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 | 9 | mce=bootlog Enable logging of machine checks left over from booting. |
e583538f | 10 | Disabled by default on AMD because some BIOS leave bogus ones. |
d5172f26 AK |
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 | |
e583538f AK |
13 | in a reboot. On Intel systems it is enabled by default. |
14 | mce=nobootlog | |
15 | Disable boot machine check logging. | |
8c566ef5 AK |
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) | |
19 | default is 1 | |
20 | Can be also set using sysfs which is preferable. | |
1da177e4 LT |
21 | |
22 | nomce (for compatibility with i386): same as mce=off | |
23 | ||
24 | Everything else is in sysfs now. | |
25 | ||
26 | APICs | |
27 | ||
28 | apic Use IO-APIC. Default | |
29 | ||
30 | noapic Don't use the IO-APIC. | |
31 | ||
32 | disableapic Don't use the local APIC | |
33 | ||
34 | nolapic Don't use the local APIC (alias for i386 compatibility) | |
35 | ||
36 | pirq=... See Documentation/i386/IO-APIC.txt | |
37 | ||
38 | noapictimer Don't set up the APIC timer | |
39 | ||
14d98cad AK |
40 | no_timer_check Don't check the IO-APIC timer. This can work around |
41 | problems with incorrect timer initialization on some boards. | |
42 | ||
73dea47f AK |
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. | |
46 | ||
47 | noapicmaintimer Don't do time keeping using the APIC timer. | |
48 | Useful when this option was auto selected, but doesn't work. | |
49 | ||
0c3749c4 AK |
50 | apicpmtimer |
51 | Do APIC timer calibration using the pmtimer. Implies | |
52 | apicmaintimer. Useful when your PIT timer is totally | |
53 | broken. | |
54 | ||
fea5f1e1 LT |
55 | disable_8254_timer / enable_8254_timer |
56 | Enable interrupt 0 timer routing over the 8254 in addition to over | |
57 | the IO-APIC. The kernel tries to set a sensible default. | |
58 | ||
1da177e4 LT |
59 | Early Console |
60 | ||
61 | syntax: earlyprintk=vga | |
62 | earlyprintk=serial[,ttySn[,baudrate]] | |
63 | ||
64 | The early console is useful when the kernel crashes before the | |
65 | normal console is initialized. It is not enabled by | |
66 | default because it has some cosmetic problems. | |
67 | Append ,keep to not disable it when the real console takes over. | |
68 | Only vga or serial at a time, not both. | |
69 | Currently only ttyS0 and ttyS1 are supported. | |
70 | Interaction with the standard serial driver is not very good. | |
71 | The VGA output is eventually overwritten by the real console. | |
72 | ||
73 | Timing | |
74 | ||
75 | notsc | |
76 | Don't use the CPU time stamp counter to read the wall time. | |
77 | This can be used to work around timing problems on multiprocessor systems | |
ef4d7cbe | 78 | with not properly synchronized CPUs. |
1da177e4 LT |
79 | |
80 | report_lost_ticks | |
81 | Report when timer interrupts are lost because some code turned off | |
82 | interrupts for too long. | |
83 | ||
84 | nmi_watchdog=NUMBER[,panic] | |
85 | NUMBER can be: | |
86 | 0 don't use an NMI watchdog | |
87 | 1 use the IO-APIC timer for the NMI watchdog | |
88 | 2 use the local APIC for the NMI watchdog using a performance counter. Note | |
89 | This will use one performance counter and the local APIC's performance | |
90 | vector. | |
91 | When panic is specified panic when an NMI watchdog timeout occurs. | |
92 | This is useful when you use a panic=... timeout and need the box | |
93 | quickly up again. | |
94 | ||
95 | nohpet | |
96 | Don't use the HPET timer. | |
97 | ||
98 | Idle loop | |
99 | ||
100 | idle=poll | |
101 | Don't do power saving in the idle loop using HLT, but poll for rescheduling | |
102 | event. This will make the CPUs eat a lot more power, but may be useful | |
103 | to get slightly better performance in multiprocessor benchmarks. It also | |
104 | makes some profiling using performance counters more accurate. | |
ef4d7cbe AK |
105 | Please note that on systems with MONITOR/MWAIT support (like Intel EM64T |
106 | CPUs) this option has no performance advantage over the normal idle loop. | |
107 | It may also interact badly with hyperthreading. | |
1da177e4 LT |
108 | |
109 | Rebooting | |
110 | ||
111 | reboot=b[ios] | t[riple] | k[bd] [, [w]arm | [c]old] | |
d6bc8ac9 | 112 | bios Use the CPU reboot vector for warm reset |
1da177e4 LT |
113 | warm Don't set the cold reboot flag |
114 | cold Set the cold reboot flag | |
