1 HXCOMM Use
DEFHEADING() to define headings
in both help text and rST
.
2 HXCOMM Text between SRST and ERST is copied to the rST version and
3 HXCOMM discarded from C version
.
4 HXCOMM
DEF(option
, HAS_ARG
/0, opt_enum
, opt_help
, arch_mask
) is used to
5 HXCOMM construct option structures
, enums and help message
for specified
7 HXCOMM HXCOMM can be used
for comments
, discarded from both rST and C
.
9 DEFHEADING(Standard options
:)
11 DEF("help", 0, QEMU_OPTION_h
,
12 "-h or -help display this help and exit\n", QEMU_ARCH_ALL
)
18 DEF("version", 0, QEMU_OPTION_version
,
19 "-version display version information and exit\n", QEMU_ARCH_ALL
)
22 Display version information and exit
25 DEF("machine", HAS_ARG
, QEMU_OPTION_machine
, \
26 "-machine [type=]name[,prop[=value][,...]]\n"
27 " selects emulated machine ('-machine help' for list)\n"
28 " property accel=accel1[:accel2[:...]] selects accelerator\n"
29 " supported accelerators are kvm, xen, hax, hvf, nvmm, whpx or tcg (default: tcg)\n"
30 " vmport=on|off|auto controls emulation of vmport (default: auto)\n"
31 " dump-guest-core=on|off include guest memory in a core dump (default=on)\n"
32 " mem-merge=on|off controls memory merge support (default: on)\n"
33 " aes-key-wrap=on|off controls support for AES key wrapping (default=on)\n"
34 " dea-key-wrap=on|off controls support for DEA key wrapping (default=on)\n"
35 " suppress-vmdesc=on|off disables self-describing migration (default=off)\n"
36 " nvdimm=on|off controls NVDIMM support (default=off)\n"
37 " memory-encryption=@var{} memory encryption object to use (default=none)\n"
38 " hmat=on|off controls ACPI HMAT support (default=off)\n"
39 " memory-backend='backend-id' specifies explicitly provided backend for main RAM (default=none)\n",
42 ``
-machine
[type
=]name
[,prop
=value
[,...]]``
43 Select the emulated machine by name
. Use ``
-machine help`` to list
46 For architectures which aim to support live migration compatibility
47 across releases
, each release will introduce a
new versioned machine
48 type
. For example
, the
2.8.0 release introduced machine types
49 "pc-i440fx-2.8" and
"pc-q35-2.8" for the x86\_64
/i686 architectures
.
51 To allow live migration of guests from QEMU version
2.8.0, to QEMU
52 version
2.9.0, the
2.9.0 version must support the
"pc-i440fx-2.8"
53 and
"pc-q35-2.8" machines too
. To allow users live migrating VMs to
54 skip multiple intermediate releases when upgrading
, new releases of
55 QEMU will support machine types from many previous versions
.
57 Supported machine properties are
:
59 ``accel
=accels1
[:accels2
[:...]]``
60 This is used to enable an accelerator
. Depending on the target
61 architecture
, kvm
, xen
, hax
, hvf
, nvmm
, whpx or tcg can be available
.
62 By
default, tcg is used
. If there is more than one accelerator
63 specified
, the next one is used
if the previous one fails to
66 ``vmport
=on|off|auto``
67 Enables emulation of VMWare IO port
, for vmmouse etc
. auto says
68 to select the value based on accel
. For accel
=xen the
default is
69 off otherwise the
default is on
.
71 ``dump
-guest
-core
=on|off``
72 Include guest memory
in a core dump
. The
default is on
.
75 Enables or disables memory merge support
. This feature
, when
76 supported by the host
, de
-duplicates identical memory pages
77 among VMs
instances (enabled by
default).
79 ``aes
-key
-wrap
=on|off``
80 Enables or disables AES key wrapping support on s390
-ccw hosts
.
81 This feature controls whether AES wrapping keys will be created
82 to allow execution of AES cryptographic functions
. The
default
85 ``dea
-key
-wrap
=on|off``
86 Enables or disables DEA key wrapping support on s390
-ccw hosts
.
87 This feature controls whether DEA wrapping keys will be created
88 to allow execution of DEA cryptographic functions
. The
default
92 Enables or disables NVDIMM support
. The
default is off
.
94 ``memory
-encryption
=``
95 Memory encryption object to use
. The
default is none
.
98 Enables or disables ACPI Heterogeneous Memory Attribute Table
99 (HMAT
) support
. The
default is off
.
101 ``memory
-backend
='id'``
102 An alternative to legacy ``
-mem
-path`` and ``mem
-prealloc`` options
.
103 Allows to use a memory backend as main RAM
.
108 -object memory
-backend
-file
,id
=pc
.ram
,size
=512M
,mem
-path
=/hugetlbfs
,prealloc
=on
,share
=on
109 -machine memory
-backend
=pc
.ram
112 Migration compatibility note
:
114 * as backend id one shall use value of
'default-ram-id', advertised by
115 machine
type (available via ``query
-machines`` QMP command
), if migration
116 to
/from old
QEMU (<5.0) is expected
.
117 * for machine types
4.0 and older
, user shall
118 use ``x
-use
-canonical
-path
-for-ramblock
-id
=off`` backend option
119 if migration to
/from old
QEMU (<5.0) is expected
.
124 -object memory
-backend
-ram
,id
=pc
.ram
,size
=512M
,x
-use
-canonical
-path
-for-ramblock
-id
=off
125 -machine memory
-backend
=pc
.ram
129 DEF("M", HAS_ARG
, QEMU_OPTION_M
,
130 " sgx-epc.0.memdev=memid,sgx-epc.0.node=numaid\n",
134 ``sgx
-epc
.0.memdev
=@
var{memid
},sgx
-epc
.0.node
=@
var{numaid
}``
135 Define an SGX EPC section
.
138 DEF("cpu", HAS_ARG
, QEMU_OPTION_cpu
,
139 "-cpu cpu select CPU ('-cpu help' for list)\n", QEMU_ARCH_ALL
)
142 Select CPU
model (``
-cpu help``
for list and additional feature
146 DEF("accel", HAS_ARG
, QEMU_OPTION_accel
,
147 "-accel [accel=]accelerator[,prop[=value][,...]]\n"
148 " select accelerator (kvm, xen, hax, hvf, nvmm, whpx or tcg; use 'help' for a list)\n"
149 " igd-passthru=on|off (enable Xen integrated Intel graphics passthrough, default=off)\n"
150 " kernel-irqchip=on|off|split controls accelerated irqchip support (default=on)\n"
151 " kvm-shadow-mem=size of KVM shadow MMU in bytes\n"
152 " split-wx=on|off (enable TCG split w^x mapping)\n"
153 " tb-size=n (TCG translation block cache size)\n"
154 " dirty-ring-size=n (KVM dirty ring GFN count, default 0)\n"
155 " thread=single|multi (enable multi-threaded TCG)\n", QEMU_ARCH_ALL
)
157 ``
-accel name
[,prop
=value
[,...]]``
158 This is used to enable an accelerator
. Depending on the target
159 architecture
, kvm
, xen
, hax
, hvf
, nvmm
, whpx or tcg can be available
. By
160 default, tcg is used
. If there is more than one accelerator
161 specified
, the next one is used
if the previous one fails to
164 ``igd
-passthru
=on|off``
165 When Xen is
in use
, this option controls whether Intel
166 integrated graphics devices can be passed through to the guest
169 ``kernel
-irqchip
=on|off|split``
170 Controls KVM
in-kernel irqchip support
. The
default is full
171 acceleration of the interrupt controllers
. On x86
, split irqchip
172 reduces the kernel attack surface
, at a performance cost
for
173 non
-MSI interrupts
. Disabling the
in-kernel irqchip completely
174 is not recommended except
for debugging purposes
.
176 ``kvm
-shadow
-mem
=size``
177 Defines the size of the KVM shadow MMU
.
180 Controls the use of split w^x mapping
for the TCG code generation
181 buffer
. Some operating systems require
this to be enabled
, and
in
182 such a
case this will
default on
. On other operating systems
, this
183 will
default off
, but one may enable
this for testing or debugging
.
186 Controls the
size (in MiB
) of the TCG translation block cache
.
188 ``thread
=single|multi``
189 Controls number of TCG threads
. When the TCG is multi
-threaded
190 there will be one thread per vCPU therefore taking advantage of
191 additional host cores
. The
default is to enable multi
-threading
192 where both the back
-end and front
-ends support it and no
193 incompatible TCG features have been
enabled (e
.g
.
196 ``dirty
-ring
-size
=n``
197 When the KVM accelerator is used
, it controls the size of the per
-vCPU
198 dirty page ring
buffer (number of entries
for each vCPU
). It should
199 be a value that is power of two
, and it should be
1024 or
bigger (but
200 still less than the maximum value that the kernel supports
). 4096
201 could be a good initial value
if you have no idea which is the best
.
202 Set
this value to
0 to disable the feature
. By
default, this feature
203 is
disabled (dirty
-ring
-size
=0). When enabled
, KVM will instead
204 record dirty pages
in a bitmap
.
208 DEF("smp", HAS_ARG
, QEMU_OPTION_smp
,
209 "-smp [[cpus=]n][,maxcpus=maxcpus][,sockets=sockets][,dies=dies][,clusters=clusters][,cores=cores][,threads=threads]\n"
210 " set the number of initial CPUs to 'n' [default=1]\n"
211 " maxcpus= maximum number of total CPUs, including\n"
212 " offline CPUs for hotplug, etc\n"
213 " sockets= number of sockets on the machine board\n"
214 " dies= number of dies in one socket\n"
215 " clusters= number of clusters in one die\n"
216 " cores= number of cores in one cluster\n"
217 " threads= number of threads in one core\n"
218 "Note: Different machines may have different subsets of the CPU topology\n"
219 " parameters supported, so the actual meaning of the supported parameters\n"
220 " will vary accordingly. For example, for a machine type that supports a\n"
221 " three-level CPU hierarchy of sockets/cores/threads, the parameters will\n"
222 " sequentially mean as below:\n"
223 " sockets means the number of sockets on the machine board\n"
224 " cores means the number of cores in one socket\n"
225 " threads means the number of threads in one core\n"
226 " For a particular machine type board, an expected CPU topology hierarchy\n"
227 " can be defined through the supported sub-option. Unsupported parameters\n"
228 " can also be provided in addition to the sub-option, but their values\n"
229 " must be set as 1 in the purpose of correct parsing.\n",
232 ``
-smp
[[cpus
=]n
][,maxcpus
=maxcpus
][,sockets
=sockets
][,dies
=dies
][,clusters
=clusters
][,cores
=cores
][,threads
=threads
]``
233 Simulate a SMP system with
'\ ``n``\ ' CPUs initially present on
234 the machine type board
. On boards supporting CPU hotplug
, the optional
235 '\ ``maxcpus``\ ' parameter can be set to enable further CPUs to be
236 added at runtime
. When both parameters are omitted
, the maximum number
237 of CPUs will be calculated from the provided topology members and the
238 initial CPU count will match the maximum number
. When only one of them
239 is given then the omitted one will be set to its counterpart
's value.
240 Both parameters may be specified, but the maximum number of CPUs must
241 be equal to or greater than the initial CPU count. Product of the
242 CPU topology hierarchy must be equal to the maximum number of CPUs.
243 Both parameters are subject to an upper limit that is determined by
244 the specific machine type chosen.
246 To control reporting of CPU topology information, values of the topology
247 parameters can be specified. Machines may only support a subset of the
248 parameters and different machines may have different subsets supported
249 which vary depending on capacity of the corresponding CPU targets. So
250 for a particular machine type board, an expected topology hierarchy can
251 be defined through the supported sub-option. Unsupported parameters can
252 also be provided in addition to the sub-option, but their values must be
253 set as 1 in the purpose of correct parsing.
255 Either the initial CPU count, or at least one of the topology parameters
256 must be specified. The specified parameters must be greater than zero,
257 explicit configuration like "cpus=0" is not allowed. Values for any
258 omitted parameters will be computed from those which are given.
260 For example, the following sub-option defines a CPU topology hierarchy
261 (2 sockets totally on the machine, 2 cores per socket, 2 threads per
262 core) for a machine that only supports sockets/cores/threads.
263 Some members of the option can be omitted but their values will be
264 automatically computed:
268 -smp 8,sockets=2,cores=2,threads=2,maxcpus=8
270 The following sub-option defines a CPU topology hierarchy (2 sockets
271 totally on the machine, 2 dies per socket, 2 cores per die, 2 threads
272 per core) for PC machines which support sockets/dies/cores/threads.
273 Some members of the option can be omitted but their values will be
274 automatically computed:
278 -smp 16,sockets=2,dies=2,cores=2,threads=2,maxcpus=16
280 Historically preference was given to the coarsest topology parameters
281 when computing missing values (ie sockets preferred over cores, which
282 were preferred over threads), however, this behaviour is considered
283 liable to change. Prior to 6.2 the preference was sockets over cores
284 over threads. Since 6.2 the preference is cores over sockets over threads.
286 For example, the following option defines a machine board with 2 sockets
287 of 1 core before 6.2 and 1 socket of 2 cores after 6.2:
294 DEF("numa", HAS_ARG, QEMU_OPTION_numa,
295 "-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
296 "-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
297 "-numa dist,src=source,dst=destination,val=distance\n"
298 "-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]\n"
299 "-numa hmat-lb,initiator=node,target=node,hierarchy=memory|first-level|second-level|third-level,data-type=access-latency|read-latency|write-latency[,latency=lat][,bandwidth=bw]\n"
300 "-numa hmat-cache,node-id=node,size=size,level=level[,associativity=none|direct|complex][,policy=none|write-back|write-through][,line=size]\n",
303 ``-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
305 ``-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
307 ``-numa dist,src=source,dst=destination,val=distance``
309 ``-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]``
311 ``-numa hmat-lb,initiator=node,target=node,hierarchy=hierarchy,data-type=tpye[,latency=lat][,bandwidth=bw]``
313 ``-numa hmat-cache,node-id=node,size=size,level=level[,associativity=str][,policy=str][,line=size]``
314 Define a NUMA node and assign RAM and VCPUs to it. Set the NUMA
315 distance from a source node to a destination node. Set the ACPI
316 Heterogeneous Memory Attributes for the given nodes.
318 Legacy VCPU assignment uses '\ ``cpus``\
' option where firstcpu and
319 lastcpu are CPU indexes. Each '\ ``cpus``\
' option represent a
320 contiguous range of CPU indexes (or a single VCPU if lastcpu is
321 omitted). A non-contiguous set of VCPUs can be represented by
322 providing multiple '\ ``cpus``\
' options. If '\ ``cpus``\
' is
323 omitted on all nodes, VCPUs are automatically split between them.
325 For example, the following option assigns VCPUs 0, 1, 2 and 5 to a
330 -numa node,cpus=0-2,cpus=5
332 '\ ``cpu``\
' option is a new alternative to '\ ``cpus``\
' option
333 which uses '\ ``socket
-id|core
-id|thread
-id``\
' properties to
334 assign CPU objects to a node using topology layout properties of
335 CPU. The set of properties is machine specific, and depends on used
336 machine type/'\ ``smp``\
' options. It could be queried with
337 '\ ``hotpluggable
-cpus``\
' monitor command. '\ ``node
-id``\
'
338 property specifies node to which CPU object will be assigned, it's
339 required
for node to be declared with
'\ ``node``\ ' option before
340 it
's used with '\ ``cpu``\
' option.
347 -smp 1,sockets=2,maxcpus=2 \
348 -numa node,nodeid=0 -numa node,nodeid=1 \
349 -numa cpu,node-id=0,socket-id=0 -numa cpu,node-id=1,socket-id=1
351 Legacy '\ ``mem``\
' assigns a given RAM amount to a node (not supported
352 for 5.1 and newer machine types). '\ ``memdev``\
' assigns RAM from
353 a given memory backend device to a node. If '\ ``mem``\
' and
354 '\ ``memdev``\
' are omitted in all nodes, RAM is split equally between them.
357 '\ ``mem``\
' and '\ ``memdev``\
' are mutually exclusive.
358 Furthermore, if one node uses '\ ``memdev``\
', all of them have to
361 '\ ``initiator``\
' is an additional option that points to an
362 initiator NUMA node that has best performance (the lowest latency or
363 largest bandwidth) to this NUMA node. Note that this option can be
364 set only when the machine property 'hmat
' is set to 'on
'.
366 Following example creates a machine with 2 NUMA nodes, node 0 has
367 CPU. node 1 has only memory, and its initiator is node 0. Note that
368 because node 0 has CPU, by default the initiator of node 0 is itself
374 -m 2G,slots=2,maxmem=4G \
375 -object memory-backend-ram,size=1G,id=m0 \
376 -object memory-backend-ram,size=1G,id=m1 \
377 -numa node,nodeid=0,memdev=m0 \
378 -numa node,nodeid=1,memdev=m1,initiator=0 \
379 -smp 2,sockets=2,maxcpus=2 \
380 -numa cpu,node-id=0,socket-id=0 \
381 -numa cpu,node-id=0,socket-id=1
383 source and destination are NUMA node IDs. distance is the NUMA
384 distance from source to destination. The distance from a node to
385 itself is always 10. If any pair of nodes is given a distance, then
386 all pairs must be given distances. Although, when distances are only
387 given in one direction for each pair of nodes, then the distances in
388 the opposite directions are assumed to be the same. If, however, an
389 asymmetrical pair of distances is given for even one node pair, then
390 all node pairs must be provided distance values for both directions,
391 even when they are symmetrical. When a node is unreachable from
392 another node, set the pair's distance to
255.
394 Note that the
-``numa`` option doesn
't allocate any of the specified
395 resources, it just assigns existing resources to NUMA nodes. This
396 means that one still has to use the ``-m``, ``-smp`` options to
397 allocate RAM and VCPUs respectively.
399 Use '\ ``hmat
-lb``\
' to set System Locality Latency and Bandwidth
400 Information between initiator and target NUMA nodes in ACPI
401 Heterogeneous Attribute Memory Table (HMAT). Initiator NUMA node can
402 create memory requests, usually it has one or more processors.
403 Target NUMA node contains addressable memory.
405 In '\ ``hmat
-lb``\
' option, node are NUMA node IDs. hierarchy is
406 the memory hierarchy of the target NUMA node: if hierarchy is
407 'memory
', the structure represents the memory performance; if
408 hierarchy is 'first
-level\|second
-level\|third
-level
', this
409 structure represents aggregated performance of memory side caches
410 for each domain. type of 'data
-type
' is type of data represented by
411 this structure instance: if 'hierarchy
' is 'memory
', 'data
-type
' is
412 'access\|read\|write
' latency or 'access\|read\|write
' bandwidth of
413 the target memory; if 'hierarchy
' is
414 'first
-level\|second
-level\|third
-level
', 'data
-type
' is
415 'access\|read\|write
' hit latency or 'access\|read\|write
' hit
416 bandwidth of the target memory side cache.
418 lat is latency value in nanoseconds. bw is bandwidth value, the
419 possible value and units are NUM[M\|G\|T], mean that the bandwidth
420 value are NUM byte per second (or MB/s, GB/s or TB/s depending on
421 used suffix). Note that if latency or bandwidth value is 0, means
422 the corresponding latency or bandwidth information is not provided.
424 In '\ ``hmat
-cache``\
' option, node-id is the NUMA-id of the memory
425 belongs. size is the size of memory side cache in bytes. level is
426 the cache level described in this structure, note that the cache
427 level 0 should not be used with '\ ``hmat
-cache``\
' option.
428 associativity is the cache associativity, the possible value is
429 'none
/direct(direct
-mapped
)/complex(complex cache indexing
)'. policy
430 is the write policy. line is the cache Line size in bytes.
432 For example, the following options describe 2 NUMA nodes. Node 0 has
433 2 cpus and a ram, node 1 has only a ram. The processors in node 0
434 access memory in node 0 with access-latency 5 nanoseconds,
435 access-bandwidth is 200 MB/s; The processors in NUMA node 0 access
436 memory in NUMA node 1 with access-latency 10 nanoseconds,
437 access-bandwidth is 100 MB/s. And for memory side cache information,
438 NUMA node 0 and 1 both have 1 level memory cache, size is 10KB,
439 policy is write-back, the cache Line size is 8 bytes:
445 -object memory-backend-ram,size=1G,id=m0 \
446 -object memory-backend-ram,size=1G,id=m1 \
447 -smp 2,sockets=2,maxcpus=2 \
448 -numa node,nodeid=0,memdev=m0 \
449 -numa node,nodeid=1,memdev=m1,initiator=0 \
450 -numa cpu,node-id=0,socket-id=0 \
451 -numa cpu,node-id=0,socket-id=1 \
452 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-latency,latency=5 \
453 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-bandwidth,bandwidth=200M \
454 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-latency,latency=10 \
455 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-bandwidth,bandwidth=100M \
456 -numa hmat-cache,node-id=0,size=10K,level=1,associativity=direct,policy=write-back,line=8 \
457 -numa hmat-cache,node-id=1,size=10K,level=1,associativity=direct,policy=write-back,line=8
460 DEF("add-fd", HAS_ARG, QEMU_OPTION_add_fd,
461 "-add-fd fd=fd,set=set[,opaque=opaque]\n"
462 " Add 'fd
' to fd 'set
'\n", QEMU_ARCH_ALL)
464 ``-add-fd fd=fd,set=set[,opaque=opaque]``
465 Add a file descriptor to an fd set. Valid options are:
468 This option defines the file descriptor of which a duplicate is
469 added to fd set. The file descriptor cannot be stdin, stdout, or
473 This option defines the ID of the fd set to add the file
477 This option defines a free-form string that can be used to
480 You can open an image using pre-opened file descriptors from an fd
486 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
487 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
488 -drive file=/dev/fdset/2,index=0,media=disk
491 DEF("set", HAS_ARG, QEMU_OPTION_set,
492 "-set group.id.arg=value\n"
493 " set <arg> parameter for item <id> of type <group>\n"
494 " i.e. -set drive.$id.file=/path/to/image\n", QEMU_ARCH_ALL)
496 ``-set group.id.arg=value``
497 Set parameter arg for item id of type group
500 DEF("global", HAS_ARG, QEMU_OPTION_global,
501 "-global driver.property=value\n"
502 "-global driver=driver,property=property,value=value\n"
503 " set a global default for a driver property\n",
506 ``-global driver.prop=value``
508 ``-global driver=driver,property=property,value=value``
509 Set default value of driver's property prop to value
, e
.g
.:
513 |qemu_system_x86|
-global ide
-hd
.physical_block_size
=4096 disk
-image
.img
515 In particular
, you can use
this to set driver properties
for devices
516 which are created automatically by the machine model
. To create a
517 device which is not created automatically and set properties on it
,
520 -global driver
.prop
=value is shorthand
for -global
521 driver
=driver
,property
=prop
,value
=value
. The longhand syntax works
522 even when driver contains a dot
.
525 DEF("boot", HAS_ARG
, QEMU_OPTION_boot
,
526 "-boot [order=drives][,once=drives][,menu=on|off]\n"
527 " [,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_time][,strict=on|off]\n"
528 " 'drives': floppy (a), hard disk (c), CD-ROM (d), network (n)\n"
529 " 'sp_name': the file's name that would be passed to bios as logo picture, if menu=on\n"
530 " 'sp_time': the period that splash picture last if menu=on, unit is ms\n"
531 " 'rb_timeout': the timeout before guest reboot when boot failed, unit is ms\n",
534 ``
-boot
[order
=drives
][,once
=drives
][,menu
=on|off
][,splash
=sp_name
][,splash
-time
=sp_time
][,reboot
-timeout
=rb_timeout
][,strict
=on|off
]``
535 Specify boot order drives as a string of drive letters
. Valid drive
536 letters depend on the target architecture
. The x86 PC uses
: a
, b
537 (floppy
1 and
2), c (first hard disk
), d (first CD
-ROM
), n
-p
538 (Etherboot from network adapter
1-4), hard disk boot is the
default.
