 | ||
139 | acpi=noirq Don't route interrupts | |
140 | ||
141 | PCI | |
142 | ||
143 | pci=off Don't use PCI | |
144 | pci=conf1 Use conf1 access. | |
145 | pci=conf2 Use conf2 access. | |
146 | pci=rom Assign ROMs. | |
147 | pci=assign-busses Assign busses | |
148 | pci=irqmask=MASK Set PCI interrupt mask to MASK | |
149 | pci=lastbus=NUMBER Scan upto NUMBER busses, no matter what the mptable says. | |
150 | pci=noacpi Don't use ACPI to set up PCI interrupt routing. | |
151 | ||
152 | IOMMU | |
153 | ||
154 | iommu=[size][,noagp][,off][,force][,noforce][,leak][,memaper[=order]][,merge] | |
155 | [,forcesac][,fullflush][,nomerge][,noaperture] | |
156 | size set size of iommu (in bytes) | |
157 | noagp don't initialize the AGP driver and use full aperture. | |
158 | off don't use the IOMMU | |
159 | leak turn on simple iommu leak tracing (only when CONFIG_IOMMU_LEAK is on) | |
160 | memaper[=order] allocate an own aperture over RAM with size 32MB^order. | |
161 | noforce don't force IOMMU usage. Default. | |
162 | force Force IOMMU. | |
163 | merge Do SG merging. Implies force (experimental) | |
164 | nomerge Don't do SG merging. | |
165 | forcesac For SAC mode for masks <40bits (experimental) | |
166 | fullflush Flush IOMMU on each allocation (default) | |
167 | nofullflush Don't use IOMMU fullflush | |
168 | allowed overwrite iommu off workarounds for specific chipsets. | |
169 | soft Use software bounce buffering (default for Intel machines) | |
170 | noaperture Don't touch the aperture for AGP. | |
171 | ||
172 | swiotlb=pages[,force] | |
173 | ||
174 | pages Prereserve that many 128K pages for the software IO bounce buffering. | |
175 | force Force all IO through the software TLB. | |
176 | ||
177 | Debugging | |
178 | ||
179 | oops=panic Always panic on oopses. Default is to just kill the process, | |
180 | but there is a small probability of deadlocking the machine. | |
181 | This will also cause panics on machine check exceptions. | |
182 | Useful together with panic=30 to trigger a reboot. | |
183 | ||
184 | kstack=N Print that many words from the kernel stack in oops dumps. | |
185 | ||
186 | Misc | |
187 | ||
188 | noreplacement Don't replace instructions with more appropiate ones | |
ef4d7cbe AK |
189 | for the CPU. This may be useful on asymmetric MP systems |
190 | where some CPU have less capabilities than the others. |