115 | triple Force a triple fault (init) | |
116 | kbd Use the keyboard controller. cold reset (default) | |
117 | ||
118 | Using warm reset will be much faster especially on big memory | |
119 | systems because the BIOS will not go through the memory check. | |
120 | Disadvantage is that not all hardware will be completely reinitialized | |
121 | on reboot so there may be boot problems on some systems. | |
122 | ||
123 | reboot=force | |
124 | ||
125 | Don't stop other CPUs on reboot. This can make reboot more reliable | |
126 | in some cases. | |
127 | ||
128 | Non Executable Mappings | |
129 | ||
130 | noexec=on|off | |
131 | ||
132 | on Enable(default) | |
133 | off Disable | |
134 | ||
135 | SMP | |
136 | ||
137 | nosmp Only use a single CPU | |
138 | ||
139 | maxcpus=NUMBER only use upto NUMBER CPUs | |
140 | ||
141 | cpumask=MASK only use cpus with bits set in mask | |
142 | ||
420f8f68 | 143 | additional_cpus=NUM Allow NUM more CPUs for hotplug |
f62a91f6 | 144 | (defaults are specified by the BIOS, see Documentation/x86_64/cpu-hotplug-spec) |
420f8f68 | 145 | |
1da177e4 LT |
146 | NUMA |
147 | ||
148 | numa=off Only set up a single NUMA node spanning all memory. | |
149 | ||
150 | numa=noacpi Don't parse the SRAT table for NUMA setup | |
151 | ||
8b8ca80e DR |
152 | numa=fake=CMDLINE |
153 | If a number, fakes CMDLINE nodes and ignores NUMA setup of the | |
154 | actual machine. Otherwise, system memory is configured | |
155 | depending on the sizes and coefficients listed. For example: | |
156 | numa=fake=2*512,1024,4*256 | |
157 | gives two 512M nodes, a 1024M node, and four 256M nodes. The | |
14694d73 DR |
158 | remaining system RAM is allocated to an additional node. If |
159 | the last character of CMDLINE is a *, the remaining system RAM | |
160 | is instead divided up equally among its coefficient. | |
1da177e4 | 161 | |
68a3a7fe AK |
162 | numa=hotadd=percent |
163 | Only allow hotadd memory to preallocate page structures upto | |
164 | percent of already available memory. | |
165 | numa=hotadd=0 will disable hotadd memory. | |
166 | ||
1da177e4 LT |
167 | ACPI |
168 | ||
169 | acpi=off Don't enable ACPI | |
170 | acpi=ht Use ACPI boot table parsing, but don't enable ACPI | |
171 | interpreter | |
172 | acpi=force Force ACPI on (currently not needed) | |
173 | ||
174 | acpi=strict Disable out of spec ACPI workarounds. | |
175 | ||
176 | acpi_sci={edge,level,high,low} Set up ACPI SCI interrupt. | |
177 | ||
178 | acpi=noirq Don't route interrupts | |
179 | ||
180 | PCI | |
181 | ||
182 | pci=off Don't use PCI | |
183 | pci=conf1 Use conf1 access. | |
184 | pci=conf2 Use conf2 access. | |
185 | pci=rom Assign ROMs. | |
186 | pci=assign-busses Assign busses | |
187 | pci=irqmask=MASK Set PCI interrupt mask to MASK | |
188 | pci=lastbus=NUMBER Scan upto NUMBER busses, no matter what the mptable says. | |
189 | pci=noacpi Don't use ACPI to set up PCI interrupt routing. | |
190 | ||
5558870b KW |
191 | IOMMU (input/output memory management unit) |
192 | ||
193 | Currently four x86-64 PCI-DMA mapping implementations exist: | |
194 | ||
195 | 1. <arch/x86_64/kernel/pci-nommu.c>: use no hardware/software IOMMU at all | |
196 | (e.g. because you have < 3 GB memory). | |
197 | Kernel boot message: "PCI-DMA: Disabling IOMMU" | |
198 | ||
199 | 2. <arch/x86_64/kernel/pci-gart.c>: AMD GART based hardware IOMMU. | |
200 | Kernel boot message: "PCI-DMA: using GART IOMMU" | |
201 | ||
202 | 3. <arch/x86_64/kernel/pci-swiotlb.c> : Software IOMMU implementation. Used | |
203 | e.g. if there is no hardware IOMMU in the system and it is need because | |
204 | you have >3GB memory or told the kernel to us it (iommu=soft)) | |
205 | Kernel boot message: "PCI-DMA: Using software bounce buffering | |
206 | for IO (SWIOTLB)" | |
207 | ||
208 | 4. <arch/x86_64/pci-calgary.c> : IBM Calgary hardware IOMMU. Used in IBM | |
209 | pSeries and xSeries servers. This hardware IOMMU supports DMA address | |
210 | mapping with memory protection, etc. | |
211 | Kernel boot message: "PCI-DMA: Using Calgary IOMMU" | |
212 | ||
213 | iommu=[<size>][,noagp][,off][,force][,noforce][,leak[=<nr_of_leak_pages>] | |
214 | [,memaper[=<order>]][,merge][,forcesac][,fullflush][,nomerge] | |
215 | [,noaperture][,calgary] | |
216 | ||
217 | General iommu options: | |
218 | off Don't initialize and use any kind of IOMMU. | |
219 | noforce Don't force hardware IOMMU usage when it is not needed. | |