539 To apply a particular boot order only on the first startup
, specify
540 it via ``once``
. Note that the ``order`` or ``once`` parameter
541 should not be used together with the ``bootindex`` property of
542 devices
, since the firmware implementations normally
do not support
543 both at the same time
.
545 Interactive boot menus
/prompts can be enabled via ``menu
=on`` as far
546 as firmware
/BIOS supports them
. The
default is non
-interactive boot
.
548 A splash picture could be passed to bios
, enabling user to show it
549 as logo
, when option splash
=sp\_name is given and menu
=on
, If
550 firmware
/BIOS supports them
. Currently Seabios
for X86 system
551 support it
. limitation
: The splash file could be a jpeg file or a
552 BMP file
in 24 BPP
format(true color
). The resolution should be
553 supported by the SVGA mode
, so the recommended is
320x240
, 640x480
,
556 A timeout could be passed to bios
, guest will pause
for rb\_timeout
557 ms when boot failed
, then reboot
. If rb\_timeout is
'-1', guest will
558 not reboot
, qemu passes
'-1' to bios by
default. Currently Seabios
559 for X86 system support it
.
561 Do strict boot via ``strict
=on`` as far as firmware
/BIOS supports
562 it
. This only effects when boot priority is changed by bootindex
563 options
. The
default is non
-strict boot
.
567 #
try to boot from network first
, then from hard disk
568 |qemu_system_x86|
-boot order
=nc
569 # boot from CD
-ROM first
, switch back to
default order after reboot
570 |qemu_system_x86|
-boot once
=d
571 # boot with a splash picture
for 5 seconds
.
572 |qemu_system_x86|
-boot menu
=on
,splash
=/root
/boot
.bmp
,splash
-time
=5000
574 Note
: The legacy format
'-boot drives' is still supported but its
575 use is discouraged as it may be removed from future versions
.
578 DEF("m", HAS_ARG
, QEMU_OPTION_m
,
579 "-m [size=]megs[,slots=n,maxmem=size]\n"
580 " configure guest RAM\n"
581 " size: initial amount of guest memory\n"
582 " slots: number of hotplug slots (default: none)\n"
583 " maxmem: maximum amount of guest memory (default: none)\n"
584 "NOTE: Some architectures might enforce a specific granularity\n",
587 ``
-m
[size
=]megs
[,slots
=n
,maxmem
=size
]``
588 Sets guest startup RAM size to megs megabytes
. Default is
128 MiB
.
589 Optionally
, a suffix of
"M" or
"G" can be used to signify a value
in
590 megabytes or gigabytes respectively
. Optional pair slots
, maxmem
591 could be used to set amount of hotpluggable memory slots and maximum
592 amount of memory
. Note that maxmem must be aligned to the page size
.
594 For example
, the following command
-line sets the guest startup RAM
595 size to
1GB
, creates
3 slots to hotplug additional memory and sets
596 the maximum memory the guest can reach to
4GB
:
600 |qemu_system|
-m
1G
,slots
=3,maxmem
=4G
602 If slots and maxmem are not specified
, memory hotplug won
't be
603 enabled and the guest startup RAM will never increase.
606 DEF("mem-path", HAS_ARG, QEMU_OPTION_mempath,
607 "-mem-path FILE provide backing storage for guest RAM\n", QEMU_ARCH_ALL)
610 Allocate guest RAM from a temporarily created file in path.
613 DEF("mem-prealloc", 0, QEMU_OPTION_mem_prealloc,
614 "-mem-prealloc preallocate guest memory (use with -mem-path)\n",
618 Preallocate memory when using -mem-path.
621 DEF("k", HAS_ARG, QEMU_OPTION_k,
622 "-k language use keyboard layout (for example 'fr
' for French)\n",
626 Use keyboard layout language (for example ``fr`` for French). This
627 option is only needed where it is not easy to get raw PC keycodes
628 (e.g. on Macs, with some X11 servers or with a VNC or curses
629 display). You don't normally need to use it on PC
/Linux or
632 The available layouts are
:
636 ar de
-ch es fo fr
-ca hu ja mk no pt
-br sv
637 da en
-gb et fr fr
-ch is lt nl pl ru th
638 de en
-us fi fr
-be hr it lv nl
-be pt sl tr
640 The
default is ``en
-us``
.
644 HXCOMM Deprecated by
-audiodev
645 DEF("audio-help", 0, QEMU_OPTION_audio_help
,
646 "-audio-help show -audiodev equivalent of the currently specified audio settings\n",
650 Will show the
-audiodev equivalent of the currently specified
651 (deprecated
) environment variables
.
654 DEF("audiodev", HAS_ARG
, QEMU_OPTION_audiodev
,
655 "-audiodev [driver=]driver,id=id[,prop[=value][,...]]\n"
656 " specifies the audio backend to use\n"
657 " id= identifier of the backend\n"
658 " timer-period= timer period in microseconds\n"
659 " in|out.mixing-engine= use mixing engine to mix streams inside QEMU\n"
660 " in|out.fixed-settings= use fixed settings for host audio\n"
661 " in|out.frequency= frequency to use with fixed settings\n"
662 " in|out.channels= number of channels to use with fixed settings\n"
663 " in|out.format= sample format to use with fixed settings\n"
664 " valid values: s8, s16, s32, u8, u16, u32, f32\n"
665 " in|out.voices= number of voices to use\n"
666 " in|out.buffer-length= length of buffer in microseconds\n"
667 "-audiodev none,id=id,[,prop[=value][,...]]\n"
668 " dummy driver that discards all output\n"
669 #ifdef CONFIG_AUDIO_ALSA
670 "-audiodev alsa,id=id[,prop[=value][,...]]\n"
671 " in|out.dev= name of the audio device to use\n"
672 " in|out.period-length= length of period in microseconds\n"
673 " in|out.try-poll= attempt to use poll mode\n"
674 " threshold= threshold (in microseconds) when playback starts\n"
676 #ifdef CONFIG_AUDIO_COREAUDIO
677 "-audiodev coreaudio,id=id[,prop[=value][,...]]\n"
678 " in|out.buffer-count= number of buffers\n"
680 #ifdef CONFIG_AUDIO_DSOUND
681 "-audiodev dsound,id=id[,prop[=value][,...]]\n"
682 " latency= add extra latency to playback in microseconds\n"
684 #ifdef CONFIG_AUDIO_OSS
685 "-audiodev oss,id=id[,prop[=value][,...]]\n"
686 " in|out.dev= path of the audio device to use\n"
687 " in|out.buffer-count= number of buffers\n"
688 " in|out.try-poll= attempt to use poll mode\n"
689 " try-mmap= try using memory mapped access\n"
690 " exclusive= open device in exclusive mode\n"
691 " dsp-policy= set timing policy (0..10), -1 to use fragment mode\n"
693 #ifdef CONFIG_AUDIO_PA
694 "-audiodev pa,id=id[,prop[=value][,...]]\n"
695 " server= PulseAudio server address\n"
696 " in|out.name= source/sink device name\n"
697 " in|out.latency= desired latency in microseconds\n"
699 #ifdef CONFIG_AUDIO_SDL
700 "-audiodev sdl,id=id[,prop[=value][,...]]\n"
701 " in|out.buffer-count= number of buffers\n"
704 "-audiodev spice,id=id[,prop[=value][,...]]\n"
706 #ifdef CONFIG_DBUS_DISPLAY
707 "-audiodev dbus,id=id[,prop[=value][,...]]\n"
709 "-audiodev wav,id=id[,prop[=value][,...]]\n"
710 " path= path of wav file to record\n",
713 ``
-audiodev
[driver
=]driver
,id
=id
[,prop
[=value
][,...]]``
714 Adds a
new audio backend driver identified by id
. There are global
715 and driver specific properties
. Some values can be set differently
716 for input and output
, they
're marked with ``in|out.``. You can set
717 the input's property with ``
in.prop`` and the output
's property with
718 ``out.prop``. For example:
722 -audiodev alsa,id=example,in.frequency=44110,out.frequency=8000
723 -audiodev alsa,id=example,out.channels=1 # leaves in.channels unspecified
725 NOTE: parameter validation is known to be incomplete, in many cases
726 specifying an invalid option causes QEMU to print an error message
727 and continue emulation without sound.
729 Valid global options are:
732 Identifies the audio backend.
734 ``timer-period=period``
735 Sets the timer period used by the audio subsystem in
736 microseconds. Default is 10000 (10 ms).
738 ``in|out.mixing-engine=on|off``
739 Use QEMU's mixing engine to mix all streams inside QEMU and
740 convert audio formats when not supported by the backend
. When
741 off
, fixed
-settings must be off too
. Note that disabling
this
742 option means that the selected backend must support multiple
743 streams and the audio formats used by the virtual cards
,
744 otherwise you
'll get no sound. It's not recommended to disable
745 this option unless you want to use
5.1 or
7.1 audio
, as mixing
746 engine only supports mono and stereo audio
. Default is on
.
748 ``
in|out
.fixed
-settings
=on|off``
749 Use fixed settings
for host audio
. When off
, it will change
750 based on how the guest opens the sound card
. In
this case you
751 must not specify frequency
, channels or format
. Default is on
.
753 ``
in|out
.frequency
=frequency``
754 Specify the frequency to use when
using fixed
-settings
. Default
757 ``
in|out
.channels
=channels``
758 Specify the number of channels to use when
using fixed
-settings
.
759 Default is
2 (stereo
).
761 ``
in|out
.format
=format``
762 Specify the sample format to use when
using fixed
-settings
.
763 Valid values are
: ``s8``
, ``s16``
, ``s32``
, ``u8``
, ``u16``
,
764 ``u32``
, ``f32``
. Default is ``s16``
.
766 ``
in|out
.voices
=voices``
767 Specify the number of voices to use
. Default is
1.
769 ``
in|out
.buffer
-length
=usecs``
770 Sets the size of the buffer
in microseconds
.
772 ``
-audiodev none
,id
=id
[,prop
[=value
][,...]]``
773 Creates a dummy backend that discards all outputs
. This backend has
774 no backend specific properties
.
776 ``
-audiodev alsa
,id
=id
[,prop
[=value
][,...]]``
777 Creates backend
using the ALSA
. This backend is only available on
780 ALSA specific options are
:
782 ``
in|out
.dev
=device``
783 Specify the ALSA device to use
for input and
/or output
. Default
786 ``
in|out
.period
-length
=usecs``
787 Sets the period length
in microseconds
.
789 ``
in|out
.try-poll
=on|off``
790 Attempt to use poll mode with the device
. Default is on
.
792 ``threshold
=threshold``
793 Threshold (in microseconds
) when playback starts
. Default is
0.
795 ``
-audiodev coreaudio
,id
=id
[,prop
[=value
][,...]]``
796 Creates a backend
using Apple
's Core Audio. This backend is only
797 available on Mac OS and only supports playback.
799 Core Audio specific options are:
801 ``in|out.buffer-count=count``
802 Sets the count of the buffers.
804 ``-audiodev dsound,id=id[,prop[=value][,...]]``
805 Creates a backend using Microsoft's DirectSound
. This backend is
806 only available on Windows and only supports playback
.
808 DirectSound specific options are
:
811 Add extra usecs microseconds latency to playback
. Default is
814 ``
-audiodev oss
,id
=id
[,prop
[=value
][,...]]``
815 Creates a backend
using OSS
. This backend is available on most
818 OSS specific options are
:
820 ``
in|out
.dev
=device``
821 Specify the file name of the OSS device to use
. Default is
824 ``
in|out
.buffer
-count
=count``
825 Sets the count of the buffers
.
827 ``
in|out
.try-poll
=on|of``
828 Attempt to use poll mode with the device
. Default is on
.
831 Try
using memory mapped device access
. Default is off
.
834 Open the device
in exclusive
mode (vmix won
't work in this
835 case). Default is off.
837 ``dsp-policy=policy``
838 Sets the timing policy (between 0 and 10, where smaller number
839 means smaller latency but higher CPU usage). Use -1 to use
840 buffer sizes specified by ``buffer`` and ``buffer-count``. This
841 option is ignored if you do not have OSS 4. Default is 5.
843 ``-audiodev pa,id=id[,prop[=value][,...]]``
844 Creates a backend using PulseAudio. This backend is available on
847 PulseAudio specific options are:
850 Sets the PulseAudio server to connect to.
853 Use the specified source/sink for recording/playback.
855 ``in|out.latency=usecs``
856 Desired latency in microseconds. The PulseAudio server will try
857 to honor this value but actual latencies may be lower or higher.
859 ``-audiodev sdl,id=id[,prop[=value][,...]]``
860 Creates a backend using SDL. This backend is available on most
861 systems, but you should use your platform's native backend
if
864 SDL specific options are
:
866 ``
in|out
.buffer
-count
=count``
867 Sets the count of the buffers
.
869 ``
-audiodev spice
,id
=id
[,prop
[=value
][,...]]``
870 Creates a backend that sends audio through SPICE
. This backend
871 requires ``
-spice`` and automatically selected
in that
case, so
872 usually you can ignore
this option
. This backend has no backend
875 ``
-audiodev wav
,id
=id
[,prop
[=value
][,...]]``
876 Creates a backend that writes audio to a WAV file
.
878 Backend specific options are
:
881 Write recorded audio into the specified file
. Default is
885 DEF("soundhw", HAS_ARG
, QEMU_OPTION_soundhw
,
886 "-soundhw c1,... enable audio support\n"
887 " and only specified sound cards (comma separated list)\n"
888 " use '-soundhw help' to get the list of supported cards\n"
889 " use '-soundhw all' to enable all of them\n", QEMU_ARCH_ALL
)
891 ``
-soundhw card1
[,card2
,...] or
-soundhw all``
892 Enable audio and selected sound hardware
. Use
'help' to print all
893 available sound hardware
. For example
:
897 |qemu_system_x86|
-soundhw sb16
,adlib disk
.img
898 |qemu_system_x86|
-soundhw es1370 disk
.img
899 |qemu_system_x86|
-soundhw ac97 disk
.img
900 |qemu_system_x86|
-soundhw hda disk
.img
901 |qemu_system_x86|
-soundhw all disk
.img
902 |qemu_system_x86|
-soundhw help
904 Note that Linux
's i810\_audio OSS kernel (for AC97) module might
905 require manually specifying clocking.
909 modprobe i810_audio clocking=48000
912 DEF("device", HAS_ARG, QEMU_OPTION_device,
913 "-device driver[,prop[=value][,...]]\n"
914 " add device (based on driver)\n"
915 " prop=value,... sets driver properties\n"
916 " use '-device help
' to print all possible drivers\n"
917 " use '-device driver
,help
' to print all possible properties\n",
920 ``-device driver[,prop[=value][,...]]``
921 Add device driver. prop=value sets driver properties. Valid
922 properties depend on the driver. To get help on possible drivers and
923 properties, use ``-device help`` and ``-device driver,help``.
927 ``-device ipmi-bmc-sim,id=id[,prop[=value][,...]]``
928 Add an IPMI BMC. This is a simulation of a hardware management
929 interface processor that normally sits on a system. It provides a
930 watchdog and the ability to reset and power control the system. You
931 need to connect this to an IPMI interface to make it useful
933 The IPMI slave address to use for the BMC. The default is 0x20. This
934 address is the BMC's address on the I2C network of management
935 controllers
. If you don
't know what this means, it is safe to ignore
939 The BMC id for interfaces to use this device.
942 Define slave address to use for the BMC. The default is 0x20.
945 file containing raw Sensor Data Records (SDR) data. The default
949 size of a Field Replaceable Unit (FRU) area. The default is
953 file containing raw Field Replaceable Unit (FRU) inventory data.
957 value for the GUID for the BMC, in standard UUID format. If this
958 is set, get "Get GUID" command to the BMC will return it.
959 Otherwise "Get GUID" will return an error.
961 ``-device ipmi-bmc-extern,id=id,chardev=id[,slave_addr=val]``
962 Add a connection to an external IPMI BMC simulator. Instead of
963 locally emulating the BMC like the above item, instead connect to an
964 external entity that provides the IPMI services.
966 A connection is made to an external BMC simulator. If you do this,
967 it is strongly recommended that you use the "reconnect=" chardev
968 option to reconnect to the simulator if the connection is lost. Note
969 that if this is not used carefully, it can be a security issue, as
970 the interface has the ability to send resets, NMIs, and power off
971 the VM. It's best
if QEMU makes a connection to an external
972 simulator running on a secure port on localhost
, so neither the
973 simulator nor QEMU is exposed to any outside network
.
975 See the
"lanserv/README.vm" file
in the OpenIPMI library
for more
976 details on the external
interface.
978 ``
-device isa
-ipmi
-kcs
,bmc
=id
[,ioport
=val
][,irq
=val
]``
979 Add a KCS IPMI interafce on the ISA bus
. This also adds a
980 corresponding ACPI and SMBIOS entries
, if appropriate
.
983 The BMC to connect to
, one of ipmi
-bmc
-sim or ipmi
-bmc
-extern
987 Define the I
/O address of the
interface. The
default is
0xca0
991 Define the interrupt to use
. The
default is
5. To disable
992 interrupts
, set
this to
0.
994 ``
-device isa
-ipmi
-bt
,bmc
=id
[,ioport
=val
][,irq
=val
]``
995 Like the KCS
interface, but defines a BT
interface. The
default port
996 is
0xe4 and the
default interrupt is
5.
998 ``
-device pci
-ipmi
-kcs
,bmc
=id``
999 Add a KCS IPMI interafce on the PCI bus
.
1002 The BMC to connect to
, one of ipmi
-bmc
-sim or ipmi
-bmc
-extern above
.
1004 ``
-device pci
-ipmi
-bt
,bmc
=id``
1005 Like the KCS
interface, but defines a BT
interface on the PCI bus
.
1007 ``
-device intel
-iommu
[,option
=...]``
1008 This is only supported by ``
-machine q35``
, which will enable Intel VT
-d
1009 emulation within the guest
. It supports below options
:
1011 ``intremap
=on|off``
(default: auto
)
1012 This enables interrupt remapping feature
. It
's required to enable
1013 complete x2apic. Currently it only supports kvm kernel-irqchip modes
1014 ``off`` or ``split``, while full kernel-irqchip is not yet supported.
1015 The default value is "auto", which will be decided by the mode of
1018 ``caching-mode=on|off`` (default: off)
1019 This enables caching mode for the VT-d emulated device. When
1020 caching-mode is enabled, each guest DMA buffer mapping will generate an
1021 IOTLB invalidation from the guest IOMMU driver to the vIOMMU device in
1022 a synchronous way. It is required for ``-device vfio-pci`` to work
1023 with the VT-d device, because host assigned devices requires to setup
1024 the DMA mapping on the host before guest DMA starts.
1026 ``device-iotlb=on|off`` (default: off)
1027 This enables device-iotlb capability for the emulated VT-d device. So
1028 far virtio/vhost should be the only real user for this parameter,
1029 paired with ats=on configured for the device.
1031 ``aw-bits=39|48`` (default: 39)
1032 This decides the address width of IOVA address space. The address
1033 space has 39 bits width for 3-level IOMMU page tables, and 48 bits for
1034 4-level IOMMU page tables.
1036 Please also refer to the wiki page for general scenarios of VT-d
1037 emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
1041 DEF("name", HAS_ARG, QEMU_OPTION_name,
1042 "-name string1[,process=string2][,debug-threads=on|off]\n"
1043 " set the name of the guest\n"
1044 " string1 sets the window title and string2 the process name\n"
1045 " When debug-threads is enabled, individual threads are given a separate name\n"
1046 " NOTE: The thread names are for debugging and not a stable API.\n",
1050 Sets the name of the guest. This name will be displayed in the SDL
1051 window caption. The name will also be used for the VNC server. Also
1052 optionally set the top visible process name in Linux. Naming of
1053 individual threads can also be enabled on Linux to aid debugging.
1056 DEF("uuid", HAS_ARG, QEMU_OPTION_uuid,
1057 "-uuid %08x-%04x-%04x-%04x-%012x\n"
1058 " specify machine UUID\n", QEMU_ARCH_ALL)
1066 DEFHEADING(Block device options:)
1068 DEF("fda", HAS_ARG, QEMU_OPTION_fda,
1069 "-fda/-fdb file use 'file
' as floppy disk 0/1 image\n", QEMU_ARCH_ALL)
1070 DEF("fdb", HAS_ARG, QEMU_OPTION_fdb, "", QEMU_ARCH_ALL)
1075 Use file as floppy disk 0/1 image (see the :ref:`disk images` chapter in
1076 the System Emulation Users Guide).
1079 DEF("hda", HAS_ARG, QEMU_OPTION_hda,
1080 "-hda/-hdb file use 'file
' as IDE hard disk 0/1 image\n", QEMU_ARCH_ALL)
1081 DEF("hdb", HAS_ARG, QEMU_OPTION_hdb, "", QEMU_ARCH_ALL)
1082 DEF("hdc", HAS_ARG, QEMU_OPTION_hdc,
1083 "-hdc/-hdd file use 'file
' as IDE hard disk 2/3 image\n", QEMU_ARCH_ALL)
1084 DEF("hdd", HAS_ARG, QEMU_OPTION_hdd, "", QEMU_ARCH_ALL)
1093 Use file as hard disk 0, 1, 2 or 3 image (see the :ref:`disk images`
1094 chapter in the System Emulation Users Guide).
1097 DEF("cdrom", HAS_ARG, QEMU_OPTION_cdrom,
1098 "-cdrom file use 'file
' as IDE cdrom image (cdrom is ide1 master)\n",
1102 Use file as CD-ROM image (you cannot use ``-hdc`` and ``-cdrom`` at
1103 the same time). You can use the host CD-ROM by using ``/dev/cdrom``
1107 DEF("blockdev", HAS_ARG, QEMU_OPTION_blockdev,
1108 "-blockdev [driver=]driver[,node-name=N][,discard=ignore|unmap]\n"
1109 " [,cache.direct=on|off][,cache.no-flush=on|off]\n"
1110 " [,read-only=on|off][,auto-read-only=on|off]\n"
1111 " [,force-share=on|off][,detect-zeroes=on|off|unmap]\n"
1112 " [,driver specific parameters...]\n"
1113 " configure a block backend\n", QEMU_ARCH_ALL)
1115 ``-blockdev option[,option[,option[,...]]]``
1116 Define a new block driver node. Some of the options apply to all
1117 block drivers, other options are only accepted for a specific block
1118 driver. See below for a list of generic options and options for the
1119 most common block drivers.
1121 Options that expect a reference to another node (e.g. ``file``) can
1122 be given in two ways. Either you specify the node name of an already
1123 existing node (file=node-name), or you define a new node inline,
1124 adding options for the referenced node after a dot
1125 (file.filename=path,file.aio=native).
1127 A block driver node created with ``-blockdev`` can be used for a
1128 guest device by specifying its node name for the ``drive`` property
1129 in a ``-device`` argument that defines a block device.
1131 ``Valid options for any block driver node:``
1133 Specifies the block driver to use for the given node.
1136 This defines the name of the block driver node by which it
1137 will be referenced later. The name must be unique, i.e. it
1138 must not match the name of a different block driver node, or
1139 (if you use ``-drive`` as well) the ID of a drive.
1141 If no node name is specified, it is automatically generated.
1142 The generated node name is not intended to be predictable
1143 and changes between QEMU invocations. For the top level, an
1144 explicit node name must be specified.
1147 Open the node read-only. Guest write attempts will fail.
1149 Note that some block drivers support only read-only access,
1150 either generally or in certain configurations. In this case,
1151 the default value ``read-only=off`` does not work and the
1152 option must be specified explicitly.
1155 If ``auto-read-only=on`` is set, QEMU may fall back to
1156 read-only usage even when ``read-only=off`` is requested, or
1157 even switch between modes as needed, e.g. depending on
1158 whether the image file is writable or whether a writing user
1159 is attached to the node.
1162 Override the image locking system of QEMU by forcing the
1163 node to utilize weaker shared access for permissions where
1164 it would normally request exclusive access. When there is
1165 the potential for multiple instances to have the same file
1166 open (whether this invocation of QEMU is the first or the
1167 second instance), both instances must permit shared access
1168 for the second instance to succeed at opening the file.
1170 Enabling ``force-share=on`` requires ``read-only=on``.
1173 The host page cache can be avoided with ``cache.direct=on``.
1174 This will attempt to do disk IO directly to the guest's
1175 memory
. QEMU may still perform an internal copy of the data
.
1178 In
case you don
't care about data integrity over host
1179 failures, you can use ``cache.no-flush=on``. This option
1180 tells QEMU that it never needs to write any data to the disk
1181 but can instead keep things in cache. If anything goes
1182 wrong, like your host losing power, the disk storage getting
1183 disconnected accidentally, etc. your image will most
1184 probably be rendered unusable.
1187 discard is one of "ignore" (or "off") or "unmap" (or "on")
1188 and controls whether ``discard`` (also known as ``trim`` or
1189 ``unmap``) requests are ignored or passed to the filesystem.
1190 Some machine types may not support discard requests.
1192 ``detect-zeroes=detect-zeroes``
1193 detect-zeroes is "off", "on" or "unmap" and enables the
1194 automatic conversion of plain zero writes by the OS to
1195 driver specific optimized zero write commands. You may even
1196 choose "unmap" if discard is set to "unmap" to allow a zero
1197 write to be converted to an ``unmap`` operation.
1199 ``Driver-specific options for file``
1200 This is the protocol-level block driver for accessing regular
1204 The path to the image file in the local filesystem
1207 Specifies the AIO backend (threads/native/io_uring,
1211 Specifies whether the image file is protected with Linux OFD
1212 / POSIX locks. The default is to use the Linux Open File
1213 Descriptor API if available, otherwise no lock is applied.
1214 (auto/on/off, default: auto)
1220 -blockdev driver=file,node-name=disk,filename=disk.img
1222 ``Driver-specific options for raw``
1223 This is the image format block driver for raw images. It is
1224 usually stacked on top of a protocol level block driver such as
1228 Reference to or definition of the data source block driver
1229 node (e.g. a ``file`` driver node)
1235 -blockdev driver=file,node-name=disk_file,filename=disk.img
1236 -blockdev driver=raw,node-name=disk,file=disk_file
1242 -blockdev driver=raw,node-name=disk,file.driver=file,file.filename=disk.img
1244 ``Driver-specific options for qcow2``
1245 This is the image format block driver for qcow2 images. It is
1246 usually stacked on top of a protocol level block driver such as
1250 Reference to or definition of the data source block driver
1251 node (e.g. a ``file`` driver node)
1254 Reference to or definition of the backing file block device
1255 (default is taken from the image file). It is allowed to
1256 pass ``null`` here in order to disable the default backing
1260 Whether to enable the lazy refcounts feature (on/off;
1261 default is taken from the image file)
1264 The maximum total size of the L2 table and refcount block
1265 caches in bytes (default: the sum of l2-cache-size and
1266 refcount-cache-size)
1269 The maximum size of the L2 table cache in bytes (default: if
1270 cache-size is not specified - 32M on Linux platforms, and 8M
1271 on non-Linux platforms; otherwise, as large as possible
1272 within the cache-size, while permitting the requested or the
1273 minimal refcount cache size)
1275 ``refcount-cache-size``
1276 The maximum size of the refcount block cache in bytes
1277 (default: 4 times the cluster size; or if cache-size is
1278 specified, the part of it which is not used for the L2
1281 ``cache-clean-interval``
1282 Clean unused entries in the L2 and refcount caches. The
1283 interval is in seconds. The default value is 600 on
1284 supporting platforms, and 0 on other platforms. Setting it
1285 to 0 disables this feature.
1287 ``pass-discard-request``
1288 Whether discard requests to the qcow2 device should be
1289 forwarded to the data source (on/off; default: on if
1290 discard=unmap is specified, off otherwise)
1292 ``pass-discard-snapshot``
1293 Whether discard requests for the data source should be
1294 issued when a snapshot operation (e.g. deleting a snapshot)
1295 frees clusters in the qcow2 file (on/off; default: on)
1297 ``pass-discard-other``
1298 Whether discard requests for the data source should be
1299 issued on other occasions where a cluster gets freed
1300 (on/off; default: off)
1303 Which overlap checks to perform for writes to the image
1304 (none/constant/cached/all; default: cached). For details or
1305 finer granularity control refer to the QAPI documentation of
1312 -blockdev driver=file,node-name=my_file,filename=/tmp/disk.qcow2
1313 -blockdev driver=qcow2,node-name=hda,file=my_file,overlap-check=none,cache-size=16777216
1319 -blockdev driver=qcow2,node-name=disk,file.driver=http,file.filename=http://example.com/image.qcow2
1321 ``Driver-specific options for other drivers``
1322 Please refer to the QAPI documentation of the ``blockdev-add``
1326 DEF("drive", HAS_ARG, QEMU_OPTION_drive,
1327 "-drive [file=file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n"
1328 " [,cache=writethrough|writeback|none|directsync|unsafe][,format=f]\n"
1329 " [,snapshot=on|off][,rerror=ignore|stop|report]\n"
1330 " [,werror=ignore|stop|report|enospc][,id=name]\n"
1331 " [,aio=threads|native|io_uring]\n"
1332 " [,readonly=on|off][,copy-on-read=on|off]\n"
1333 " [,discard=ignore|unmap][,detect-zeroes=on|off|unmap]\n"
1334 " [[,bps=b]|[[,bps_rd=r][,bps_wr=w]]]\n"
1335 " [[,iops=i]|[[,iops_rd=r][,iops_wr=w]]]\n"
1336 " [[,bps_max=bm]|[[,bps_rd_max=rm][,bps_wr_max=wm]]]\n"
1337 " [[,iops_max=im]|[[,iops_rd_max=irm][,iops_wr_max=iwm]]]\n"
1338 " [[,iops_size=is]]\n"
1340 " use 'file
' as a drive image\n", QEMU_ARCH_ALL)
1342 ``-drive option[,option[,option[,...]]]``
1343 Define a new drive. This includes creating a block driver node (the
1344 backend) as well as a guest device, and is mostly a shortcut for
1345 defining the corresponding ``-blockdev`` and ``-device`` options.
1347 ``-drive`` accepts all options that are accepted by ``-blockdev``.
1348 In addition, it knows the following options:
1351 This option defines which disk image (see the :ref:`disk images`
1352 chapter in the System Emulation Users Guide) to use with this drive.
1353 If the filename contains comma, you must double it (for instance,
1354 "file=my,,file" to use file "my,file").
1356 Special files such as iSCSI devices can be specified using
1357 protocol specific URLs. See the section for "Device URL Syntax"
1358 for more information.
1361 This option defines on which type on interface the drive is
1362 connected. Available types are: ide, scsi, sd, mtd, floppy,
1363 pflash, virtio, none.
1365 ``bus=bus,unit=unit``
1366 These options define where is connected the drive by defining
1367 the bus number and the unit id.
1370 This option defines where is connected the drive by using an
1371 index in the list of available connectors of a given interface
1375 This option defines the type of the media: disk or cdrom.
1377 ``snapshot=snapshot``
1378 snapshot is "on" or "off" and controls snapshot mode for the
1379 given drive (see ``-snapshot``).
1382 cache is "none", "writeback", "unsafe", "directsync" or
1383 "writethrough" and controls how the host cache is used to access
1384 block data. This is a shortcut that sets the ``cache.direct``
1385 and ``cache.no-flush`` options (as in ``-blockdev``), and
1386 additionally ``cache.writeback``, which provides a default for
1387 the ``write-cache`` option of block guest devices (as in
1388 ``-device``). The modes correspond to the following settings:
1390 ============= =============== ============ ==============
1391 \ cache.writeback cache.direct cache.no-flush
1392 ============= =============== ============ ==============
1393 writeback on off off
1395 writethrough off off off
1396 directsync off on off
1398 ============= =============== ============ ==============
1400 The default mode is ``cache=writeback``.
1403 aio is "threads", "native", or "io_uring" and selects between pthread
1404 based disk I/O, native Linux AIO, or Linux io_uring API.
1407 Specify which disk format will be used rather than detecting the
1408 format. Can be used to specify format=raw to avoid interpreting
1409 an untrusted format header.
1411 ``werror=action,rerror=action``
1412 Specify which action to take on write and read errors. Valid
1413 actions are: "ignore" (ignore the error and try to continue),
1414 "stop" (pause QEMU), "report" (report the error to the guest),
1415 "enospc" (pause QEMU only if the host disk is full; report the
1416 error to the guest otherwise). The default setting is
1417 ``werror=enospc`` and ``rerror=report``.
1419 ``copy-on-read=copy-on-read``
1420 copy-on-read is "on" or "off" and enables whether to copy read
1421 backing file sectors into the image file.
1423 ``bps=b,bps_rd=r,bps_wr=w``
1424 Specify bandwidth throttling limits in bytes per second, either
1425 for all request types or for reads or writes only. Small values
1426 can lead to timeouts or hangs inside the guest. A safe minimum
1427 for disks is 2 MB/s.
1429 ``bps_max=bm,bps_rd_max=rm,bps_wr_max=wm``
1430 Specify bursts in bytes per second, either for all request types
1431 or for reads or writes only. Bursts allow the guest I/O to spike
1432 above the limit temporarily.
1434 ``iops=i,iops_rd=r,iops_wr=w``
1435 Specify request rate limits in requests per second, either for
1436 all request types or for reads or writes only.
1438 ``iops_max=bm,iops_rd_max=rm,iops_wr_max=wm``
1439 Specify bursts in requests per second, either for all request
1440 types or for reads or writes only. Bursts allow the guest I/O to
1441 spike above the limit temporarily.
1444 Let every is bytes of a request count as a new request for iops
1445 throttling purposes. Use this option to prevent guests from
1446 circumventing iops limits by sending fewer but larger requests.
1449 Join a throttling quota group with given name g. All drives that
1450 are members of the same group are accounted for together. Use
1451 this option to prevent guests from circumventing throttling
1452 limits by using many small disks instead of a single larger
1455 By default, the ``cache.writeback=on`` mode is used. It will report
1456 data writes as completed as soon as the data is present in the host
1457 page cache. This is safe as long as your guest OS makes sure to
1458 correctly flush disk caches where needed. If your guest OS does not
1459 handle volatile disk write caches correctly and your host crashes or
1460 loses power, then the guest may experience data corruption.
1462 For such guests, you should consider using ``cache.writeback=off``.
1463 This means that the host page cache will be used to read and write
1464 data, but write notification will be sent to the guest only after
1465 QEMU has made sure to flush each write to the disk. Be aware that
1466 this has a major impact on performance.
1468 When using the ``-snapshot`` option, unsafe caching is always used.
1470 Copy-on-read avoids accessing the same backing file sectors
1471 repeatedly and is useful when the backing file is over a slow
1472 network. By default copy-on-read is off.
1474 Instead of ``-cdrom`` you can use:
1478 |qemu_system| -drive file=file,index=2,media=cdrom
1480 Instead of ``-hda``, ``-hdb``, ``-hdc``, ``-hdd``, you can use:
1484 |qemu_system| -drive file=file,index=0,media=disk
1485 |qemu_system| -drive file=file,index=1,media=disk
1486 |qemu_system| -drive file=file,index=2,media=disk
1487 |qemu_system| -drive file=file,index=3,media=disk
1489 You can open an image using pre-opened file descriptors from an fd
1495 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
1496 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
1497 -drive file=/dev/fdset/2,index=0,media=disk
1499 You can connect a CDROM to the slave of ide0:
1503 |qemu_system_x86| -drive file=file,if=ide,index=1,media=cdrom
1505 If you don't specify the
"file=" argument
, you define an empty
1510 |qemu_system_x86|
-drive
if=ide
,index
=1,media
=cdrom
1512 Instead of ``
-fda``
, ``
-fdb``
, you can use
:
1516 |qemu_system_x86|
-drive file
=file
,index
=0,if=floppy
1517 |qemu_system_x86|
-drive file
=file
,index
=1,if=floppy
1519 By
default, interface is
"ide" and index is automatically
1524 |qemu_system_x86|
-drive file
=a
-drive file
=b
"
1526 is interpreted like:
1530 |qemu_system_x86| -hda a -hdb b
1533 DEF("mtdblock
", HAS_ARG, QEMU_OPTION_mtdblock,
1534 "-mtdblock file use
'file' as on
-board Flash memory image
\n",
1538 Use file as on-board Flash memory image.
1541 DEF("sd
", HAS_ARG, QEMU_OPTION_sd,
1542 "-sd file use
'file' as SecureDigital card image
\n", QEMU_ARCH_ALL)
1545 Use file as SecureDigital card image.
1548 DEF("pflash
", HAS_ARG, QEMU_OPTION_pflash,
1549 "-pflash file use
'file' as a parallel flash image
\n", QEMU_ARCH_ALL)
1552 Use file as a parallel flash image.
1555 DEF("snapshot
", 0, QEMU_OPTION_snapshot,
1556 "-snapshot write to temporary files instead of disk image files
\n",
1560 Write to temporary files instead of disk image files. In this case,
1561 the raw disk image you use is not written back. You can however
1562 force the write back by pressing C-a s (see the :ref:`disk images`
1563 chapter in the System Emulation Users Guide).
1566 DEF("fsdev
", HAS_ARG, QEMU_OPTION_fsdev,
1567 "-fsdev local
,id
=id
,path
=path
,security_model
=mapped
-xattr|mapped
-file|passthrough|none
\n"
1568 " [,writeout
=immediate
][,readonly
=on
][,fmode
=fmode
][,dmode
=dmode
]\n"
1569 " [[,throttling
.bps
-total
=b
]|
[[,throttling
.bps
-read
=r
][,throttling
.bps
-write
=w
]]]\n"
1570 " [[,throttling
.iops
-total
=i
]|
[[,throttling
.iops
-read
=r
][,throttling
.iops
-write
=w
]]]\n"
1571 " [[,throttling
.bps
-total
-max
=bm
]|
[[,throttling
.bps
-read
-max
=rm
][,throttling
.bps
-write
-max
=wm
]]]\n"
1572 " [[,throttling
.iops
-total
-max
=im
]|
[[,throttling
.iops
-read
-max
=irm
][,throttling
.iops
-write
-max
=iwm
]]]\n"
1573 " [[,throttling
.iops
-size
=is
]]\n"
1574 "-fsdev proxy
,id
=id
,socket
=socket
[,writeout
=immediate
][,readonly
=on
]\n"
1575 "-fsdev proxy
,id
=id
,sock_fd
=sock_fd
[,writeout
=immediate
][,readonly
=on
]\n"
1576 "-fsdev synth
,id
=id
\n",
1580 ``-fsdev local,id=id,path=path,security_model=security_model [,writeout=writeout][,readonly=on][,fmode=fmode][,dmode=dmode] [,throttling.option=value[,throttling.option=value[,...]]]``
1582 ``-fsdev proxy,id=id,socket=socket[,writeout=writeout][,readonly=on]``
1584 ``-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=writeout][,readonly=on]``
1586 ``-fsdev synth,id=id[,readonly=on]``
1587 Define a new file system device. Valid options are:
1590 Accesses to the filesystem are done by QEMU.
1593 Accesses to the filesystem are done by virtfs-proxy-helper(1).
1596 Synthetic filesystem, only used by QTests.
1599 Specifies identifier for this device.
1602 Specifies the export path for the file system device. Files
1603 under this path will be available to the 9p client on the guest.
1605 ``security_model=security_model``
1606 Specifies the security model to be used for this export path.
1607 Supported security models are "passthrough
", "mapped
-xattr
",
1608 "mapped
-file
" and "none
". In "passthrough
" security model, files
1609 are stored using the same credentials as they are created on the
1610 guest. This requires QEMU to run as root. In "mapped
-xattr
"
1611 security model, some of the file attributes like uid, gid, mode
1612 bits and link target are stored as file attributes. For
1613 "mapped
-file
" these attributes are stored in the hidden
1614 .virtfs\_metadata directory. Directories exported by this
1615 security model cannot interact with other unix tools. "none
"
1616 security model is same as passthrough except the sever won't
1617 report failures if it fails to set file attributes like
1618 ownership. Security model is mandatory only for local fsdriver.
1619 Other fsdrivers (like proxy) don't take security model as a
1622 ``writeout=writeout``
1623 This is an optional argument. The only supported value is
1624 "immediate
". This means that host page cache will be used to
1625 read and write data but write notification will be sent to the
1626 guest only when the data has been reported as written by the
1630 Enables exporting 9p share as a readonly mount for guests. By
1631 default read-write access is given.
1634 Enables proxy filesystem driver to use passed socket file for
1635 communicating with virtfs-proxy-helper(1).
1638 Enables proxy filesystem driver to use passed socket descriptor
1639 for communicating with virtfs-proxy-helper(1). Usually a helper
1640 like libvirt will create socketpair and pass one of the fds as
1644 Specifies the default mode for newly created files on the host.
1645 Works only with security models "mapped
-xattr
" and
1649 Specifies the default mode for newly created directories on the
1650 host. Works only with security models "mapped
-xattr
" and
1653 ``throttling.bps-total=b,throttling.bps-read=r,throttling.bps-write=w``
1654 Specify bandwidth throttling limits in bytes per second, either
1655 for all request types or for reads or writes only.
1657 ``throttling.bps-total-max=bm,bps-read-max=rm,bps-write-max=wm``
1658 Specify bursts in bytes per second, either for all request types
1659 or for reads or writes only. Bursts allow the guest I/O to spike
1660 above the limit temporarily.
1662 ``throttling.iops-total=i,throttling.iops-read=r, throttling.iops-write=w``
1663 Specify request rate limits in requests per second, either for
1664 all request types or for reads or writes only.
1666 ``throttling.iops-total-max=im,throttling.iops-read-max=irm, throttling.iops-write-max=iwm``
1667 Specify bursts in requests per second, either for all request
1668 types or for reads or writes only. Bursts allow the guest I/O to
1669 spike above the limit temporarily.
1671 ``throttling.iops-size=is``
1672 Let every is bytes of a request count as a new request for iops
1673 throttling purposes.
1675 -fsdev option is used along with -device driver "virtio
-9p
-...".
1677 ``-device virtio-9p-type,fsdev=id,mount_tag=mount_tag``
1678 Options for virtio-9p-... driver are:
1681 Specifies the variant to be used. Supported values are "pci
",
1682 "ccw
" or "device
", depending on the machine type.
1685 Specifies the id value specified along with -fsdev option.
1687 ``mount_tag=mount_tag``
1688 Specifies the tag name to be used by the guest to mount this
1692 DEF("virtfs
", HAS_ARG, QEMU_OPTION_virtfs,
1693 "-virtfs local
,path
=path
,mount_tag
=tag
,security_model
=mapped
-xattr|mapped
-file|passthrough|none
\n"
1694 " [,id
=id
][,writeout
=immediate
][,readonly
=on
][,fmode
=fmode
][,dmode
=dmode
][,multidevs
=remap|forbid|warn
]\n"
1695 "-virtfs proxy
,mount_tag
=tag
,socket
=socket
[,id
=id
][,writeout
=immediate
][,readonly
=on
]\n"
1696 "-virtfs proxy
,mount_tag
=tag
,sock_fd
=sock_fd
[,id
=id
][,writeout
=immediate
][,readonly
=on
]\n"
1697 "-virtfs synth
,mount_tag
=tag
[,id
=id
][,readonly
=on
]\n",
1701 ``-virtfs local,path=path,mount_tag=mount_tag ,security_model=security_model[,writeout=writeout][,readonly=on] [,fmode=fmode][,dmode=dmode][,multidevs=multidevs]``
1703 ``-virtfs proxy,socket=socket,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1705 ``-virtfs proxy,sock_fd=sock_fd,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1707 ``-virtfs synth,mount_tag=mount_tag``
1708 Define a new virtual filesystem device and expose it to the guest using
1709 a virtio-9p-device (a.k.a. 9pfs), which essentially means that a certain
1710 directory on host is made directly accessible by guest as a pass-through
1711 file system by using the 9P network protocol for communication between
1712 host and guests, if desired even accessible, shared by several guests
1715 Note that ``-virtfs`` is actually just a convenience shortcut for its
1716 generalized form ``-fsdev -device virtio-9p-pci``.
1718 The general form of pass-through file system options are:
1721 Accesses to the filesystem are done by QEMU.
1724 Accesses to the filesystem are done by virtfs-proxy-helper(1).
1727 Synthetic filesystem, only used by QTests.
1730 Specifies identifier for the filesystem device
1733 Specifies the export path for the file system device. Files
1734 under this path will be available to the 9p client on the guest.
1736 ``security_model=security_model``
1737 Specifies the security model to be used for this export path.
1738 Supported security models are "passthrough
", "mapped
-xattr
",
1739 "mapped
-file
" and "none
". In "passthrough
" security model, files
1740 are stored using the same credentials as they are created on the
1741 guest. This requires QEMU to run as root. In "mapped
-xattr
"
1742 security model, some of the file attributes like uid, gid, mode
1743 bits and link target are stored as file attributes. For
1744 "mapped
-file
" these attributes are stored in the hidden
1745 .virtfs\_metadata directory. Directories exported by this
1746 security model cannot interact with other unix tools. "none
"
1747 security model is same as passthrough except the sever won't
1748 report failures if it fails to set file attributes like
1749 ownership. Security model is mandatory only for local fsdriver.
1750 Other fsdrivers (like proxy) don't take security model as a
1753 ``writeout=writeout``
1754 This is an optional argument. The only supported value is
1755 "immediate
". This means that host page cache will be used to
1756 read and write data but write notification will be sent to the
1757 guest only when the data has been reported as written by the
1761 Enables exporting 9p share as a readonly mount for guests. By
1762 default read-write access is given.
1765 Enables proxy filesystem driver to use passed socket file for
1766 communicating with virtfs-proxy-helper(1). Usually a helper like
1767 libvirt will create socketpair and pass one of the fds as
1771 Enables proxy filesystem driver to use passed 'sock\_fd' as the
1772 socket descriptor for interfacing with virtfs-proxy-helper(1).
1775 Specifies the default mode for newly created files on the host.
1776 Works only with security models "mapped
-xattr
" and
1780 Specifies the default mode for newly created directories on the
1781 host. Works only with security models "mapped
-xattr
" and
1784 ``mount_tag=mount_tag``
1785 Specifies the tag name to be used by the guest to mount this
1788 ``multidevs=multidevs``
1789 Specifies how to deal with multiple devices being shared with a
1790 9p export. Supported behaviours are either "remap
", "forbid
" or
1791 "warn
". The latter is the default behaviour on which virtfs 9p
1792 expects only one device to be shared with the same export, and
1793 if more than one device is shared and accessed via the same 9p
1794 export then only a warning message is logged (once) by qemu on
1795 host side. In order to avoid file ID collisions on guest you
1796 should either create a separate virtfs export for each device to
1797 be shared with guests (recommended way) or you might use "remap
"
1798 instead which allows you to share multiple devices with only one
1799 export instead, which is achieved by remapping the original
1800 inode numbers from host to guest in a way that would prevent
1801 such collisions. Remapping inodes in such use cases is required
1802 because the original device IDs from host are never passed and
1803 exposed on guest. Instead all files of an export shared with
1804 virtfs always share the same device id on guest. So two files
1805 with identical inode numbers but from actually different devices
1806 on host would otherwise cause a file ID collision and hence
1807 potential misbehaviours on guest. "forbid
" on the other hand
1808 assumes like "warn
" that only one device is shared by the same
1809 export, however it will not only log a warning message but also
1810 deny access to additional devices on guest. Note though that
1811 "forbid
" does currently not block all possible file access
1812 operations (e.g. readdir() would still return entries from other
1816 DEF("iscsi
", HAS_ARG, QEMU_OPTION_iscsi,
1817 "-iscsi
[user
=user
][,password
=password
]\n"
1818 " [,header
-digest
=CRC32C|CR32C
-NONE|NONE
-CRC32C|NONE
\n"
1819 " [,initiator
-name
=initiator
-iqn
][,id
=target
-iqn
]\n"
1820 " [,timeout
=timeout
]\n"
1821 " iSCSI session parameters
\n", QEMU_ARCH_ALL)
1825 Configure iSCSI session parameters.
1830 DEFHEADING(USB convenience options:)
1832 DEF("usb
", 0, QEMU_OPTION_usb,
1833 "-usb enable on
-board USB host
controller (if not enabled by
default)\n",
1837 Enable USB emulation on machine types with an on-board USB host
1838 controller (if not enabled by default). Note that on-board USB host
1839 controllers may not support USB 3.0. In this case
1840 ``-device qemu-xhci`` can be used instead on machines with PCI.