220 | (default). | |
221 | force Force the use of the hardware IOMMU even when it is | |
222 | not actually needed (e.g. because < 3 GB memory). | |
223 | soft Use software bounce buffering (SWIOTLB) (default for | |
224 | Intel machines). This can be used to prevent the usage | |
225 | of an available hardware IOMMU. | |
226 | ||
227 | iommu options only relevant to the AMD GART hardware IOMMU: | |
228 | <size> Set the size of the remapping area in bytes. | |
229 | allowed Overwrite iommu off workarounds for specific chipsets. | |
230 | fullflush Flush IOMMU on each allocation (default). | |
231 | nofullflush Don't use IOMMU fullflush. | |
232 | leak Turn on simple iommu leak tracing (only when | |
233 | CONFIG_IOMMU_LEAK is on). Default number of leak pages | |
234 | is 20. | |
235 | memaper[=<order>] Allocate an own aperture over RAM with size 32MB<<order. | |
236 | (default: order=1, i.e. 64MB) | |
57d30772 | 237 | merge Do scatter-gather (SG) merging. Implies "force" |
5558870b | 238 | (experimental). |
57d30772 | 239 | nomerge Don't do scatter-gather (SG) merging. |
5558870b KW |
240 | noaperture Ask the IOMMU not to touch the aperture for AGP. |
241 | forcesac Force single-address cycle (SAC) mode for masks <40bits | |
242 | (experimental). | |
243 | noagp Don't initialize the AGP driver and use full aperture. | |
244 | allowdac Allow double-address cycle (DAC) mode, i.e. DMA >4GB. | |
245 | DAC is used with 32-bit PCI to push a 64-bit address in | |
246 | two cycles. When off all DMA over >4GB is forced through | |
247 | an IOMMU or software bounce buffering. | |
248 | nodac Forbid DAC mode, i.e. DMA >4GB. | |
249 | panic Always panic when IOMMU overflows. | |
250 | calgary Use the Calgary IOMMU if it is available | |
251 | ||
252 | iommu options only relevant to the software bounce buffering (SWIOTLB) IOMMU | |
253 | implementation: | |
254 | swiotlb=<pages>[,force] | |
255 | <pages> Prereserve that many 128K pages for the software IO | |
256 | bounce buffering. | |
257 | force Force all IO through the software TLB. | |
258 | ||
259 | Settings for the IBM Calgary hardware IOMMU currently found in IBM | |
260 | pSeries and xSeries machines: | |
261 | ||
262 | calgary=[64k,128k,256k,512k,1M,2M,4M,8M] | |
263 | calgary=[translate_empty_slots] | |
264 | calgary=[disable=<PCI bus number>] | |
265 | panic Always panic when IOMMU overflows | |
e465058d JM |
266 | |
267 | 64k,...,8M - Set the size of each PCI slot's translation table | |
268 | when using the Calgary IOMMU. This is the size of the translation | |
269 | table itself in main memory. The smallest table, 64k, covers an IO | |
270 | space of 32MB; the largest, 8MB table, can cover an IO space of | |
271 | 4GB. Normally the kernel will make the right choice by itself. | |
272 | ||
273 | translate_empty_slots - Enable translation even on slots that have | |
274 | no devices attached to them, in case a device will be hotplugged | |
275 | in the future. | |
276 | ||
277 | disable=<PCI bus number> - Disable translation on a given PHB. For | |
278 | example, the built-in graphics adapter resides on the first bridge | |
279 | (PCI bus number 0); if translation (isolation) is enabled on this | |
280 | bridge, X servers that access the hardware directly from user | |
281 | space might stop working. Use this option if you have devices that | |
282 | are accessed from userspace directly on some PCI host bridge. | |
283 | ||
1da177e4 LT |
284 | Debugging |
285 | ||
57d30772 RD |
286 | oops=panic Always panic on oopses. Default is to just kill the process, |
287 | but there is a small probability of deadlocking the machine. | |
288 | This will also cause panics on machine check exceptions. | |
289 | Useful together with panic=30 to trigger a reboot. | |
1da177e4 | 290 | |
57d30772 | 291 | kstack=N Print N words from the kernel stack in oops dumps. |
1da177e4 | 292 | |
57d30772 | 293 | pagefaulttrace Dump all page faults. Only useful for extreme debugging |
9e43e1b7 AK |
294 | and will create a lot of output. |
295 | ||
b783fd92 AK |
296 | call_trace=[old|both|newfallback|new] |
297 | old: use old inexact backtracer | |
298 | new: use new exact dwarf2 unwinder | |
299 | both: print entries from both | |
300 | newfallback: use new unwinder but fall back to old if it gets | |
301 | stuck (default) | |
302 | ||
57d30772 | 303 | Miscellaneous |