1843 DEF("usbdevice
", HAS_ARG, QEMU_OPTION_usbdevice,
1844 "-usbdevice name add the host or guest USB device
'name'\n",
1847 ``-usbdevice devname``
1848 Add the USB device devname, and enable an on-board USB controller
1849 if possible and necessary (just like it can be done via
1850 ``-machine usb=on``). Note that this option is mainly intended for
1851 the user's convenience only. More fine-grained control can be
1852 achieved by selecting a USB host controller (if necessary) and the
1853 desired USB device via the ``-device`` option instead. For example,
1854 instead of using ``-usbdevice mouse`` it is possible to use
1855 ``-device qemu-xhci -device usb-mouse`` to connect the USB mouse
1856 to a USB 3.0 controller instead (at least on machines that support
1857 PCI and do not have an USB controller enabled by default yet).
1858 For more details, see the chapter about
1859 :ref:`Connecting USB devices` in the System Emulation Users Guide.
1860 Possible devices for devname are:
1863 Braille device. This will use BrlAPI to display the braille
1864 output on a real or fake device (i.e. it also creates a
1865 corresponding ``braille`` chardev automatically beside the
1866 ``usb-braille`` USB device).
1869 Standard USB keyboard. Will override the PS/2 keyboard (if present).
1872 Virtual Mouse. This will override the PS/2 mouse emulation when
1876 Pointer device that uses absolute coordinates (like a
1877 touchscreen). This means QEMU is able to report the mouse
1878 position without having to grab the mouse. Also overrides the
1879 PS/2 mouse emulation when activated.
1882 Wacom PenPartner USB tablet.
1889 DEFHEADING(Display options:)
1891 DEF("display
", HAS_ARG, QEMU_OPTION_display,
1892 #if defined(CONFIG_SPICE)
1893 "-display spice
-app
[,gl
=on|off
]\n"
1895 #if defined(CONFIG_SDL)
1896 "-display sdl
[,alt_grab
=on|off
][,ctrl_grab
=on|off
][,gl
=on|core|es|off
]\n"
1897 " [,grab
-mod
=<mod
>][,show
-cursor
=on|off
][,window
-close
=on|off
]\n"
1899 #if defined(CONFIG_GTK)
1900 "-display gtk
[,full
-screen
=on|off
][,gl
=on|off
][,grab
-on
-hover
=on|off
]\n"
1901 " [,show
-cursor
=on|off
][,window
-close
=on|off
]\n"
1903 #if defined(CONFIG_VNC)
1904 "-display vnc
=<display
>[,<optargs
>]\n"
1906 #if defined(CONFIG_CURSES)
1907 "-display curses
[,charset
=<encoding
>]\n"
1909 #if defined(CONFIG_OPENGL)
1910 "-display egl
-headless
[,rendernode
=<file
>]\n"
1912 #if defined(CONFIG_DBUS_DISPLAY)
1913 "-display dbus
[,addr
=<dbusaddr
>]\n"
1914 " [,gl
=on|core|es|off
][,rendernode
=<file
>]\n"
1917 " select display backend type
\n"
1918 " The
default display is equivalent to
\n "
1919 #if defined(CONFIG_GTK)
1920 "\"-display gtk
\"\n"
1921 #elif defined(CONFIG_SDL)
1922 "\"-display sdl
\"\n"
1923 #elif defined(CONFIG_COCOA)
1924 "\"-display cocoa
\"\n"
1925 #elif defined(CONFIG_VNC)
1926 "\"-vnc localhost
:0,to
=99,id
=default\"\n"
1928 "\"-display none
\"\n"
1933 Select type of display to use. This option is a replacement for the
1934 old style -sdl/-curses/... options. Use ``-display help`` to list
1935 the available display types. Valid values for type are
1937 ``spice-app[,gl=on|off]``
1938 Start QEMU as a Spice server and launch the default Spice client
1939 application. The Spice server will redirect the serial consoles
1940 and QEMU monitors. (Since 4.0)
1943 Export the display over D-Bus interfaces. (Since 7.0)
1945 The connection is registered with the "org
.qemu
" name (and queued when
1948 ``addr=<dbusaddr>`` : D-Bus bus address to connect to.
1950 ``p2p=yes|no`` : Use peer-to-peer connection, accepted via QMP ``add_client``.
1952 ``gl=on|off|core|es`` : Use OpenGL for rendering (the D-Bus interface
1953 will share framebuffers with DMABUF file descriptors).
1956 Display video output via SDL (usually in a separate graphics
1957 window; see the SDL documentation for other possibilities).
1958 Valid parameters are:
1960 ``grab-mod=<mods>`` : Used to select the modifier keys for toggling
1961 the mouse grabbing in conjunction with the "g
" key. ``<mods>`` can be
1962 either ``lshift-lctrl-lalt`` or ``rctrl``.
1964 ``alt_grab=on|off`` : Use Control+Alt+Shift-g to toggle mouse grabbing.
1965 This parameter is deprecated - use ``grab-mod`` instead.
1967 ``ctrl_grab=on|off`` : Use Right-Control-g to toggle mouse grabbing.
1968 This parameter is deprecated - use ``grab-mod`` instead.
1970 ``gl=on|off|core|es`` : Use OpenGL for displaying
1972 ``show-cursor=on|off`` : Force showing the mouse cursor
1974 ``window-close=on|off`` : Allow to quit qemu with window close button
1977 Display video output in a GTK window. This interface provides
1978 drop-down menus and other UI elements to configure and control
1979 the VM during runtime. Valid parameters are:
1981 ``full-screen=on|off`` : Start in fullscreen mode
1983 ``gl=on|off`` : Use OpenGL for displaying
1985 ``grab-on-hover=on|off`` : Grab keyboard input on mouse hover
1987 ``show-cursor=on|off`` : Force showing the mouse cursor
1989 ``window-close=on|off`` : Allow to quit qemu with window close button
1991 ``curses[,charset=<encoding>]``
1992 Display video output via curses. For graphics device models
1993 which support a text mode, QEMU can display this output using a
1994 curses/ncurses interface. Nothing is displayed when the graphics
1995 device is in graphical mode or if the graphics device does not
1996 support a text mode. Generally only the VGA device models
1997 support text mode. The font charset used by the guest can be
1998 specified with the ``charset`` option, for example
1999 ``charset=CP850`` for IBM CP850 encoding. The default is
2002 ``egl-headless[,rendernode=<file>]``
2003 Offload all OpenGL operations to a local DRI device. For any
2004 graphical display, this display needs to be paired with either
2005 VNC or SPICE displays.
2008 Start a VNC server on display <display>
2011 Do not display video output. The guest will still see an
2012 emulated graphics card, but its output will not be displayed to
2013 the QEMU user. This option differs from the -nographic option in
2014 that it only affects what is done with video output; -nographic
2015 also changes the destination of the serial and parallel port
2019 DEF("nographic
", 0, QEMU_OPTION_nographic,
2020 "-nographic disable graphical output and redirect serial I
/Os to console
\n",
2024 Normally, if QEMU is compiled with graphical window support, it
2025 displays output such as guest graphics, guest console, and the QEMU
2026 monitor in a window. With this option, you can totally disable
2027 graphical output so that QEMU is a simple command line application.
2028 The emulated serial port is redirected on the console and muxed with
2029 the monitor (unless redirected elsewhere explicitly). Therefore, you
2030 can still use QEMU to debug a Linux kernel with a serial console.
2031 Use C-a h for help on switching between the console and monitor.
2034 DEF("curses
", 0, QEMU_OPTION_curses,
2035 "-curses shorthand
for -display curses
\n",
2039 Normally, if QEMU is compiled with graphical window support, it
2040 displays output such as guest graphics, guest console, and the QEMU
2041 monitor in a window. With this option, QEMU can display the VGA
2042 output when in text mode using a curses/ncurses interface. Nothing
2043 is displayed in graphical mode.
2046 DEF("alt
-grab
", 0, QEMU_OPTION_alt_grab,
2047 "-alt
-grab use Ctrl
-Alt
-Shift to grab
mouse (instead of Ctrl
-Alt
)\n",
2051 Use Ctrl-Alt-Shift to grab mouse (instead of Ctrl-Alt). Note that
2052 this also affects the special keys (for fullscreen, monitor-mode
2053 switching, etc). This option is deprecated - please use
2054 ``-display sdl,grab-mod=lshift-lctrl-lalt`` instead.
2057 DEF("ctrl
-grab
", 0, QEMU_OPTION_ctrl_grab,
2058 "-ctrl
-grab use Right
-Ctrl to grab
mouse (instead of Ctrl
-Alt
)\n",
2062 Use Right-Ctrl to grab mouse (instead of Ctrl-Alt). Note that this
2063 also affects the special keys (for fullscreen, monitor-mode
2064 switching, etc). This option is deprecated - please use
2065 ``-display sdl,grab-mod=rctrl`` instead.
2068 DEF("no
-quit
", 0, QEMU_OPTION_no_quit,
2069 "-no
-quit disable SDL
/GTK window close
capability (deprecated
)\n", QEMU_ARCH_ALL)
2072 Disable window close capability (SDL and GTK only). This option is
2073 deprecated, please use ``-display ...,window-close=off`` instead.
2076 DEF("sdl
", 0, QEMU_OPTION_sdl,
2077 "-sdl shorthand
for -display sdl
\n", QEMU_ARCH_ALL)
2083 DEF("spice
", HAS_ARG, QEMU_OPTION_spice,
2084 "-spice
[port
=port
][,tls
-port
=secured
-port
][,x509
-dir
=<dir
>]\n"
2085 " [,x509
-key
-file
=<file
>][,x509
-key
-password
=<file
>]\n"
2086 " [,x509
-cert
-file
=<file
>][,x509
-cacert
-file
=<file
>]\n"
2087 " [,x509
-dh
-key
-file
=<file
>][,addr
=addr
]\n"
2088 " [,ipv4
=on|off
][,ipv6
=on|off
][,unix
=on|off
]\n"
2089 " [,tls
-ciphers
=<list
>]\n"
2090 " [,tls
-channel
=[main|display|cursor|inputs|record|playback
]]\n"
2091 " [,plaintext
-channel
=[main|display|cursor|inputs|record|playback
]]\n"
2092 " [,sasl
=on|off
][,disable
-ticketing
=on|off
]\n"
2093 " [,password
=<string
>][,password
-secret
=<secret
-id
>]\n"
2094 " [,image
-compression
=[auto_glz|auto_lz|quic|glz|lz|off
]]\n"
2095 " [,jpeg
-wan
-compression
=[auto|
never|always
]]\n"
2096 " [,zlib
-glz
-wan
-compression
=[auto|
never|always
]]\n"
2097 " [,streaming
-video
=[off|all|filter
]][,disable
-copy
-paste
=on|off
]\n"
2098 " [,disable
-agent
-file
-xfer
=on|off
][,agent
-mouse
=[on|off
]]\n"
2099 " [,playback
-compression
=[on|off
]][,seamless
-migration
=[on|off
]]\n"
2100 " [,gl
=[on|off
]][,rendernode
=<file
>]\n"
2102 " at least one of
{port
, tls
-port
} is mandatory
\n",
2105 ``-spice option[,option[,...]]``
2106 Enable the spice remote desktop protocol. Valid options are
2109 Set the TCP port spice is listening on for plaintext channels.
2112 Set the IP address spice is listening on. Default is any
2115 ``ipv4=on|off``; \ ``ipv6=on|off``; \ ``unix=on|off``
2116 Force using the specified IP version.
2118 ``password=<string>``
2119 Set the password you need to authenticate.
2121 This option is deprecated and insecure because it leaves the
2122 password visible in the process listing. Use ``password-secret``
2125 ``password-secret=<secret-id>``
2126 Set the ID of the ``secret`` object containing the password
2127 you need to authenticate.
2130 Require that the client use SASL to authenticate with the spice.
2131 The exact choice of authentication method used is controlled
2132 from the system / user's SASL configuration file for the 'qemu'
2133 service. This is typically found in /etc/sasl2/qemu.conf. If
2134 running QEMU as an unprivileged user, an environment variable
2135 SASL\_CONF\_PATH can be used to make it search alternate
2136 locations for the service config. While some SASL auth methods
2137 can also provide data encryption (eg GSSAPI), it is recommended
2138 that SASL always be combined with the 'tls' and 'x509' settings
2139 to enable use of SSL and server certificates. This ensures a
2140 data encryption preventing compromise of authentication
2143 ``disable-ticketing=on|off``
2144 Allow client connects without authentication.
2146 ``disable-copy-paste=on|off``
2147 Disable copy paste between the client and the guest.
2149 ``disable-agent-file-xfer=on|off``
2150 Disable spice-vdagent based file-xfer between the client and the
2154 Set the TCP port spice is listening on for encrypted channels.
2157 Set the x509 file directory. Expects same filenames as -vnc
2160 ``x509-key-file=<file>``; \ ``x509-key-password=<file>``; \ ``x509-cert-file=<file>``; \ ``x509-cacert-file=<file>``; \ ``x509-dh-key-file=<file>``
2161 The x509 file names can also be configured individually.
2163 ``tls-ciphers=<list>``
2164 Specify which ciphers to use.
2166 ``tls-channel=[main|display|cursor|inputs|record|playback]``; \ ``plaintext-channel=[main|display|cursor|inputs|record|playback]``
2167 Force specific channel to be used with or without TLS
2168 encryption. The options can be specified multiple times to
2169 configure multiple channels. The special name "default" can be
2170 used to set the default mode. For channels which are not
2171 explicitly forced into one mode the spice client is allowed to
2172 pick tls/plaintext as he pleases.
2174 ``image-compression=[auto_glz|auto_lz|quic|glz|lz|off]``
2175 Configure image compression (lossless). Default is auto\_glz.
2177 ``jpeg-wan-compression=[auto|never|always]``; \ ``zlib-glz-wan-compression=[auto|never|always]``
2178 Configure wan image compression (lossy for slow links). Default
2181 ``streaming-video=[off|all|filter]``
2182 Configure video stream detection. Default is off.
2184 ``agent-mouse=[on|off]``
2185 Enable/disable passing mouse events via vdagent. Default is on.
2187 ``playback-compression=[on|off]``
2188 Enable/disable audio stream compression (using celt 0.5.1).
2191 ``seamless-migration=[on|off]``
2192 Enable/disable spice seamless migration. Default is off.
2195 Enable/disable OpenGL context. Default is off.
2197 ``rendernode=<file>``
2198 DRM render node for OpenGL rendering. If not specified, it will
2199 pick the first available. (Since 2.9)
2202 DEF("portrait
", 0, QEMU_OPTION_portrait,
2203 "-portrait rotate graphical output
90 deg
left (only PXA LCD
)\n",
2207 Rotate graphical output 90 deg left (only PXA LCD).
2210 DEF("rotate
", HAS_ARG, QEMU_OPTION_rotate,
2211 "-rotate
<deg
> rotate graphical output some deg
left (only PXA LCD
)\n",
2215 Rotate graphical output some deg left (only PXA LCD).
2218 DEF("vga
", HAS_ARG, QEMU_OPTION_vga,
2219 "-vga
[std|cirrus|vmware|qxl|xenfb|tcx|cg3|virtio|none
]\n"
2220 " select video card type
\n", QEMU_ARCH_ALL)
2223 Select type of VGA card to emulate. Valid values for type are
2226 Cirrus Logic GD5446 Video card. All Windows versions starting
2227 from Windows 95 should recognize and use this graphic card. For
2228 optimal performances, use 16 bit color depth in the guest and
2229 the host OS. (This card was the default before QEMU 2.2)
2232 Standard VGA card with Bochs VBE extensions. If your guest OS
2233 supports the VESA 2.0 VBE extensions (e.g. Windows XP) and if
2234 you want to use high resolution modes (>= 1280x1024x16) then you
2235 should use this option. (This card is the default since QEMU
2239 VMWare SVGA-II compatible adapter. Use it if you have
2240 sufficiently recent XFree86/XOrg server or Windows guest with a
2241 driver for this card.
2244 QXL paravirtual graphic card. It is VGA compatible (including
2245 VESA 2.0 VBE support). Works best with qxl guest drivers
2246 installed though. Recommended choice when using the spice
2250 (sun4m only) Sun TCX framebuffer. This is the default
2251 framebuffer for sun4m machines and offers both 8-bit and 24-bit
2252 colour depths at a fixed resolution of 1024x768.
2255 (sun4m only) Sun cgthree framebuffer. This is a simple 8-bit
2256 framebuffer for sun4m machines available in both 1024x768
2257 (OpenBIOS) and 1152x900 (OBP) resolutions aimed at people
2258 wishing to run older Solaris versions.
2267 DEF("full
-screen
", 0, QEMU_OPTION_full_screen,
2268 "-full
-screen start
in full screen
\n", QEMU_ARCH_ALL)
2271 Start in full screen.
2274 DEF("g
", HAS_ARG, QEMU_OPTION_g ,
2275 "-g WxH
[xDEPTH
] Set the initial graphical resolution and depth
\n",
2276 QEMU_ARCH_PPC | QEMU_ARCH_SPARC | QEMU_ARCH_M68K)
2278 ``-g`` *width*\ ``x``\ *height*\ ``[x``\ *depth*\ ``]``
2279 Set the initial graphical resolution and depth (PPC, SPARC only).
2281 For PPC the default is 800x600x32.
2283 For SPARC with the TCX graphics device, the default is 1024x768x8
2284 with the option of 1024x768x24. For cgthree, the default is
2285 1024x768x8 with the option of 1152x900x8 for people who wish to use
2289 DEF("vnc
", HAS_ARG, QEMU_OPTION_vnc ,
2290 "-vnc
<display
> shorthand
for -display vnc
=<display
>\n", QEMU_ARCH_ALL)
2292 ``-vnc display[,option[,option[,...]]]``
2293 Normally, if QEMU is compiled with graphical window support, it
2294 displays output such as guest graphics, guest console, and the QEMU
2295 monitor in a window. With this option, you can have QEMU listen on
2296 VNC display display and redirect the VGA display over the VNC
2297 session. It is very useful to enable the usb tablet device when
2298 using this option (option ``-device usb-tablet``). When using the
2299 VNC display, you must use the ``-k`` parameter to set the keyboard
2300 layout if you are not using en-us. Valid syntax for the display is
2303 With this option, QEMU will try next available VNC displays,
2304 until the number L, if the origianlly defined "-vnc display
" is
2305 not available, e.g. port 5900+display is already used by another
2306 application. By default, to=0.
2309 TCP connections will only be allowed from host on display d. By
2310 convention the TCP port is 5900+d. Optionally, host can be
2311 omitted in which case the server will accept connections from
2315 Connections will be allowed over UNIX domain sockets where path
2316 is the location of a unix socket to listen for connections on.
2319 VNC is initialized but not started. The monitor ``change``
2320 command can be used to later start the VNC server.
2322 Following the display value there may be one or more option flags
2323 separated by commas. Valid options are
2326 Connect to a listening VNC client via a "reverse
" connection.
2327 The client is specified by the display. For reverse network
2328 connections (host:d,``reverse``), the d argument is a TCP port
2329 number, not a display number.
2331 ``websocket=on|off``
2332 Opens an additional TCP listening port dedicated to VNC
2333 Websocket connections. If a bare websocket option is given, the
2334 Websocket port is 5700+display. An alternative port can be
2335 specified with the syntax ``websocket``\ =port.
2337 If host is specified connections will only be allowed from this
2338 host. It is possible to control the websocket listen address
2339 independently, using the syntax ``websocket``\ =host:port.
2341 If no TLS credentials are provided, the websocket connection
2342 runs in unencrypted mode. If TLS credentials are provided, the
2343 websocket connection requires encrypted client connections.
2346 Require that password based authentication is used for client
2349 The password must be set separately using the ``set_password``
2350 command in the :ref:`QEMU monitor`. The
2351 syntax to change your password is:
2352 ``set_password <protocol> <password>`` where <protocol> could be
2353 either "vnc
" or "spice
".
2355 If you would like to change <protocol> password expiration, you
2356 should use ``expire_password <protocol> <expiration-time>``
2357 where expiration time could be one of the following options:
2358 now, never, +seconds or UNIX time of expiration, e.g. +60 to
2359 make password expire in 60 seconds, or 1335196800 to make
2360 password expire on "Mon Apr
23 12:00:00 EDT
2012" (UNIX time for
2361 this date and time).
2363 You can also use keywords "now
" or "never" for the expiration
2364 time to allow <protocol> password to expire immediately or never
2367 ``password-secret=<secret-id>``
2368 Require that password based authentication is used for client
2369 connections, using the password provided by the ``secret``
2370 object identified by ``secret-id``.
2373 Provides the ID of a set of TLS credentials to use to secure the
2374 VNC server. They will apply to both the normal VNC server socket
2375 and the websocket socket (if enabled). Setting TLS credentials
2376 will cause the VNC server socket to enable the VeNCrypt auth
2377 mechanism. The credentials should have been previously created
2378 using the ``-object tls-creds`` argument.
2381 Provides the ID of the QAuthZ authorization object against which
2382 the client's x509 distinguished name will validated. This object
2383 is only resolved at time of use, so can be deleted and recreated
2384 on the fly while the VNC server is active. If missing, it will
2385 default to denying access.
2388 Require that the client use SASL to authenticate with the VNC
2389 server. The exact choice of authentication method used is
2390 controlled from the system / user's SASL configuration file for
2391 the 'qemu' service. This is typically found in
2392 /etc/sasl2/qemu.conf. If running QEMU as an unprivileged user,
2393 an environment variable SASL\_CONF\_PATH can be used to make it
2394 search alternate locations for the service config. While some
2395 SASL auth methods can also provide data encryption (eg GSSAPI),
2396 it is recommended that SASL always be combined with the 'tls'
2397 and 'x509' settings to enable use of SSL and server
2398 certificates. This ensures a data encryption preventing
2399 compromise of authentication credentials. See the
2400 :ref:`VNC security` section in the System Emulation Users Guide
2401 for details on using SASL authentication.
2404 Provides the ID of the QAuthZ authorization object against which
2405 the client's SASL username will validated. This object is only
2406 resolved at time of use, so can be deleted and recreated on the
2407 fly while the VNC server is active. If missing, it will default
2411 Legacy method for enabling authorization of clients against the
2412 x509 distinguished name and SASL username. It results in the
2413 creation of two ``authz-list`` objects with IDs of
2414 ``vnc.username`` and ``vnc.x509dname``. The rules for these
2415 objects must be configured with the HMP ACL commands.
2417 This option is deprecated and should no longer be used. The new
2418 ``sasl-authz`` and ``tls-authz`` options are a replacement.
2421 Enable lossy compression methods (gradient, JPEG, ...). If this
2422 option is set, VNC client may receive lossy framebuffer updates
2423 depending on its encoding settings. Enabling this option can
2424 save a lot of bandwidth at the expense of quality.
2426 ``non-adaptive=on|off``
2427 Disable adaptive encodings. Adaptive encodings are enabled by
2428 default. An adaptive encoding will try to detect frequently
2429 updated screen regions, and send updates in these regions using
2430 a lossy encoding (like JPEG). This can be really helpful to save
2431 bandwidth when playing videos. Disabling adaptive encodings
2432 restores the original static behavior of encodings like Tight.
2434 ``share=[allow-exclusive|force-shared|ignore]``
2435 Set display sharing policy. 'allow-exclusive' allows clients to
2436 ask for exclusive access. As suggested by the rfb spec this is
2437 implemented by dropping other connections. Connecting multiple
2438 clients in parallel requires all clients asking for a shared
2439 session (vncviewer: -shared switch). This is the default.
2440 'force-shared' disables exclusive client access. Useful for
2441 shared desktop sessions, where you don't want someone forgetting
2442 specify -shared disconnect everybody else. 'ignore' completely
2443 ignores the shared flag and allows everybody connect
2444 unconditionally. Doesn't conform to the rfb spec but is
2445 traditional QEMU behavior.
2448 Set keyboard delay, for key down and key up events, in
2449 milliseconds. Default is 10. Keyboards are low-bandwidth
2450 devices, so this slowdown can help the device and guest to keep
2451 up and not lose events in case events are arriving in bulk.
2452 Possible causes for the latter are flaky network connections, or
2453 scripts for automated testing.
2455 ``audiodev=audiodev``
2456 Use the specified audiodev when the VNC client requests audio
2457 transmission. When not using an -audiodev argument, this option
2458 must be omitted, otherwise is must be present and specify a
2461 ``power-control=on|off``
2462 Permit the remote client to issue shutdown, reboot or reset power
2466 ARCHHEADING(, QEMU_ARCH_I386)
2468 ARCHHEADING(i386 target only:, QEMU_ARCH_I386)
2470 DEF("win2k
-hack
", 0, QEMU_OPTION_win2k_hack,
2471 "-win2k
-hack use it when installing Windows
2000 to avoid a disk full bug
\n",
2475 Use it when installing Windows 2000 to avoid a disk full bug. After
2476 Windows 2000 is installed, you no longer need this option (this
2477 option slows down the IDE transfers).
2480 DEF("no
-fd
-bootchk
", 0, QEMU_OPTION_no_fd_bootchk,
2481 "-no
-fd
-bootchk disable boot signature checking
for floppy disks
\n",
2485 Disable boot signature checking for floppy disks in BIOS. May be
2486 needed to boot from old floppy disks.
2489 DEF("no
-acpi
", 0, QEMU_OPTION_no_acpi,
2490 "-no
-acpi disable ACPI
\n", QEMU_ARCH_I386 | QEMU_ARCH_ARM)
2493 Disable ACPI (Advanced Configuration and Power Interface) support.
2494 Use it if your guest OS complains about ACPI problems (PC target
2498 DEF("no
-hpet
", 0, QEMU_OPTION_no_hpet,
2499 "-no
-hpet disable HPET
\n", QEMU_ARCH_I386)
2502 Disable HPET support.
2505 DEF("acpitable
", HAS_ARG, QEMU_OPTION_acpitable,
2506 "-acpitable
[sig
=str
][,rev
=n
][,oem_id
=str
][,oem_table_id
=str
][,oem_rev
=n
][,asl_compiler_id
=str
][,asl_compiler_rev
=n
][,{data|file
}=file1
[:file2
]...]\n"
2507 " ACPI table description
\n", QEMU_ARCH_I386)
2509 ``-acpitable [sig=str][,rev=n][,oem_id=str][,oem_table_id=str][,oem_rev=n] [,asl_compiler_id=str][,asl_compiler_rev=n][,data=file1[:file2]...]``
2510 Add ACPI table with specified header fields and context from
2511 specified files. For file=, take whole ACPI table from the specified
2512 files, including all ACPI headers (possible overridden by other
2513 options). For data=, only data portion of the table is used, all
2514 header information is specified in the command line. If a SLIC table
2515 is supplied to QEMU, then the SLIC's oem\_id and oem\_table\_id
2516 fields will override the same in the RSDT and the FADT (a.k.a.
2517 FACP), in order to ensure the field matches required by the
2518 Microsoft SLIC spec and the ACPI spec.
2521 DEF("smbios
", HAS_ARG, QEMU_OPTION_smbios,
2522 "-smbios file
=binary
\n"
2523 " load SMBIOS entry from binary file
\n"
2524 "-smbios type
=0[,vendor
=str
][,version
=str
][,date
=str
][,release
=%d
.%d
]\n"
2526 " specify SMBIOS type
0 fields
\n"
2527 "-smbios type
=1[,manufacturer
=str
][,product
=str
][,version
=str
][,serial
=str
]\n"
2528 " [,uuid
=uuid
][,sku
=str
][,family
=str
]\n"
2529 " specify SMBIOS type
1 fields
\n"
2530 "-smbios type
=2[,manufacturer
=str
][,product
=str
][,version
=str
][,serial
=str
]\n"
2531 " [,asset
=str
][,location
=str
]\n"
2532 " specify SMBIOS type
2 fields
\n"
2533 "-smbios type
=3[,manufacturer
=str
][,version
=str
][,serial
=str
][,asset
=str
]\n"
2535 " specify SMBIOS type
3 fields
\n"
2536 "-smbios type
=4[,sock_pfx
=str
][,manufacturer
=str
][,version
=str
][,serial
=str
]\n"
2537 " [,asset
=str
][,part
=str
][,max
-speed
=%d
][,current
-speed
=%d
]\n"
2538 " specify SMBIOS type
4 fields
\n"
2539 "-smbios type
=11[,value
=str
][,path
=filename
]\n"
2540 " specify SMBIOS type
11 fields
\n"
2541 "-smbios type
=17[,loc_pfx
=str
][,bank
=str
][,manufacturer
=str
][,serial
=str
]\n"
2542 " [,asset
=str
][,part
=str
][,speed
=%d
]\n"
2543 " specify SMBIOS type
17 fields
\n"
2544 "-smbios type
=41[,designation
=str
][,kind
=str
][,instance
=%d
][,pcidev
=str
]\n"
2545 " specify SMBIOS type
41 fields
\n",
2546 QEMU_ARCH_I386 | QEMU_ARCH_ARM)
2548 ``-smbios file=binary``
2549 Load SMBIOS entry from binary file.
2551 ``-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d][,uefi=on|off]``
2552 Specify SMBIOS type 0 fields
2554 ``-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str][,uuid=uuid][,sku=str][,family=str]``
2555 Specify SMBIOS type 1 fields
2557 ``-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str][,asset=str][,location=str]``
2558 Specify SMBIOS type 2 fields
2560 ``-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str][,sku=str]``
2561 Specify SMBIOS type 3 fields
2563 ``-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str][,asset=str][,part=str]``
2564 Specify SMBIOS type 4 fields
2566 ``-smbios type=11[,value=str][,path=filename]``
2567 Specify SMBIOS type 11 fields
2569 This argument can be repeated multiple times, and values are added in the order they are parsed.
2570 Applications intending to use OEM strings data are encouraged to use their application name as
2571 a prefix for the value string. This facilitates passing information for multiple applications
2574 The ``value=str`` syntax provides the string data inline, while the ``path=filename`` syntax
2575 loads data from a file on disk. Note that the file is not permitted to contain any NUL bytes.
2577 Both the ``value`` and ``path`` options can be repeated multiple times and will be added to
2578 the SMBIOS table in the order in which they appear.
2580 Note that on the x86 architecture, the total size of all SMBIOS tables is limited to 65535
2581 bytes. Thus the OEM strings data is not suitable for passing large amounts of data into the
2582 guest. Instead it should be used as a indicator to inform the guest where to locate the real
2583 data set, for example, by specifying the serial ID of a block device.
2585 An example passing three strings is
2589 -smbios type=11,value=cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/,\\
2590 value=anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os,\\
2591 path=/some/file/with/oemstringsdata.txt
2593 In the guest OS this is visible with the ``dmidecode`` command
2598 Handle 0x0E00, DMI type 11, 5 bytes
2600 String 1: cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/
2601 String 2: anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os
2602 String 3: myapp:some extra data
2605 ``-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str][,asset=str][,part=str][,speed=%d]``
2606 Specify SMBIOS type 17 fields
2608 ``-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]``
2609 Specify SMBIOS type 41 fields
2611 This argument can be repeated multiple times. Its main use is to allow network interfaces be created
2612 as ``enoX`` on Linux, with X being the instance number, instead of the name depending on the interface
2613 position on the PCI bus.
2615 Here is an example of use:
2619 -netdev user,id=internet \\
2620 -device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \\
2621 -smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev
2623 In the guest OS, the device should then appear as ``eno1``:
2628 lo UNKNOWN 00:00:00:00:00:00 <LOOPBACK,UP,LOWER_UP>
2629 eno1 UP 50:54:00:00:00:42 <BROADCAST,MULTICAST,UP,LOWER_UP>
2631 Currently, the PCI device has to be attached to the root bus.
2637 DEFHEADING(Network options:)
2639 DEF("netdev
", HAS_ARG, QEMU_OPTION_netdev,
2641 "-netdev user
,id
=str
[,ipv4
=on|off
][,net
=addr
[/mask
]][,host
=addr
]\n"
2642 " [,ipv6
=on|off
][,ipv6
-net
=addr
[/int
]][,ipv6
-host
=addr
]\n"
2643 " [,restrict
=on|off
][,hostname
=host
][,dhcpstart
=addr
]\n"
2644 " [,dns
=addr
][,ipv6
-dns
=addr
][,dnssearch
=domain
][,domainname
=domain
]\n"
2645 " [,tftp
=dir
][,tftp
-server
-name
=name
][,bootfile
=f
][,hostfwd
=rule
][,guestfwd
=rule
]"
2647 "[,smb
=dir
[,smbserver
=addr
]]\n"
2649 " configure a user mode network backend with ID
'str',\n"
2650 " its DHCP server and optional services
\n"
2653 "-netdev tap
,id
=str
,ifname
=name
\n"
2654 " configure a host TAP network backend with ID
'str'\n"
2656 "-netdev tap
,id
=str
[,fd
=h
][,fds
=x
:y
:...:z
][,ifname
=name
][,script
=file
][,downscript
=dfile
]\n"
2657 " [,br
=bridge
][,helper
=helper
][,sndbuf
=nbytes
][,vnet_hdr
=on|off
][,vhost
=on|off
]\n"
2658 " [,vhostfd
=h
][,vhostfds
=x
:y
:...:z
][,vhostforce
=on|off
][,queues
=n
]\n"
2660 " configure a host TAP network backend with ID
'str'\n"
2661 " connected to a
bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2662 " use network scripts
'file' (default=" DEFAULT_NETWORK_SCRIPT ")\n"
2663 " to configure it and
'dfile' (default=" DEFAULT_NETWORK_DOWN_SCRIPT ")\n"
2664 " to deconfigure it
\n"
2665 " use
'[down]script=no' to disable script execution
\n"
2666 " use network helper
'helper' (default=" DEFAULT_BRIDGE_HELPER ") to
\n"
2668 " use
'fd=h' to connect to an already opened TAP
interface\n"
2669 " use
'fds=x:y:...:z' to connect to already opened multiqueue capable TAP interfaces
\n"
2670 " use
'sndbuf=nbytes' to limit the size of the send
buffer (the
\n"
2671 " default is disabled
'sndbuf=0' to enable flow control set
'sndbuf=1048576')\n"
2672 " use vnet_hdr
=off to avoid enabling the IFF_VNET_HDR tap flag
\n"
2673 " use vnet_hdr
=on to make the lack of IFF_VNET_HDR support an error condition
\n"
2674 " use vhost
=on to enable experimental
in kernel accelerator
\n"
2675 " (only has effect
for virtio guests which use MSIX
)\n"
2676 " use vhostforce
=on to force vhost on
for non
-MSIX virtio guests
\n"
2677 " use
'vhostfd=h' to connect to an already opened vhost net device
\n"
2678 " use
'vhostfds=x:y:...:z to connect to multiple already opened vhost net devices\n"
2679 " use 'queues
=n
' to specify the number of queues to be created for multiqueue TAP\n"
2680 " use 'poll
-us
=n
' to specify the maximum number of microseconds that could be\n"
2681 " spent on busy polling for vhost net\n"
2682 "-netdev bridge,id=str[,br=bridge][,helper=helper]\n"
2683 " configure a host TAP network backend with ID 'str
' that is\n"
2684 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2685 " using the program 'helper (default=" DEFAULT_BRIDGE_HELPER ")\n"
2688 "-netdev l2tpv3
,id
=str
,src
=srcaddr
,dst
=dstaddr
[,srcport
=srcport
][,dstport
=dstport
]\n"
2689 " [,rxsession
=rxsession
],txsession
=txsession
[,ipv6
=on|off
][,udp
=on|off
]\n"
2690 " [,cookie64
=on|off
][,counter
][,pincounter
][,txcookie
=txcookie
]\n"
2691 " [,rxcookie
=rxcookie
][,offset
=offset
]\n"
2692 " configure a network backend with ID
'str' connected to
\n"
2693 " an Ethernet over L2TPv3 pseudowire
.\n"
2694 " Linux kernel
3.3+ as well as most routers can talk
\n"
2695 " L2TPv3
. This transport allows connecting a VM to a VM
,\n"
2696 " VM to a router and even VM to Host
. It is a nearly
-universal
\n"
2697 " standard (RFC3931
). Note
- this implementation uses
static\n"
2698 " pre
-configured
tunnels (same as the Linux kernel
).\n"
2699 " use
'src=' to specify source address
\n"
2700 " use
'dst=' to specify destination address
\n"
2701 " use
'udp=on' to specify udp encapsulation
\n"
2702 " use
'srcport=' to specify source udp port
\n"
2703 " use
'dstport=' to specify destination udp port
\n"
2704 " use
'ipv6=on' to force v6
\n"
2705 " L2TPv3 uses cookies to prevent misconfiguration as
\n"
2706 " well as a weak security measure
\n"
2707 " use
'rxcookie=0x012345678' to specify a rxcookie
\n"
2708 " use
'txcookie=0x012345678' to specify a txcookie
\n"
2709 " use
'cookie64=on' to set cookie size to
64 bit
, otherwise
32\n"
2710 " use
'counter=off' to force a
'cut-down' L2TPv3 with no counter
\n"
2711 " use
'pincounter=on' to work around broken counter handling
in peer
\n"
2712 " use
'offset=X' to add an extra offset between header and data
\n"
2714 "-netdev socket
,id
=str
[,fd
=h
][,listen
=[host
]:port
][,connect
=host
:port
]\n"
2715 " configure a network backend to connect to another network
\n"
2716 " using a socket connection
\n"
2717 "-netdev socket
,id
=str
[,fd
=h
][,mcast
=maddr
:port
[,localaddr
=addr
]]\n"
2718 " configure a network backend to connect to a multicast maddr and port
\n"
2719 " use
'localaddr=addr' to specify the host address to send packets from
\n"
2720 "-netdev socket
,id
=str
[,fd
=h
][,udp
=host
:port
][,localaddr
=host
:port
]\n"
2721 " configure a network backend to connect to another network
\n"
2722 " using an UDP tunnel
\n"
2724 "-netdev vde
,id
=str
[,sock
=socketpath
][,port
=n
][,group
=groupname
][,mode
=octalmode
]\n"
2725 " configure a network backend to connect to port
'n' of a vde
switch\n"
2726 " running on host and listening
for incoming connections on
'socketpath'.\n"
2727 " Use group
'groupname' and mode
'octalmode' to change
default\n"
2728 " ownership and permissions
for communication port
.\n"
2730 #ifdef CONFIG_NETMAP
2731 "-netdev netmap
,id
=str
,ifname
=name
[,devname
=nmname
]\n"
2732 " attach to the existing netmap
-enabled network
interface 'name', or to a
\n"
2733 " VALE
port (created on the fly
) called
'name' ('nmname' is name of the
\n"
2734 " netmap device
, defaults to
'/dev/netmap')\n"
2737 "-netdev vhost
-user
,id
=str
,chardev
=dev
[,vhostforce
=on|off
]\n"
2738 " configure a vhost
-user network
, backed by a chardev
'dev'\n"
2741 "-netdev vhost
-vdpa
,id
=str
,vhostdev
=/path
/to
/dev
\n"
2742 " configure a vhost
-vdpa network
,Establish a vhost
-vdpa netdev
\n"
2744 "-netdev hubport
,id
=str
,hubid
=n
[,netdev
=nd
]\n"
2745 " configure a hub port on the hub with ID
'n'\n", QEMU_ARCH_ALL)
2746 DEF("nic
", HAS_ARG, QEMU_OPTION_nic,
2757 #ifdef CONFIG_NETMAP
2763 "socket
][,option
][,...][mac
=macaddr
]\n"
2764 " initialize an on
-board
/ default host
NIC (using MAC address
\n"
2765 " macaddr
) and connect it to the given host network backend
\n"
2766 "-nic none use it alone to have zero network
devices (the
default is to
\n"
2767 " provided a
'user' network connection
)\n",
2769 DEF("net
", HAS_ARG, QEMU_OPTION_net,
2770 "-net nic
[,macaddr
=mac
][,model
=type
][,name
=str
][,addr
=str
][,vectors
=v
]\n"
2771 " configure or create an on
-board (or machine
default) NIC and
\n"
2772 " connect it to hub
0 (please use
-nic unless you need a hub
)\n"
2782 #ifdef CONFIG_NETMAP
2785 "socket
][,option
][,option
][,...]\n"
2786 " old way to initialize a host network
interface\n"
2787 " (use the
-netdev option
if possible instead
)\n", QEMU_ARCH_ALL)
2789 ``-nic [tap|bridge|user|l2tpv3|vde|netmap|vhost-user|socket][,...][,mac=macaddr][,model=mn]``
2790 This option is a shortcut for configuring both the on-board
2791 (default) guest NIC hardware and the host network backend in one go.
2792 The host backend options are the same as with the corresponding
2793 ``-netdev`` options below. The guest NIC model can be set with
2794 ``model=modelname``. Use ``model=help`` to list the available device
2795 types. The hardware MAC address can be set with ``mac=macaddr``.
2797 The following two example do exactly the same, to show how ``-nic``
2798 can be used to shorten the command line length:
2802 |qemu_system| -netdev user,id=n1,ipv6=off -device e1000,netdev=n1,mac=52:54:98:76:54:32
2803 |qemu_system| -nic user,ipv6=off,model=e1000,mac=52:54:98:76:54:32
2806 Indicate that no network devices should be configured. It is used to
2807 override the default configuration (default NIC with "user
" host
2808 network backend) which is activated if no other networking options
2811 ``-netdev user,id=id[,option][,option][,...]``
2812 Configure user mode host network backend which requires no
2813 administrator privilege to run. Valid options are:
2816 Assign symbolic name for use in monitor commands.
2818 ``ipv4=on|off and ipv6=on|off``
2819 Specify that either IPv4 or IPv6 must be enabled. If neither is
2820 specified both protocols are enabled.
2823 Set IP network address the guest will see. Optionally specify
2824 the netmask, either in the form a.b.c.d or as number of valid
2825 top-most bits. Default is 10.0.2.0/24.
2828 Specify the guest-visible address of the host. Default is the
2829 2nd IP in the guest network, i.e. x.x.x.2.
2831 ``ipv6-net=addr[/int]``
2832 Set IPv6 network address the guest will see (default is
2833 fec0::/64). The network prefix is given in the usual hexadecimal
2834 IPv6 address notation. The prefix size is optional, and is given
2835 as the number of valid top-most bits (default is 64).
2838 Specify the guest-visible IPv6 address of the host. Default is
2839 the 2nd IPv6 in the guest network, i.e. xxxx::2.
2842 If this option is enabled, the guest will be isolated, i.e. it
2843 will not be able to contact the host and no guest IP packets
2844 will be routed over the host to the outside. This option does
2845 not affect any explicitly set forwarding rules.
2848 Specifies the client hostname reported by the built-in DHCP
2852 Specify the first of the 16 IPs the built-in DHCP server can
2853 assign. Default is the 15th to 31st IP in the guest network,
2854 i.e. x.x.x.15 to x.x.x.31.
2857 Specify the guest-visible address of the virtual nameserver. The
2858 address must be different from the host address. Default is the
2859 3rd IP in the guest network, i.e. x.x.x.3.
2862 Specify the guest-visible address of the IPv6 virtual
2863 nameserver. The address must be different from the host address.
2864 Default is the 3rd IP in the guest network, i.e. xxxx::3.
2866 ``dnssearch=domain``
2867 Provides an entry for the domain-search list sent by the
2868 built-in DHCP server. More than one domain suffix can be
2869 transmitted by specifying this option multiple times. If
2870 supported, this will cause the guest to automatically try to
2871 append the given domain suffix(es) in case a domain name can not
2878 |qemu_system| -nic user,dnssearch=mgmt.example.org,dnssearch=example.org
2880 ``domainname=domain``
2881 Specifies the client domain name reported by the built-in DHCP
2885 When using the user mode network stack, activate a built-in TFTP
2886 server. The files in dir will be exposed as the root of a TFTP
2887 server. The TFTP client on the guest must be configured in
2888 binary mode (use the command ``bin`` of the Unix TFTP client).
2890 ``tftp-server-name=name``
2891 In BOOTP reply, broadcast name as the "TFTP server name
"
2892 (RFC2132 option 66). This can be used to advise the guest to
2893 load boot files or configurations from a different server than
2897 When using the user mode network stack, broadcast file as the
2898 BOOTP filename. In conjunction with ``tftp``, this can be used
2899 to network boot a guest from a local directory.
2901 Example (using pxelinux):
2905 |qemu_system| -hda linux.img -boot n -device e1000,netdev=n1 \\
2906 -netdev user,id=n1,tftp=/path/to/tftp/files,bootfile=/pxelinux.0
2908 ``smb=dir[,smbserver=addr]``
2909 When using the user mode network stack, activate a built-in SMB
2910 server so that Windows OSes can access to the host files in
2911 ``dir`` transparently. The IP address of the SMB server can be
2912 set to addr. By default the 4th IP in the guest network is used,
2915 In the guest Windows OS, the line:
2921 must be added in the file ``C:\WINDOWS\LMHOSTS`` (for windows
2922 9x/Me) or ``C:\WINNT\SYSTEM32\DRIVERS\ETC\LMHOSTS`` (Windows
2925 Then ``dir`` can be accessed in ``\\smbserver\qemu``.
2927 Note that a SAMBA server must be installed on the host OS.
2929 ``hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport``
2930 Redirect incoming TCP or UDP connections to the host port
2931 hostport to the guest IP address guestaddr on guest port
2932 guestport. If guestaddr is not specified, its value is x.x.x.15
2933 (default first address given by the built-in DHCP server). By
2934 specifying hostaddr, the rule can be bound to a specific host
2935 interface. If no connection type is set, TCP is used. This
2936 option can be given multiple times.
2938 For example, to redirect host X11 connection from screen 1 to
2939 guest screen 0, use the following:
2944 |qemu_system| -nic user,hostfwd=tcp:127.0.0.1:6001-:6000
2945 # this host xterm should open in the guest X11 server
2948 To redirect telnet connections from host port 5555 to telnet
2949 port on the guest, use the following:
2954 |qemu_system| -nic user,hostfwd=tcp::5555-:23
2955 telnet localhost 5555
2957 Then when you use on the host ``telnet localhost 5555``, you
2958 connect to the guest telnet server.
2960 ``guestfwd=[tcp]:server:port-dev``; \ ``guestfwd=[tcp]:server:port-cmd:command``
2961 Forward guest TCP connections to the IP address server on port
2962 port to the character device dev or to a program executed by
2963 cmd:command which gets spawned for each connection. This option
2964 can be given multiple times.
2966 You can either use a chardev directly and have that one used
2967 throughout QEMU's lifetime, like in the following example:
2971 # open 10.10.1.1:4321 on bootup, connect 10.0.2.100:1234 to it whenever
2972 # the guest accesses it
2973 |qemu_system| -nic user,guestfwd=tcp:10.0.2.100:1234-tcp:10.10.1.1:4321
2975 Or you can execute a command on every TCP connection established
2976 by the guest, so that QEMU behaves similar to an inetd process
2977 for that virtual server:
2981 # call "netcat
10.10.1.1 4321" on every TCP connection to 10.0.2.100:1234
2982 # and connect the TCP stream to its stdin/stdout
2983 |qemu_system| -nic 'user,id=n1,guestfwd=tcp:10.0.2.100:1234-cmd:netcat 10.10.1.1 4321'
2985 ``-netdev tap,id=id[,fd=h][,ifname=name][,script=file][,downscript=dfile][,br=bridge][,helper=helper]``
2986 Configure a host TAP network backend with ID id.
2988 Use the network script file to configure it and the network script
2989 dfile to deconfigure it. If name is not provided, the OS
2990 automatically provides one. The default network configure script is
2991 ``/etc/qemu-ifup`` and the default network deconfigure script is
2992 ``/etc/qemu-ifdown``. Use ``script=no`` or ``downscript=no`` to
2993 disable script execution.
2995 If running QEMU as an unprivileged user, use the network helper
2996 to configure the TAP interface and attach it to the bridge.
2997 The default network helper executable is
2998 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3001 ``fd``\ =h can be used to specify the handle of an already opened
3008 #launch a QEMU instance with the default network script
3009 |qemu_system| linux.img -nic tap
3013 #launch a QEMU instance with two NICs, each one connected
3015 |qemu_system| linux.img \\
3016 -netdev tap,id=nd0,ifname=tap0 -device e1000,netdev=nd0 \\
3017 -netdev tap,id=nd1,ifname=tap1 -device rtl8139,netdev=nd1
3021 #launch a QEMU instance with the default network helper to
3022 #connect a TAP device to bridge br0
3023 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3024 -netdev tap,id=n1,"helper
=/path
/to
/qemu
-bridge
-helper
"
3026 ``-netdev bridge,id=id[,br=bridge][,helper=helper]``
3027 Connect a host TAP network interface to a host bridge device.
3029 Use the network helper helper to configure the TAP interface and
3030 attach it to the bridge. The default network helper executable is
3031 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3038 #launch a QEMU instance with the default network helper to
3039 #connect a TAP device to bridge br0
3040 |qemu_system| linux.img -netdev bridge,id=n1 -device virtio-net,netdev=n1
3044 #launch a QEMU instance with the default network helper to
3045 #connect a TAP device to bridge qemubr0
3046 |qemu_system| linux.img -netdev bridge,br=qemubr0,id=n1 -device virtio-net,netdev=n1
3048 ``-netdev socket,id=id[,fd=h][,listen=[host]:port][,connect=host:port]``
3049 This host network backend can be used to connect the guest's network
3050 to another QEMU virtual machine using a TCP socket connection. If
3051 ``listen`` is specified, QEMU waits for incoming connections on port
3052 (host is optional). ``connect`` is used to connect to another QEMU
3053 instance using the ``listen`` option. ``fd``\ =h specifies an
3054 already opened TCP socket.
3060 # launch a first QEMU instance
3061 |qemu_system| linux.img \\
3062 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3063 -netdev socket,id=n1,listen=:1234
3064 # connect the network of this instance to the network of the first instance
3065 |qemu_system| linux.img \\
3066 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3067 -netdev socket,id=n2,connect=127.0.0.1:1234
3069 ``-netdev socket,id=id[,fd=h][,mcast=maddr:port[,localaddr=addr]]``
3070 Configure a socket host network backend to share the guest's network
3071 traffic with another QEMU virtual machines using a UDP multicast
3072 socket, effectively making a bus for every QEMU with same multicast
3073 address maddr and port. NOTES:
3075 1. Several QEMU can be running on different hosts and share same bus
3076 (assuming correct multicast setup for these hosts).
3078 2. mcast support is compatible with User Mode Linux (argument
3079 ``ethN=mcast``), see http://user-mode-linux.sf.net.
3081 3. Use ``fd=h`` to specify an already opened UDP multicast socket.
3087 # launch one QEMU instance
3088 |qemu_system| linux.img \\
3089 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3090 -netdev socket,id=n1,mcast=230.0.0.1:1234
3091 # launch another QEMU instance on same "bus
"
3092 |qemu_system| linux.img \\
3093 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3094 -netdev socket,id=n2,mcast=230.0.0.1:1234
3095 # launch yet another QEMU instance on same "bus
"
3096 |qemu_system| linux.img \\
3097 -device e1000,netdev=n3,mac=52:54:00:12:34:58 \\
3098 -netdev socket,id=n3,mcast=230.0.0.1:1234
3100 Example (User Mode Linux compat.):
3104 # launch QEMU instance (note mcast address selected is UML's default)
3105 |qemu_system| linux.img \\
3106 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3107 -netdev socket,id=n1,mcast=239.192.168.1:1102
3109 /path/to/linux ubd0=/path/to/root_fs eth0=mcast
3111 Example (send packets from host's 1.2.3.4):
3115 |qemu_system| linux.img \\
3116 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3117 -netdev socket,id=n1,mcast=239.192.168.1:1102,localaddr=1.2.3.4
3119 ``-netdev l2tpv3,id=id,src=srcaddr,dst=dstaddr[,srcport=srcport][,dstport=dstport],txsession=txsession[,rxsession=rxsession][,ipv6=on|off][,udp=on|off][,cookie64][,counter][,pincounter][,txcookie=txcookie][,rxcookie=rxcookie][,offset=offset]``
3120 Configure a L2TPv3 pseudowire host network backend. L2TPv3 (RFC3931)
3121 is a popular protocol to transport Ethernet (and other Layer 2) data
3122 frames between two systems. It is present in routers, firewalls and
3123 the Linux kernel (from version 3.3 onwards).
3125 This transport allows a VM to communicate to another VM, router or
3129 source address (mandatory)
3132 destination address (mandatory)
3135 select udp encapsulation (default is ip).
3141 destination udp port.
3144 force v6, otherwise defaults to v4.
3146 ``rxcookie=rxcookie``; \ ``txcookie=txcookie``
3147 Cookies are a weak form of security in the l2tpv3 specification.
3148 Their function is mostly to prevent misconfiguration. By default
3152 Set cookie size to 64 bit instead of the default 32
3155 Force a 'cut-down' L2TPv3 with no counter as in
3156 draft-mkonstan-l2tpext-keyed-ipv6-tunnel-00
3159 Work around broken counter handling in peer. This may also help
3160 on networks which have packet reorder.
3163 Add an extra offset between header and data
3165 For example, to attach a VM running on host 4.3.2.1 via L2TPv3 to
3166 the bridge br-lan on the remote Linux host 1.2.3.4:
3170 # Setup tunnel on linux host using raw ip as encapsulation
3172 ip l2tp add tunnel remote 4.3.2.1 local 1.2.3.4 tunnel_id 1 peer_tunnel_id 1 \\
3173 encap udp udp_sport 16384 udp_dport 16384
3174 ip l2tp add session tunnel_id 1 name vmtunnel0 session_id \\
3175 0xFFFFFFFF peer_session_id 0xFFFFFFFF
3176 ifconfig vmtunnel0 mtu 1500
3177 ifconfig vmtunnel0 up
3178 brctl addif br-lan vmtunnel0
3182 # launch QEMU instance - if your network has reorder or is very lossy add ,pincounter
3184 |qemu_system| linux.img -device e1000,netdev=n1 \\
3185 -netdev l2tpv3,id=n1,src=4.2.3.1,dst=1.2.3.4,udp,srcport=16384,dstport=16384,rxsession=0xffffffff,txsession=0xffffffff,counter
3187 ``-netdev vde,id=id[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]``
3188 Configure VDE backend to connect to PORT n of a vde switch running
3189 on host and listening for incoming connections on socketpath. Use
3190 GROUP groupname and MODE octalmode to change default ownership and
3191 permissions for communication port. This option is only available if
3192 QEMU has been compiled with vde support enabled.
3199 vde_switch -F -sock /tmp/myswitch
3200 # launch QEMU instance
3201 |qemu_system| linux.img -nic vde,sock=/tmp/myswitch
3203 ``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]``
3204 Establish a vhost-user netdev, backed by a chardev id. The chardev
3205 should be a unix domain socket backed one. The vhost-user uses a
3206 specifically defined protocol to pass vhost ioctl replacement
3207 messages to an application on the other end of the socket. On
3208 non-MSIX guests, the feature can be forced with vhostforce. Use
3209 'queues=n' to specify the number of queues to be created for
3210 multiqueue vhost-user.
3216 qemu -m 512 -object memory-backend-file,id=mem,size=512M,mem-path=/hugetlbfs,share=on \
3217 -numa node,memdev=mem \
3218 -chardev socket,id=chr0,path=/path/to/socket \
3219 -netdev type=vhost-user,id=net0,chardev=chr0 \
3220 -device virtio-net-pci,netdev=net0
3222 ``-netdev vhost-vdpa,vhostdev=/path/to/dev``
3223 Establish a vhost-vdpa netdev.
3225 vDPA device is a device that uses a datapath which complies with
3226 the virtio specifications with a vendor specific control path.
3227 vDPA devices can be both physically located on the hardware or
3228 emulated by software.
3230 ``-netdev hubport,id=id,hubid=hubid[,netdev=nd]``
3231 Create a hub port on the emulated hub with ID hubid.
3233 The hubport netdev lets you connect a NIC to a QEMU emulated hub
3234 instead of a single netdev. Alternatively, you can also connect the
3235 hubport to another netdev with ID nd by using the ``netdev=nd``
3238 ``-net nic[,netdev=nd][,macaddr=mac][,model=type] [,name=name][,addr=addr][,vectors=v]``
3239 Legacy option to configure or create an on-board (or machine
3240 default) Network Interface Card(NIC) and connect it either to the
3241 emulated hub with ID 0 (i.e. the default hub), or to the netdev nd.
3242 If model is omitted, then the default NIC model associated with the
3243 machine type is used. Note that the default NIC model may change in
3244 future QEMU releases, so it is highly recommended to always specify
3245 a model. Optionally, the MAC address can be changed to mac, the
3246 device address set to addr (PCI cards only), and a name can be
3247 assigned for use in monitor commands. Optionally, for PCI cards, you
3248 can specify the number v of MSI-X vectors that the card should have;
3249 this option currently only affects virtio cards; set v = 0 to
3250 disable MSI-X. If no ``-net`` option is specified, a single NIC is
3251 created. QEMU can emulate several different models of network card.
3252 Use ``-net nic,model=help`` for a list of available devices for your
3255 ``-net user|tap|bridge|socket|l2tpv3|vde[,...][,name=name]``
3256 Configure a host network backend (with the options corresponding to
3257 the same ``-netdev`` option) and connect it to the emulated hub 0
3258 (the default hub). Use name to specify the name of the hub port.
3263 DEFHEADING(Character device options:)
3265 DEF("chardev
", HAS_ARG, QEMU_OPTION_chardev,
3267 "-chardev
null,id
=id
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3268 "-chardev socket
,id
=id
[,host
=host
],port
=port
[,to
=to
][,ipv4
=on|off
][,ipv6
=on|off
][,nodelay
=on|off
]\n"
3269 " [,server
=on|off
][,wait
=on|off
][,telnet
=on|off
][,websocket
=on|off
][,reconnect
=seconds
][,mux
=on|off
]\n"
3270 " [,logfile
=PATH
][,logappend
=on|off
][,tls
-creds
=ID
][,tls
-authz
=ID
] (tcp
)\n"
3271 "-chardev socket
,id
=id
,path
=path
[,server
=on|off
][,wait
=on|off
][,telnet
=on|off
][,websocket
=on|off
][,reconnect
=seconds
]\n"
3272 " [,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
][,abstract
=on|off
][,tight
=on|off
] (unix
)\n"
3273 "-chardev udp
,id
=id
[,host
=host
],port
=port
[,localaddr
=localaddr
]\n"
3274 " [,localport
=localport
][,ipv4
=on|off
][,ipv6
=on|off
][,mux
=on|off
]\n"
3275 " [,logfile
=PATH
][,logappend
=on|off
]\n"
3276 "-chardev msmouse
,id
=id
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3277 "-chardev vc
,id
=id
[[,width
=width
][,height
=height
]][[,cols
=cols
][,rows
=rows
]]\n"
3278 " [,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3279 "-chardev ringbuf
,id
=id
[,size
=size
][,logfile
=PATH
][,logappend
=on|off
]\n"
3280 "-chardev file
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3281 "-chardev pipe
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3283 "-chardev console
,id
=id
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3284 "-chardev serial
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3286 "-chardev pty
,id
=id
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3287 "-chardev stdio
,id
=id
[,mux
=on|off
][,signal
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3289 #ifdef CONFIG_BRLAPI
3290 "-chardev braille
,id
=id
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3292 #if defined(__linux__) || defined(__sun__) || defined(__FreeBSD__) \
3293 || defined(__NetBSD__) || defined(__OpenBSD__) || defined(__DragonFly__)
3294 "-chardev serial
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3295 "-chardev tty
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3297 #if defined(__linux__) || defined(__FreeBSD__) || defined(__DragonFly__)
3298 "-chardev parallel
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3299 "-chardev parport
,id
=id
,path
=path
[,mux
=on|off
][,logfile
=PATH
][,logappend
=on|off
]\n"
3301 #if defined(CONFIG_SPICE)
3302 "-chardev spicevmc
,id
=id
,name
=name
[,debug
=debug
][,logfile
=PATH
][,logappend
=on|off
]\n"
3303 "-chardev spiceport
,id
=id
,name
=name
[,debug
=debug
][,logfile
=PATH
][,logappend
=on|off
]\n"
3309 The general form of a character device option is:
3311 ``-chardev backend,id=id[,mux=on|off][,options]``
3312 Backend is one of: ``null``, ``socket``, ``udp``, ``msmouse``,
3313 ``vc``, ``ringbuf``, ``file``, ``pipe``, ``console``, ``serial``,
3314 ``pty``, ``stdio``, ``braille``, ``tty``, ``parallel``, ``parport``,
3315 ``spicevmc``, ``spiceport``. The specific backend will determine the
3318 Use ``-chardev help`` to print all available chardev backend types.
3320 All devices must have an id, which can be any string up to 127
3321 characters long. It is used to uniquely identify this device in
3322 other command line directives.
3324 A character device may be used in multiplexing mode by multiple
3325 front-ends. Specify ``mux=on`` to enable this mode. A multiplexer is
3326 a "1:N
" device, and here the "1" end is your specified chardev
3327 backend, and the "N
" end is the various parts of QEMU that can talk
3328 to a chardev. If you create a chardev with ``id=myid`` and
3329 ``mux=on``, QEMU will create a multiplexer with your specified ID,
3330 and you can then configure multiple front ends to use that chardev
3331 ID for their input/output. Up to four different front ends can be
3332 connected to a single multiplexed chardev. (Without multiplexing
3333 enabled, a chardev can only be used by a single front end.) For
3334 instance you could use this to allow a single stdio chardev to be
3335 used by two serial ports and the QEMU monitor:
3339 -chardev stdio,mux=on,id=char0 \
3340 -mon chardev=char0,mode=readline \
3341 -serial chardev:char0 \
3342 -serial chardev:char0
3344 You can have more than one multiplexer in a system configuration;
3345 for instance you could have a TCP port multiplexed between UART 0
3346 and UART 1, and stdio multiplexed between the QEMU monitor and a
3351 -chardev stdio,mux=on,id=char0 \
3352 -mon chardev=char0,mode=readline \
3353 -parallel chardev:char0 \
3354 -chardev tcp,...,mux=on,id=char1 \
3355 -serial chardev:char1 \
3356 -serial chardev:char1
3358 When you're using a multiplexed character device, some escape
3359 sequences are interpreted in the input. See the chapter about
3360 :ref:`keys in the character backend multiplexer` in the
3361 System Emulation Users Guide for more details.
3363 Note that some other command line options may implicitly create
3364 multiplexed character backends; for instance ``-serial mon:stdio``
3365 creates a multiplexed stdio backend connected to the serial port and
3366 the QEMU monitor, and ``-nographic`` also multiplexes the console
3367 and the monitor to stdio.
3369 There is currently no support for multiplexing in the other
3370 direction (where a single QEMU front end takes input and output from
3373 Every backend supports the ``logfile`` option, which supplies the
3374 path to a file to record all data transmitted via the backend. The
3375 ``logappend`` option controls whether the log file will be truncated
3376 or appended to when opened.
3378 The available backends are:
3380 ``-chardev null,id=id``
3381 A void device. This device will not emit any data, and will drop any
3382 data it receives. The null backend does not take any options.
3384 ``-chardev socket,id=id[,TCP options or unix options][,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds][,tls-creds=id][,tls-authz=id]``
3385 Create a two-way stream socket, which can be either a TCP or a unix
3386 socket. A unix socket will be created if ``path`` is specified.
3387 Behaviour is undefined if TCP options are specified for a unix
3390 ``server=on|off`` specifies that the socket shall be a listening socket.
3392 ``wait=on|off`` specifies that QEMU should not block waiting for a client
3393 to connect to a listening socket.
3395 ``telnet=on|off`` specifies that traffic on the socket should interpret
3396 telnet escape sequences.
3398 ``websocket=on|off`` specifies that the socket uses WebSocket protocol for
3401 ``reconnect`` sets the timeout for reconnecting on non-server
3402 sockets when the remote end goes away. qemu will delay this many
3403 seconds and then attempt to reconnect. Zero disables reconnecting,
3406 ``tls-creds`` requests enablement of the TLS protocol for
3407 encryption, and specifies the id of the TLS credentials to use for
3408 the handshake. The credentials must be previously created with the
3409 ``-object tls-creds`` argument.
3411 ``tls-auth`` provides the ID of the QAuthZ authorization object
3412 against which the client's x509 distinguished name will be
3413 validated. This object is only resolved at time of use, so can be
3414 deleted and recreated on the fly while the chardev server is active.
3415 If missing, it will default to denying access.
3417 TCP and unix socket options are given below:
3419 ``TCP options: port=port[,host=host][,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]``
3420 ``host`` for a listening socket specifies the local address to
3421 be bound. For a connecting socket species the remote host to
3422 connect to. ``host`` is optional for listening sockets. If not
3423 specified it defaults to ``0.0.0.0``.
3425 ``port`` for a listening socket specifies the local port to be
3426 bound. For a connecting socket specifies the port on the remote
3427 host to connect to. ``port`` can be given as either a port
3428 number or a service name. ``port`` is required.
3430 ``to`` is only relevant to listening sockets. If it is
3431 specified, and ``port`` cannot be bound, QEMU will attempt to
3432 bind to subsequent ports up to and including ``to`` until it
3433 succeeds. ``to`` must be specified as a port number.
3435 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4
3436 or IPv6 must be used. If neither is specified the socket may
3437 use either protocol.
3439 ``nodelay=on|off`` disables the Nagle algorithm.
3441 ``unix options: path=path[,abstract=on|off][,tight=on|off]``
3442 ``path`` specifies the local path of the unix socket. ``path``
3444 ``abstract=on|off`` specifies the use of the abstract socket namespace,
3445 rather than the filesystem. Optional, defaults to false.
3446 ``tight=on|off`` sets the socket length of abstract sockets to their minimum,
3447 rather than the full sun_path length. Optional, defaults to true.
3449 ``-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr][,localport=localport][,ipv4=on|off][,ipv6=on|off]``
3450 Sends all traffic from the guest to a remote host over UDP.
3452 ``host`` specifies the remote host to connect to. If not specified
3453 it defaults to ``localhost``.
3455 ``port`` specifies the port on the remote host to connect to.
3456 ``port`` is required.
3458 ``localaddr`` specifies the local address to bind to. If not
3459 specified it defaults to ``0.0.0.0``.
3461 ``localport`` specifies the local port to bind to. If not specified
3462 any available local port will be used.
3464 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4 or IPv6 must be used.
3465 If neither is specified the device may use either protocol.
3467 ``-chardev msmouse,id=id``
3468 Forward QEMU's emulated msmouse events to the guest. ``msmouse``
3469 does not take any options.
3471 ``-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]``
3472 Connect to a QEMU text console. ``vc`` may optionally be given a
3475 ``width`` and ``height`` specify the width and height respectively
3476 of the console, in pixels.
3478 ``cols`` and ``rows`` specify that the console be sized to fit a
3479 text console with the given dimensions.
3481 ``-chardev ringbuf,id=id[,size=size]``
3482 Create a ring buffer with fixed size ``size``. size must be a power
3483 of two and defaults to ``64K``.
3485 ``-chardev file,id=id,path=path``
3486 Log all traffic received from the guest to a file.
3488 ``path`` specifies the path of the file to be opened. This file will
3489 be created if it does not already exist, and overwritten if it does.
3490 ``path`` is required.
3492 ``-chardev pipe,id=id,path=path``
3493 Create a two-way connection to the guest. The behaviour differs
3494 slightly between Windows hosts and other hosts:
3496 On Windows, a single duplex pipe will be created at
3499 On other hosts, 2 pipes will be created called ``path.in`` and
3500 ``path.out``. Data written to ``path.in`` will be received by the
3501 guest. Data written by the guest can be read from ``path.out``. QEMU
3502 will not create these fifos, and requires them to be present.
3504 ``path`` forms part of the pipe path as described above. ``path`` is
3507 ``-chardev console,id=id``
3508 Send traffic from the guest to QEMU's standard output. ``console``
3509 does not take any options.
3511 ``console`` is only available on Windows hosts.
3513 ``-chardev serial,id=id,path=path``
3514 Send traffic from the guest to a serial device on the host.
3516 On Unix hosts serial will actually accept any tty device, not only
3519 ``path`` specifies the name of the serial device to open.
3521 ``-chardev pty,id=id``
3522 Create a new pseudo-terminal on the host and connect to it. ``pty``
3523 does not take any options.
3525 ``pty`` is not available on Windows hosts.
3527 ``-chardev stdio,id=id[,signal=on|off]``
3528 Connect to standard input and standard output of the QEMU process.
3530 ``signal`` controls if signals are enabled on the terminal, that
3531 includes exiting QEMU with the key sequence Control-c. This option
3532 is enabled by default, use ``signal=off`` to disable it.
3534 ``-chardev braille,id=id``
3535 Connect to a local BrlAPI server. ``braille`` does not take any
3538 ``-chardev tty,id=id,path=path``
3539 ``tty`` is only available on Linux, Sun, FreeBSD, NetBSD, OpenBSD
3540 and DragonFlyBSD hosts. It is an alias for ``serial``.
3542 ``path`` specifies the path to the tty. ``path`` is required.
3544 ``-chardev parallel,id=id,path=path``
3546 ``-chardev parport,id=id,path=path``
3547 ``parallel`` is only available on Linux, FreeBSD and DragonFlyBSD
3550 Connect to a local parallel port.
3552 ``path`` specifies the path to the parallel port device. ``path`` is
3555 ``-chardev spicevmc,id=id,debug=debug,name=name``
3556 ``spicevmc`` is only available when spice support is built in.
3558 ``debug`` debug level for spicevmc
3560 ``name`` name of spice channel to connect to
3562 Connect to a spice virtual machine channel, such as vdiport.
3564 ``-chardev spiceport,id=id,debug=debug,name=name``
3565 ``spiceport`` is only available when spice support is built in.
3567 ``debug`` debug level for spicevmc
3569 ``name`` name of spice port to connect to
3571 Connect to a spice port, allowing a Spice client to handle the
3572 traffic identified by a name (preferably a fqdn).
3578 DEFHEADING(TPM device options:)
3580 DEF("tpmdev
", HAS_ARG, QEMU_OPTION_tpmdev, \
3581 "-tpmdev passthrough
,id
=id
[,path
=path
][,cancel
-path
=path
]\n"
3582 " use path to provide path to a character device
; default is
/dev
/tpm0
\n"
3583 " use cancel
-path to provide path to TPM
's cancel sysfs entry; if\n"
3584 " not provided it will be searched for in /sys/class/misc/tpm?/device\n"
3585 "-tpmdev emulator,id=id,chardev=dev\n"
3586 " configure the TPM device using chardev backend\n",
3589 The general form of a TPM device option is:
3591 ``-tpmdev backend,id=id[,options]``
3592 The specific backend type will determine the applicable options. The
3593 ``-tpmdev`` option creates the TPM backend and requires a
3594 ``-device`` option that specifies the TPM frontend interface model.
3596 Use ``-tpmdev help`` to print all available TPM backend types.
3598 The available backends are:
3600 ``-tpmdev passthrough,id=id,path=path,cancel-path=cancel-path``
3601 (Linux-host only) Enable access to the host's TPM
using the
3604 ``path`` specifies the path to the host
's TPM device, i.e., on a
3605 Linux host this would be ``/dev/tpm0``. ``path`` is optional and by
3606 default ``/dev/tpm0`` is used.
3608 ``cancel-path`` specifies the path to the host TPM device's sysfs
3609 entry allowing
for cancellation of an ongoing TPM command
.
3610 ``cancel
-path`` is optional and by
default QEMU will search
for the
3613 Some notes about
using the host
's TPM with the passthrough driver:
3615 The TPM device accessed by the passthrough driver must not be used
3616 by any other application on the host.
3618 Since the host's
firmware (BIOS
/UEFI
) has already initialized the
3619 TPM
, the VM
's firmware (BIOS/UEFI) will not be able to initialize
3620 the TPM again and may therefore not show a TPM-specific menu that
3621 would otherwise allow the user to configure the TPM, e.g., allow the
3622 user to enable/disable or activate/deactivate the TPM. Further, if
3623 TPM ownership is released from within a VM then the host's TPM will
3624 get disabled and deactivated
. To enable and activate the TPM again
3625 afterwards
, the host has to be rebooted and the user is required to
3626 enter the firmware
's menu to enable and activate the TPM. If the TPM
3627 is left disabled and/or deactivated most TPM commands will fail.
3629 To create a passthrough TPM use the following two options:
3633 -tpmdev passthrough,id=tpm0 -device tpm-tis,tpmdev=tpm0
3635 Note that the ``-tpmdev`` id is ``tpm0`` and is referenced by
3636 ``tpmdev=tpm0`` in the device option.
3638 ``-tpmdev emulator,id=id,chardev=dev``
3639 (Linux-host only) Enable access to a TPM emulator using Unix domain
3640 socket based chardev backend.
3642 ``chardev`` specifies the unique ID of a character device backend
3643 that provides connection to the software TPM server.
3645 To create a TPM emulator backend device with chardev socket backend:
3649 -chardev socket,id=chrtpm,path=/tmp/swtpm-sock -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0
3656 DEFHEADING(Linux/Multiboot boot specific:)
3658 When using these options, you can use a given Linux or Multiboot kernel
3659 without installing it in the disk image. It can be useful for easier
3660 testing of various kernels.
3665 DEF("kernel", HAS_ARG, QEMU_OPTION_kernel, \
3666 "-kernel bzImage use 'bzImage
' as kernel image\n", QEMU_ARCH_ALL)
3669 Use bzImage as kernel image. The kernel can be either a Linux kernel
3670 or in multiboot format.
3673 DEF("append", HAS_ARG, QEMU_OPTION_append, \
3674 "-append cmdline use 'cmdline
' as kernel command line\n", QEMU_ARCH_ALL)
3677 Use cmdline as kernel command line
3680 DEF("initrd", HAS_ARG, QEMU_OPTION_initrd, \
3681 "-initrd file use 'file
' as initial ram disk\n", QEMU_ARCH_ALL)
3684 Use file as initial ram disk.
3686 ``-initrd "file1 arg=foo,file2"``
3687 This syntax is only available with multiboot.
3689 Use file1 and file2 as modules and pass arg=foo as parameter to the
3693 DEF("dtb", HAS_ARG, QEMU_OPTION_dtb, \
3694 "-dtb file use 'file
' as device tree image\n", QEMU_ARCH_ALL)
3697 Use file as a device tree binary (dtb) image and pass it to the
3703 DEFHEADING(Debug/Expert options:)
3705 DEF("compat", HAS_ARG, QEMU_OPTION_compat,
3706 "-compat [deprecated-input=accept|reject|crash][,deprecated-output=accept|hide]\n"
3707 " Policy for handling deprecated management interfaces\n"
3708 "-compat [unstable-input=accept|reject|crash][,unstable-output=accept|hide]\n"
3709 " Policy for handling unstable management interfaces\n",
3712 ``-compat [deprecated-input=@var{input-policy}][,deprecated-output=@var{output-policy}]``
3713 Set policy for handling deprecated management interfaces (experimental):
3715 ``deprecated-input=accept`` (default)
3716 Accept deprecated commands and arguments
3717 ``deprecated-input=reject``
3718 Reject deprecated commands and arguments
3719 ``deprecated-input=crash``
3720 Crash on deprecated commands and arguments
3721 ``deprecated-output=accept`` (default)
3722 Emit deprecated command results and events
3723 ``deprecated-output=hide``
3724 Suppress deprecated command results and events
3726 Limitation: covers only syntactic aspects of QMP.
3728 ``-compat [unstable-input=@var{input-policy}][,unstable-output=@var{output-policy}]``
3729 Set policy for handling unstable management interfaces (experimental):
3731 ``unstable-input=accept`` (default)
3732 Accept unstable commands and arguments
3733 ``unstable-input=reject``
3734 Reject unstable commands and arguments
3735 ``unstable-input=crash``
3736 Crash on unstable commands and arguments
3737 ``unstable-output=accept`` (default)
3738 Emit unstable command results and events
3739 ``unstable-output=hide``
3740 Suppress unstable command results and events
3742 Limitation: covers only syntactic aspects of QMP.
3745 DEF("fw_cfg", HAS_ARG, QEMU_OPTION_fwcfg,
3746 "-fw_cfg [name=]<name>,file=<file>\n"
3747 " add named fw_cfg entry with contents from file\n"
3748 "-fw_cfg [name=]<name>,string=<str>\n"
3749 " add named fw_cfg entry with contents from string\n",
3752 ``-fw_cfg [name=]name,file=file``
3753 Add named fw\_cfg entry with contents from file file.
3755 ``-fw_cfg [name=]name,string=str``
3756 Add named fw\_cfg entry with contents from string str.
3758 The terminating NUL character of the contents of str will not be
3759 included as part of the fw\_cfg item data. To insert contents with
3760 embedded NUL characters, you have to use the file parameter.
3762 The fw\_cfg entries are passed by QEMU through to the guest.
3768 -fw_cfg name=opt/com.mycompany/blob,file=./my_blob.bin
3770 creates an fw\_cfg entry named opt/com.mycompany/blob with contents
3771 from ./my\_blob.bin.
3774 DEF("serial", HAS_ARG, QEMU_OPTION_serial, \
3775 "-serial dev redirect the serial port to char device 'dev
'\n",
3779 Redirect the virtual serial port to host character device dev. The
3780 default device is ``vc`` in graphical mode and ``stdio`` in non
3783 This option can be used several times to simulate up to 4 serial
3786 Use ``-serial none`` to disable all serial ports.
3788 Available character devices are:
3791 Virtual console. Optionally, a width and height can be given in
3798 It is also possible to specify width or height in characters:
3805 [Linux only] Pseudo TTY (a new PTY is automatically allocated)
3808 No device is allocated.
3814 Use a named character device defined with the ``-chardev``
3818 [Linux only] Use host tty, e.g. ``/dev/ttyS0``. The host serial
3819 port parameters are set according to the emulated ones.
3822 [Linux only, parallel port only] Use host parallel port N.
3823 Currently SPP and EPP parallel port features can be used.
3826 Write output to filename. No character can be read.
3829 [Unix only] standard input/output
3835 [Windows only] Use host serial port n
3837 ``udp:[remote_host]:remote_port[@[src_ip]:src_port]``
3838 This implements UDP Net Console. When remote\_host or src\_ip
3839 are not specified they default to ``0.0.0.0``. When not using a
3840 specified src\_port a random port is automatically chosen.
3842 If you just want a simple readonly console you can use
3843 ``netcat`` or ``nc``, by starting QEMU with:
3844 ``-serial udp::4555`` and nc as: ``nc -u -l -p 4555``. Any time
3845 QEMU writes something to that port it will appear in the
3848 If you plan to send characters back via netconsole or you want
3849 to stop and start QEMU a lot of times, you should have QEMU use
3850 the same source port each time by using something like ``-serial
3851 udp::4555@:4556`` to QEMU. Another approach is to use a patched
3852 version of netcat which can listen to a TCP port and send and
3853 receive characters via udp. If you have a patched version of
3854 netcat which activates telnet remote echo and single char
3855 transfer, then you can use the following options to set up a
3856 netcat redirector to allow telnet on port 5555 to access the
3860 -serial udp::4555@:4556
3863 -u -P 4555 -L 0.0.0.0:4556 -t -p 5555 -I -T
3868 ``tcp:[host]:port[,server=on|off][,wait=on|off][,nodelay=on|off][,reconnect=seconds]``
3869 The TCP Net Console has two modes of operation. It can send the
3870 serial I/O to a location or wait for a connection from a
3871 location. By default the TCP Net Console is sent to host at the
3872 port. If you use the ``server=on`` option QEMU will wait for a client
3873 socket application to connect to the port before continuing,
3874 unless the ``wait=on|off`` option was specified. The ``nodelay=on|off``
3875 option disables the Nagle buffering algorithm. The ``reconnect=on``
3876 option only applies if ``server=no`` is set, if the connection goes
3877 down it will attempt to reconnect at the given interval. If host
3878 is omitted, 0.0.0.0 is assumed. Only one TCP connection at a
3879 time is accepted. You can use ``telnet=on`` to connect to the
3880 corresponding character device.
3882 ``Example to send tcp console to 192.168.0.2 port 4444``
3883 -serial tcp:192.168.0.2:4444
3885 ``Example to listen and wait on port 4444 for connection``
3886 -serial tcp::4444,server=on
3888 ``Example to not wait and listen on ip 192.168.0.100 port 4444``
3889 -serial tcp:192.168.0.100:4444,server=on,wait=off
3891 ``telnet:host:port[,server=on|off][,wait=on|off][,nodelay=on|off]``
3892 The telnet protocol is used instead of raw tcp sockets. The
3893 options work the same as if you had specified ``-serial tcp``.
3894 The difference is that the port acts like a telnet server or
3895 client using telnet option negotiation. This will also allow you
3896 to send the MAGIC\_SYSRQ sequence if you use a telnet that
3897 supports sending the break sequence. Typically in unix telnet
3898 you do it with Control-] and then type "send break" followed by
3899 pressing the enter key.
3901 ``websocket:host:port,server=on[,wait=on|off][,nodelay=on|off]``
3902 The WebSocket protocol is used instead of raw tcp socket. The
3903 port acts as a WebSocket server. Client mode is not supported.
3905 ``unix:path[,server=on|off][,wait=on|off][,reconnect=seconds]``
3906 A unix domain socket is used instead of a tcp socket. The option
3907 works the same as if you had specified ``-serial tcp`` except
3908 the unix domain socket path is used for connections.
3911 This is a special option to allow the monitor to be multiplexed
3912 onto another serial port. The monitor is accessed with key
3913 sequence of Control-a and then pressing c. dev\_string should be
3914 any one of the serial devices specified above. An example to
3915 multiplex the monitor onto a telnet server listening on port
3918 ``-serial mon:telnet::4444,server=on,wait=off``
3920 When the monitor is multiplexed to stdio in this way, Ctrl+C
3921 will not terminate QEMU any more but will be passed to the guest
3925 Braille device. This will use BrlAPI to display the braille
3926 output on a real or fake device.
3929 Three button serial mouse. Configure the guest to use Microsoft
3933 DEF("parallel", HAS_ARG, QEMU_OPTION_parallel, \
3934 "-parallel dev redirect the parallel port to char device 'dev
'\n",
3938 Redirect the virtual parallel port to host device dev (same devices
3939 as the serial port). On Linux hosts, ``/dev/parportN`` can be used
3940 to use hardware devices connected on the corresponding host parallel
3943 This option can be used several times to simulate up to 3 parallel
3946 Use ``-parallel none`` to disable all parallel ports.
3949 DEF("monitor", HAS_ARG, QEMU_OPTION_monitor, \
3950 "-monitor dev redirect the monitor to char device 'dev
'\n",
3954 Redirect the monitor to host device dev (same devices as the serial
3955 port). The default device is ``vc`` in graphical mode and ``stdio``
3956 in non graphical mode. Use ``-monitor none`` to disable the default
3959 DEF("qmp", HAS_ARG, QEMU_OPTION_qmp, \
3960 "-qmp dev like -monitor but opens in 'control
' mode\n",
3964 Like -monitor but opens in 'control
' mode.
3966 DEF("qmp-pretty", HAS_ARG, QEMU_OPTION_qmp_pretty, \
3967 "-qmp-pretty dev like -qmp but uses pretty JSON formatting\n",
3971 Like -qmp but uses pretty JSON formatting.
3974 DEF("mon", HAS_ARG, QEMU_OPTION_mon, \
3975 "-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]\n", QEMU_ARCH_ALL)
3977 ``-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]``
3978 Setup monitor on chardev name. ``mode=control`` configures
3979 a QMP monitor (a JSON RPC-style protocol) and it is not the
3980 same as HMP, the human monitor that has a "(qemu)" prompt.
3981 ``pretty`` is only valid when ``mode=control``,
3982 turning on JSON pretty printing to ease
3983 human reading and debugging.
3986 DEF("debugcon", HAS_ARG, QEMU_OPTION_debugcon, \
3987 "-debugcon dev redirect the debug console to char device 'dev
'\n",
3991 Redirect the debug console to host device dev (same devices as the
3992 serial port). The debug console is an I/O port which is typically
3993 port 0xe9; writing to that I/O port sends output to this device. The
3994 default device is ``vc`` in graphical mode and ``stdio`` in non
3998 DEF("pidfile", HAS_ARG, QEMU_OPTION_pidfile, \
3999 "-pidfile file write PID to 'file
'\n", QEMU_ARCH_ALL)
4002 Store the QEMU process PID in file. It is useful if you launch QEMU
4006 DEF("singlestep", 0, QEMU_OPTION_singlestep, \
4007 "-singlestep always run in singlestep mode\n", QEMU_ARCH_ALL)
4010 Run the emulation in single step mode.
4013 DEF("preconfig", 0, QEMU_OPTION_preconfig, \
4014 "--preconfig pause QEMU before machine is initialized (experimental)\n",
4018 Pause QEMU for interactive configuration before the machine is
4019 created, which allows querying and configuring properties that will
4020 affect machine initialization. Use QMP command 'x
-exit
-preconfig
' to
4021 exit the preconfig state and move to the next state (i.e. run guest
4022 if -S isn't used or pause the second time
if -S is used
). This
4023 option is experimental
.
4026 DEF("S", 0, QEMU_OPTION_S
, \
4027 "-S freeze CPU at startup (use 'c' to start execution)\n",
4031 Do not start CPU at
startup (you must type
'c' in the monitor
).
4034 DEF("overcommit", HAS_ARG
, QEMU_OPTION_overcommit
,
4035 "-overcommit [mem-lock=on|off][cpu-pm=on|off]\n"
4036 " run qemu with overcommit hints\n"
4037 " mem-lock=on|off controls memory lock support (default: off)\n"
4038 " cpu-pm=on|off controls cpu power management (default: off)\n",
4041 ``
-overcommit mem
-lock
=on|off``
4043 ``
-overcommit cpu
-pm
=on|off``
4044 Run qemu with hints about host resource overcommit
. The
default is
4045 to assume that host overcommits all resources
.
4047 Locking qemu and guest memory can be enabled via ``mem
-lock
=on``
4048 (disabled by
default). This works when host memory is not
4049 overcommitted and reduces the worst
-case latency
for guest
.
4051 Guest ability to manage power state of host
cpus (increasing latency
4052 for other processes on the same host cpu
, but decreasing latency
for
4053 guest
) can be enabled via ``cpu
-pm
=on``
(disabled by
default). This
4054 works best when host CPU is not overcommitted
. When used
, host
4055 estimates of CPU cycle and power utilization will be incorrect
, not
4056 taking into account guest idle time
.
4059 DEF("gdb", HAS_ARG
, QEMU_OPTION_gdb
, \
4060 "-gdb dev accept gdb connection on 'dev'. (QEMU defaults to starting\n"
4061 " the guest without waiting for gdb to connect; use -S too\n"
4062 " if you want it to not start execution.)\n",
4066 Accept a gdb connection on device
dev (see the
:ref
:`GDB usage` chapter
4067 in the System Emulation Users Guide
). Note that
this option does not pause QEMU
4068 execution
-- if you want QEMU to not start the guest until you
4069 connect with gdb and issue a ``
continue`` command
, you will need to
4070 also pass the ``
-S`` option to QEMU
.
4072 The most usual configuration is to listen on a local TCP socket
::
4076 but you can specify other backends
; UDP
, pseudo TTY
, or even stdio
4077 are all reasonable use cases
. For example
, a stdio connection
4078 allows you to start QEMU from within gdb and establish the
4079 connection via a pipe
:
4083 (gdb
) target remote | exec |qemu_system|
-gdb stdio
...
4086 DEF("s", 0, QEMU_OPTION_s
, \
4087 "-s shorthand for -gdb tcp::" DEFAULT_GDBSTUB_PORT
"\n",
4091 Shorthand
for -gdb tcp
::1234, i
.e
. open a gdbserver on TCP port
1234
4092 (see the
:ref
:`GDB usage` chapter
in the System Emulation Users Guide
).
4095 DEF("d", HAS_ARG
, QEMU_OPTION_d
, \
4096 "-d item1,... enable logging of specified items (use '-d help' for a list of log items)\n",
4100 Enable logging of specified items
. Use
'-d help' for a list of log
4104 DEF("D", HAS_ARG
, QEMU_OPTION_D
, \
4105 "-D logfile output log to logfile (default stderr)\n",
4109 Output log
in logfile instead of to stderr
4112 DEF("dfilter", HAS_ARG
, QEMU_OPTION_DFILTER
, \
4113 "-dfilter range,.. filter debug output to range of addresses (useful for -d cpu,exec,etc..)\n",
4116 ``
-dfilter range1
[,...]``
4117 Filter debug output to that relevant to a range of target addresses
.
4118 The filter spec can be either start
+size
, start
-size or start
..end
4119 where start end and size are the addresses and sizes required
. For
4124 -dfilter
0x8000..0x8fff
,0xffffffc000080000+0x200,0xffffffc000060000-0x1000
4126 Will dump output
for any code
in the
0x1000 sized block starting at
4127 0x8000 and the
0x200 sized block starting at
0xffffffc000080000 and
4128 another
0x1000 sized block starting at
0xffffffc00005f000.
4131 DEF("seed", HAS_ARG
, QEMU_OPTION_seed
, \
4132 "-seed number seed the pseudo-random number generator\n",
4136 Force the guest to use a deterministic pseudo
-random number
4137 generator
, seeded with number
. This does not affect crypto routines
4141 DEF("L", HAS_ARG
, QEMU_OPTION_L
, \
4142 "-L path set the directory for the BIOS, VGA BIOS and keymaps\n",
4146 Set the directory
for the BIOS
, VGA BIOS and keymaps
.
4148 To list all the data directories
, use ``
-L help``
.
4151 DEF("bios", HAS_ARG
, QEMU_OPTION_bios
, \
4152 "-bios file set the filename for the BIOS\n", QEMU_ARCH_ALL
)
4155 Set the filename
for the BIOS
.
4158 DEF("enable-kvm", 0, QEMU_OPTION_enable_kvm
, \
4159 "-enable-kvm enable KVM full virtualization support\n", QEMU_ARCH_ALL
)
4162 Enable KVM full virtualization support
. This option is only
4163 available
if KVM support is enabled when compiling
.
4166 DEF("xen-domid", HAS_ARG
, QEMU_OPTION_xen_domid
,
4167 "-xen-domid id specify xen guest domain id\n", QEMU_ARCH_ALL
)
4168 DEF("xen-attach", 0, QEMU_OPTION_xen_attach
,
4169 "-xen-attach attach to existing xen domain\n"
4170 " libxl will use this when starting QEMU\n",
4172 DEF("xen-domid-restrict", 0, QEMU_OPTION_xen_domid_restrict
,
4173 "-xen-domid-restrict restrict set of available xen operations\n"
4174 " to specified domain id. (Does not affect\n"
4175 " xenpv machine type).\n",
4179 Specify xen guest domain
id (XEN only
).
4182 Attach to existing xen domain
. libxl will use
this when starting
4183 QEMU (XEN only
). Restrict set of available xen operations to
4184 specified domain
id (XEN only
).
4187 DEF("no-reboot", 0, QEMU_OPTION_no_reboot
, \
4188 "-no-reboot exit instead of rebooting\n", QEMU_ARCH_ALL
)
4191 Exit instead of rebooting
.
4194 DEF("no-shutdown", 0, QEMU_OPTION_no_shutdown
, \
4195 "-no-shutdown stop before shutdown\n", QEMU_ARCH_ALL
)
4198 Don
't exit QEMU on guest shutdown, but instead only stop the
4199 emulation. This allows for instance switching to monitor to commit
4200 changes to the disk image.
4203 DEF("action", HAS_ARG, QEMU_OPTION_action,
4204 "-action reboot=reset|shutdown\n"
4205 " action when guest reboots [default=reset]\n"
4206 "-action shutdown=poweroff|pause\n"
4207 " action when guest shuts down [default=poweroff]\n"
4208 "-action panic=pause|shutdown|none\n"
4209 " action when guest panics [default=shutdown]\n"
4210 "-action watchdog=reset|shutdown|poweroff|inject-nmi|pause|debug|none\n"
4211 " action when watchdog fires [default=reset]\n",
4214 ``-action event=action``
4215 The action parameter serves to modify QEMU's
default behavior when
4216 certain guest events occur
. It provides a generic method
for specifying the
4217 same behaviors that are modified by the ``
-no
-reboot`` and ``
-no
-shutdown``
4222 ``
-action panic
=none``
4223 ``
-action reboot
=shutdown
,shutdown
=pause``
4224 ``
-watchdog i6300esb
-action watchdog
=pause``
4228 DEF("loadvm", HAS_ARG
, QEMU_OPTION_loadvm
, \
4229 "-loadvm [tag|id]\n" \
4230 " start right away with a saved state (loadvm in monitor)\n",
4234 Start right away with a saved
state (``loadvm``
in monitor
)
4238 DEF("daemonize", 0, QEMU_OPTION_daemonize
, \
4239 "-daemonize daemonize QEMU after initializing\n", QEMU_ARCH_ALL
)
4243 Daemonize the QEMU process after initialization
. QEMU will not
4244 detach from standard IO until it is ready to receive connections on
4245 any of its devices
. This option is a useful way
for external
4246 programs to launch QEMU without having to cope with initialization
4250 DEF("option-rom", HAS_ARG
, QEMU_OPTION_option_rom
, \
4251 "-option-rom rom load a file, rom, into the option ROM space\n",
4254 ``
-option
-rom file``
4255 Load the contents of file as an option ROM
. This option is useful to
4256 load things like EtherBoot
.
4259 DEF("rtc", HAS_ARG
, QEMU_OPTION_rtc
, \
4260 "-rtc [base=utc|localtime|<datetime>][,clock=host|rt|vm][,driftfix=none|slew]\n" \
4261 " set the RTC base and clock, enable drift fix for clock ticks (x86 only)\n",
4265 ``
-rtc
[base
=utc|localtime|datetime
][,clock
=host|rt|vm
][,driftfix
=none|slew
]``
4266 Specify ``base`` as ``utc`` or ``localtime`` to let the RTC start at
4267 the current UTC or local time
, respectively
. ``localtime`` is
4268 required
for correct date
in MS
-DOS or Windows
. To start at a
4269 specific point
in time
, provide datetime
in the format
4270 ``
2006-06-17T16
:01:21`` or ``
2006-06-17``
. The
default base is UTC
.
4272 By
default the RTC is driven by the host system time
. This allows
4273 using of the RTC as accurate reference clock inside the guest
,
4274 specifically
if the host time is smoothly following an accurate
4275 external reference clock
, e
.g
. via NTP
. If you want to isolate the
4276 guest time from the host
, you can set ``clock`` to ``rt`` instead
,
4277 which provides a host monotonic clock
if host support it
. To even
4278 prevent the RTC from progressing during suspension
, you can set
4279 ``clock`` to ``vm``
(virtual clock
). '\ ``clock=vm``\ ' is
4280 recommended especially
in icount mode
in order to preserve
4281 determinism
; however
, note that
in icount mode the speed of the
4282 virtual clock is variable and can
in general differ from the host
4285 Enable ``driftfix``
(i386 targets only
) if you experience time drift
4286 problems
, specifically with Windows
' ACPI HAL. This option will try
4287 to figure out how many timer interrupts were not processed by the
4288 Windows guest and will re-inject them.
4291 DEF("icount", HAS_ARG, QEMU_OPTION_icount, \
4292 "-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=<filename>[,rrsnapshot=<snapshot>]]\n" \
4293 " enable virtual instruction counter with 2^N clock ticks per\n" \
4294 " instruction, enable aligning the host and virtual clocks\n" \
4295 " or disable real time cpu sleeping, and optionally enable\n" \
4296 " record-and-replay mode\n", QEMU_ARCH_ALL)
4298 ``-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=filename[,rrsnapshot=snapshot]]``
4299 Enable virtual instruction counter. The virtual cpu will execute one
4300 instruction every 2^N ns of virtual time. If ``auto`` is specified
4301 then the virtual cpu speed will be automatically adjusted to keep
4302 virtual time within a few seconds of real time.
4304 Note that while this option can give deterministic behavior, it does
4305 not provide cycle accurate emulation. Modern CPUs contain
4306 superscalar out of order cores with complex cache hierarchies. The
4307 number of instructions executed often has little or no correlation
4308 with actual performance.
4310 When the virtual cpu is sleeping, the virtual time will advance at
4311 default speed unless ``sleep=on`` is specified. With
4312 ``sleep=on``, the virtual time will jump to the next timer
4313 deadline instantly whenever the virtual cpu goes to sleep mode and
4314 will not advance if no timer is enabled. This behavior gives
4315 deterministic execution times from the guest point of view.
4316 The default if icount is enabled is ``sleep=off``.
4317 ``sleep=on`` cannot be used together with either ``shift=auto``
4320 ``align=on`` will activate the delay algorithm which will try to
4321 synchronise the host clock and the virtual clock. The goal is to
4322 have a guest running at the real frequency imposed by the shift
4323 option. Whenever the guest clock is behind the host clock and if
4324 ``align=on`` is specified then we print a message to the user to
4325 inform about the delay. Currently this option does not work when
4326 ``shift`` is ``auto``. Note: The sync algorithm will work for those
4327 shift values for which the guest clock runs ahead of the host clock.
4328 Typically this happens when the shift value is high (how high
4329 depends on the host machine). The default if icount is enabled
4332 When the ``rr`` option is specified deterministic record/replay is
4333 enabled. The ``rrfile=`` option must also be provided to
4334 specify the path to the replay log. In record mode data is written
4335 to this file, and in replay mode it is read back.
4336 If the ``rrsnapshot`` option is given then it specifies a VM snapshot
4337 name. In record mode, a new VM snapshot with the given name is created
4338 at the start of execution recording. In replay mode this option
4339 specifies the snapshot name used to load the initial VM state.
4342 DEF("watchdog", HAS_ARG, QEMU_OPTION_watchdog, \
4343 "-watchdog model\n" \
4344 " enable virtual hardware watchdog [default=none]\n",
4348 Create a virtual hardware watchdog device. Once enabled (by a guest
4349 action), the watchdog must be periodically polled by an agent inside
4350 the guest or else the guest will be restarted. Choose a model for
4351 which your guest has drivers.
4353 The model is the model of hardware watchdog to emulate. Use
4354 ``-watchdog help`` to list available hardware models. Only one
4355 watchdog can be enabled for a guest.
4357 The following models may be available:
4360 iBASE 700 is a very simple ISA watchdog with a single timer.
4363 Intel 6300ESB I/O controller hub is a much more featureful
4364 PCI-based dual-timer watchdog.
4367 A virtual watchdog for s390x backed by the diagnose 288
4368 hypercall (currently KVM only).
4371 DEF("watchdog-action", HAS_ARG, QEMU_OPTION_watchdog_action, \
4372 "-watchdog-action reset|shutdown|poweroff|inject-nmi|pause|debug|none\n" \
4373 " action when watchdog fires [default=reset]\n",
4376 ``-watchdog-action action``
4377 The action controls what QEMU will do when the watchdog timer
4378 expires. The default is ``reset`` (forcefully reset the guest).
4379 Other possible actions are: ``shutdown`` (attempt to gracefully
4380 shutdown the guest), ``poweroff`` (forcefully poweroff the guest),
4381 ``inject-nmi`` (inject a NMI into the guest), ``pause`` (pause the
4382 guest), ``debug`` (print a debug message and continue), or ``none``
4385 Note that the ``shutdown`` action requires that the guest responds
4386 to ACPI signals, which it may not be able to do in the sort of
4387 situations where the watchdog would have expired, and thus
4388 ``-watchdog-action shutdown`` is not recommended for production use.
4392 ``-watchdog i6300esb -watchdog-action pause``; \ ``-watchdog ib700``
4396 DEF("echr", HAS_ARG, QEMU_OPTION_echr, \
4397 "-echr chr set terminal escape character instead of ctrl-a\n",
4400 ``-echr numeric_ascii_value``
4401 Change the escape character used for switching to the monitor when
4402 using monitor and serial sharing. The default is ``0x01`` when using
4403 the ``-nographic`` option. ``0x01`` is equal to pressing
4404 ``Control-a``. You can select a different character from the ascii
4405 control keys where 1 through 26 map to Control-a through Control-z.
4406 For instance you could use the either of the following to change the
4407 escape character to Control-t.
4409 ``-echr 0x14``; \ ``-echr 20``
4413 DEF("incoming", HAS_ARG, QEMU_OPTION_incoming, \
4414 "-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]\n" \
4415 "-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]\n" \
4416 "-incoming unix:socketpath\n" \
4417 " prepare for incoming migration, listen on\n" \
4418 " specified protocol and socket address\n" \
4419 "-incoming fd:fd\n" \
4420 "-incoming exec:cmdline\n" \
4421 " accept incoming migration on given file descriptor\n" \
4422 " or from given external command\n" \
4423 "-incoming defer\n" \
4424 " wait for the URI to be specified via migrate_incoming\n",
4427 ``-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]``
4429 ``-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]``
4430 Prepare for incoming migration, listen on a given tcp port.
4432 ``-incoming unix:socketpath``
4433 Prepare for incoming migration, listen on a given unix socket.
4436 Accept incoming migration from a given filedescriptor.
4438 ``-incoming exec:cmdline``
4439 Accept incoming migration as an output from specified external
4443 Wait for the URI to be specified via migrate\_incoming. The monitor
4444 can be used to change settings (such as migration parameters) prior
4445 to issuing the migrate\_incoming to allow the migration to begin.
4448 DEF("only-migratable", 0, QEMU_OPTION_only_migratable, \
4449 "-only-migratable allow only migratable devices\n", QEMU_ARCH_ALL)
4451 ``-only-migratable``
4452 Only allow migratable devices. Devices will not be allowed to enter
4453 an unmigratable state.
4456 DEF("nodefaults", 0, QEMU_OPTION_nodefaults, \
4457 "-nodefaults don't create
default devices
\n", QEMU_ARCH_ALL)
4460 Don't create default devices. Normally, QEMU sets the default
4461 devices like serial port, parallel port, virtual console, monitor
4462 device, VGA adapter, floppy and CD-ROM drive and others. The
4463 ``-nodefaults`` option will disable all those default devices.
4467 DEF("chroot
", HAS_ARG, QEMU_OPTION_chroot, \
4468 "-chroot dir chroot to dir just before starting the VM
\n",
4473 Immediately before starting guest execution, chroot to the specified
4474 directory. Especially useful in combination with -runas.
4478 DEF("runas
", HAS_ARG, QEMU_OPTION_runas, \
4479 "-runas user change to user id user just before starting the VM
\n" \
4480 " user can be numeric uid
:gid instead
\n",
4485 Immediately before starting guest execution, drop root privileges,
4486 switching to the specified user.
4489 DEF("prom
-env
", HAS_ARG, QEMU_OPTION_prom_env,
4490 "-prom
-env variable
=value
\n"
4491 " set OpenBIOS nvram variables
\n",
4492 QEMU_ARCH_PPC | QEMU_ARCH_SPARC)
4494 ``-prom-env variable=value``
4495 Set OpenBIOS nvram variable to given value (PPC, SPARC only).
4499 qemu-system-sparc -prom-env 'auto-boot?=false' \
4500 -prom-env 'boot-device=sd(0,2,0):d' -prom-env 'boot-args=linux single'
4504 qemu-system-ppc -prom-env 'auto-boot?=false' \
4505 -prom-env 'boot-device=hd:2,\yaboot' \
4506 -prom-env 'boot-args=conf=hd:2,\yaboot.conf'
4508 DEF("semihosting
", 0, QEMU_OPTION_semihosting,
4509 "-semihosting semihosting mode
\n",
4510 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4511 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4514 Enable semihosting mode (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V only).
4516 Note that this allows guest direct access to the host filesystem, so
4517 should only be used with a trusted guest OS.
4519 See the -semihosting-config option documentation for further
4520 information about the facilities this enables.
4522 DEF("semihosting
-config
", HAS_ARG, QEMU_OPTION_semihosting_config,
4523 "-semihosting
-config
[enable
=on|off
][,target
=native|gdb|auto
][,chardev
=id
][,arg
=str
[,...]]\n" \
4524 " semihosting configuration
\n",
4525 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4526 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4528 ``-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,arg=str[,...]]``
4529 Enable and configure semihosting (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V
4532 Note that this allows guest direct access to the host filesystem, so
4533 should only be used with a trusted guest OS.
4535 On Arm this implements the standard semihosting API, version 2.0.
4537 On M68K this implements the "ColdFire GDB
" interface used by
4540 Xtensa semihosting provides basic file IO calls, such as
4541 open/read/write/seek/select. Tensilica baremetal libc for ISS and
4542 linux platform "sim
" use this interface.
4544 On RISC-V this implements the standard semihosting API, version 0.2.
4546 ``target=native|gdb|auto``
4547 Defines where the semihosting calls will be addressed, to QEMU
4548 (``native``) or to GDB (``gdb``). The default is ``auto``, which
4549 means ``gdb`` during debug sessions and ``native`` otherwise.
4552 Send the output to a chardev backend output for native or auto
4553 output when not in gdb
4555 ``arg=str1,arg=str2,...``
4556 Allows the user to pass input arguments, and can be used
4557 multiple times to build up a list. The old-style
4558 ``-kernel``/``-append`` method of passing a command line is
4559 still supported for backward compatibility. If both the
4560 ``--semihosting-config arg`` and the ``-kernel``/``-append`` are
4561 specified, the former is passed to semihosting as it always
4564 DEF("old
-param
", 0, QEMU_OPTION_old_param,
4565 "-old
-param old param mode
\n", QEMU_ARCH_ARM)
4568 Old param mode (ARM only).
4571 DEF("sandbox
", HAS_ARG, QEMU_OPTION_sandbox, \
4572 "-sandbox on
[,obsolete
=allow|deny
][,elevateprivileges
=allow|deny|children
]\n" \
4573 " [,spawn
=allow|deny
][,resourcecontrol
=allow|deny
]\n" \
4574 " Enable seccomp mode
2 system call
filter (default 'off').\n" \
4575 " use
'obsolete' to allow obsolete system calls that are provided
\n" \
4576 " by the kernel
, but typically no longer used by modern
\n" \
4577 " C library implementations
.\n" \
4578 " use
'elevateprivileges' to allow or deny the QEMU process ability
\n" \
4579 " to elevate privileges
using set
*uid|gid system calls
.\n" \
4580 " The value
'children' will deny set
*uid|gid system calls
for\n" \
4581 " main QEMU process but will allow forks and execves to run unprivileged
\n" \
4582 " use
'spawn' to avoid QEMU to spawn
new threads or processes by
\n" \
4583 " blocking
*fork and execve
\n" \
4584 " use
'resourcecontrol' to disable process affinity and schedular priority
\n",
4587 ``-sandbox arg[,obsolete=string][,elevateprivileges=string][,spawn=string][,resourcecontrol=string]``
4588 Enable Seccomp mode 2 system call filter. 'on' will enable syscall
4589 filtering and 'off' will disable it. The default is 'off'.
4592 Enable Obsolete system calls
4594 ``elevateprivileges=string``
4595 Disable set\*uid\|gid system calls
4598 Disable \*fork and execve
4600 ``resourcecontrol=string``
4601 Disable process affinity and schedular priority
4604 DEF("readconfig
", HAS_ARG, QEMU_OPTION_readconfig,
4605 "-readconfig
<file
>\n", QEMU_ARCH_ALL)
4607 ``-readconfig file``
4608 Read device configuration from file. This approach is useful when
4609 you want to spawn QEMU process with many command line options but
4610 you don't want to exceed the command line character limit.
4612 DEF("writeconfig
", HAS_ARG, QEMU_OPTION_writeconfig,
4613 "-writeconfig
<file
>\n"
4614 " read
/write config
file (deprecated
)\n", QEMU_ARCH_ALL)
4618 DEF("no
-user
-config
", 0, QEMU_OPTION_nouserconfig,
4620 " do not load
default user
-provided config files at startup
\n",
4624 The ``-no-user-config`` option makes QEMU not load any of the
4625 user-provided config files on sysconfdir.
4628 DEF("trace", HAS_ARG, QEMU_OPTION_trace,
4629 "-trace [[enable
=]<pattern
>][,events
=<file
>][,file
=<file
>]\n"
4630 " specify tracing options
\n",
4633 ``-trace [[enable=]pattern][,events=file][,file=file]``
4634 .. include:: ../qemu-option-trace.rst.inc
4637 DEF("plugin
", HAS_ARG, QEMU_OPTION_plugin,
4638 "-plugin
[file
=]<file
>[,<argname
>=<argvalue
>]\n"
4642 ``-plugin file=file[,argname=argvalue]``
4646 Load the given plugin from a shared library file.
4648 ``argname=argvalue``
4649 Argument passed to the plugin. (Can be given multiple times.)
4653 DEF("qtest
", HAS_ARG, QEMU_OPTION_qtest, "", QEMU_ARCH_ALL)
4654 DEF("qtest
-log
", HAS_ARG, QEMU_OPTION_qtest_log, "", QEMU_ARCH_ALL)
4657 DEF("enable
-fips
", 0, QEMU_OPTION_enablefips,
4658 "-enable
-fips enable FIPS
140-2 compliance
\n",
4663 Enable FIPS 140-2 compliance mode.
4666 DEF("msg
", HAS_ARG, QEMU_OPTION_msg,
4667 "-msg
[timestamp
[=on|off
]][,guest
-name
=[on|off
]]\n"
4668 " control error message format
\n"
4669 " timestamp
=on enables
timestamps (default: off
)\n"
4670 " guest
-name
=on enables guest name prefix but only
if\n"
4671 " -name guest option is
set (default: off
)\n",
4674 ``-msg [timestamp[=on|off]][,guest-name[=on|off]]``
4675 Control error message format.
4677 ``timestamp=on|off``
4678 Prefix messages with a timestamp. Default is off.
4680 ``guest-name=on|off``
4681 Prefix messages with guest name but only if -name guest option is set
4682 otherwise the option is ignored. Default is off.
4685 DEF("dump
-vmstate
", HAS_ARG, QEMU_OPTION_dump_vmstate,
4686 "-dump
-vmstate
<file
>\n"
4687 " Output vmstate information
in JSON format to file
.\n"
4688 " Use the scripts
/vmstate
-static-checker
.py file to
\n"
4689 " check
for possible regressions
in migration code
\n"
4690 " by comparing two such vmstate dumps
.\n",
4693 ``-dump-vmstate file``
4694 Dump json-encoded vmstate information for current machine type to
4698 DEF("enable
-sync
-profile
", 0, QEMU_OPTION_enable_sync_profile,
4699 "-enable
-sync
-profile
\n"
4700 " enable synchronization profiling
\n",
4703 ``-enable-sync-profile``
4704 Enable synchronization profiling.
4709 DEFHEADING(Generic object creation:)
4711 DEF("object
", HAS_ARG, QEMU_OPTION_object,
4712 "-object TYPENAME
[,PROP1
=VALUE1
,...]\n"
4713 " create a
new object of type TYPENAME setting properties
\n"
4714 " in the order they are specified
. Note that the
'id'\n"
4715 " property must be set
. These objects are placed
in the
\n"
4716 " '/objects' path
.\n",
4719 ``-object typename[,prop1=value1,...]``
4720 Create a new object of type typename setting properties in the order
4721 they are specified. Note that the 'id' property must be set. These
4722 objects are placed in the '/objects' path.
4724 ``-object memory-backend-file,id=id,size=size,mem-path=dir,share=on|off,discard-data=on|off,merge=on|off,dump=on|off,prealloc=on|off,host-nodes=host-nodes,policy=default|preferred|bind|interleave,align=align,readonly=on|off``
4725 Creates a memory file backend object, which can be used to back
4726 the guest RAM with huge pages.
4728 The ``id`` parameter is a unique ID that will be used to
4729 reference this memory region in other parameters, e.g. ``-numa``,
4730 ``-device nvdimm``, etc.
4732 The ``size`` option provides the size of the memory region, and
4733 accepts common suffixes, e.g. ``500M``.
4735 The ``mem-path`` provides the path to either a shared memory or
4736 huge page filesystem mount.
4738 The ``share`` boolean option determines whether the memory
4739 region is marked as private to QEMU, or shared. The latter
4740 allows a co-operating external process to access the QEMU memory
4743 The ``share`` is also required for pvrdma devices due to
4744 limitations in the RDMA API provided by Linux.
4746 Setting share=on might affect the ability to configure NUMA
4747 bindings for the memory backend under some circumstances, see
4748 Documentation/vm/numa\_memory\_policy.txt on the Linux kernel
4749 source tree for additional details.
4751 Setting the ``discard-data`` boolean option to on indicates that
4752 file contents can be destroyed when QEMU exits, to avoid
4753 unnecessarily flushing data to the backing file. Note that
4754 ``discard-data`` is only an optimization, and QEMU might not
4755 discard file contents if it aborts unexpectedly or is terminated
4758 The ``merge`` boolean option enables memory merge, also known as
4759 MADV\_MERGEABLE, so that Kernel Samepage Merging will consider
4760 the pages for memory deduplication.
4762 Setting the ``dump`` boolean option to off excludes the memory
4763 from core dumps. This feature is also known as MADV\_DONTDUMP.
4765 The ``prealloc`` boolean option enables memory preallocation.
4767 The ``host-nodes`` option binds the memory range to a list of
4770 The ``policy`` option sets the NUMA policy to one of the
4777 prefer the given host node list for allocation
4780 restrict memory allocation to the given host node list
4783 interleave memory allocations across the given host node
4786 The ``align`` option specifies the base address alignment when
4787 QEMU mmap(2) ``mem-path``, and accepts common suffixes, eg
4788 ``2M``. Some backend store specified by ``mem-path`` requires an
4789 alignment different than the default one used by QEMU, eg the
4790 device DAX /dev/dax0.0 requires 2M alignment rather than 4K. In
4791 such cases, users can specify the required alignment via this
4794 The ``pmem`` option specifies whether the backing file specified
4795 by ``mem-path`` is in host persistent memory that can be
4796 accessed using the SNIA NVM programming model (e.g. Intel
4797 NVDIMM). If ``pmem`` is set to 'on', QEMU will take necessary
4798 operations to guarantee the persistence of its own writes to
4799 ``mem-path`` (e.g. in vNVDIMM label emulation and live
4800 migration). Also, we will map the backend-file with MAP\_SYNC
4801 flag, which ensures the file metadata is in sync for
4802 ``mem-path`` in case of host crash or a power failure. MAP\_SYNC
4803 requires support from both the host kernel (since Linux kernel
4804 4.15) and the filesystem of ``mem-path`` mounted with DAX
4807 The ``readonly`` option specifies whether the backing file is opened
4808 read-only or read-write (default).
4810 ``-object memory-backend-ram,id=id,merge=on|off,dump=on|off,share=on|off,prealloc=on|off,size=size,host-nodes=host-nodes,policy=default|preferred|bind|interleave``
4811 Creates a memory backend object, which can be used to back the
4812 guest RAM. Memory backend objects offer more control than the
4813 ``-m`` option that is traditionally used to define guest RAM.
4814 Please refer to ``memory-backend-file`` for a description of the
4817 ``-object memory-backend-memfd,id=id,merge=on|off,dump=on|off,share=on|off,prealloc=on|off,size=size,host-nodes=host-nodes,policy=default|preferred|bind|interleave,seal=on|off,hugetlb=on|off,hugetlbsize=size``
4818 Creates an anonymous memory file backend object, which allows
4819 QEMU to share the memory with an external process (e.g. when
4820 using vhost-user). The memory is allocated with memfd and
4821 optional sealing. (Linux only)
4823 The ``seal`` option creates a sealed-file, that will block
4824 further resizing the memory ('on' by default).
4826 The ``hugetlb`` option specify the file to be created resides in
4827 the hugetlbfs filesystem (since Linux 4.14). Used in conjunction
4828 with the ``hugetlb`` option, the ``hugetlbsize`` option specify
4829 the hugetlb page size on systems that support multiple hugetlb
4830 page sizes (it must be a power of 2 value supported by the
4833 In some versions of Linux, the ``hugetlb`` option is
4834 incompatible with the ``seal`` option (requires at least Linux
4837 Please refer to ``memory-backend-file`` for a description of the
4840 The ``share`` boolean option is on by default with memfd.
4842 ``-object rng-builtin,id=id``
4843 Creates a random number generator backend which obtains entropy
4844 from QEMU builtin functions. The ``id`` parameter is a unique ID
4845 that will be used to reference this entropy backend from the
4846 ``virtio-rng`` device. By default, the ``virtio-rng`` device
4847 uses this RNG backend.
4849 ``-object rng-random,id=id,filename=/dev/random``
4850 Creates a random number generator backend which obtains entropy
4851 from a device on the host. The ``id`` parameter is a unique ID
4852 that will be used to reference this entropy backend from the
4853 ``virtio-rng`` device. The ``filename`` parameter specifies
4854 which file to obtain entropy from and if omitted defaults to
4857 ``-object rng-egd,id=id,chardev=chardevid``
4858 Creates a random number generator backend which obtains entropy
4859 from an external daemon running on the host. The ``id``
4860 parameter is a unique ID that will be used to reference this
4861 entropy backend from the ``virtio-rng`` device. The ``chardev``
4862 parameter is the unique ID of a character device backend that
4863 provides the connection to the RNG daemon.
4865 ``-object tls-creds-anon,id=id,endpoint=endpoint,dir=/path/to/cred/dir,verify-peer=on|off``
4866 Creates a TLS anonymous credentials object, which can be used to
4867 provide TLS support on network backends. The ``id`` parameter is
4868 a unique ID which network backends will use to access the
4869 credentials. The ``endpoint`` is either ``server`` or ``client``
4870 depending on whether the QEMU network backend that uses the
4871 credentials will be acting as a client or as a server. If
4872 ``verify-peer`` is enabled (the default) then once the handshake
4873 is completed, the peer credentials will be verified, though this
4874 is a no-op for anonymous credentials.
4876 The dir parameter tells QEMU where to find the credential files.
4877 For server endpoints, this directory may contain a file
4878 dh-params.pem providing diffie-hellman parameters to use for the
4879 TLS server. If the file is missing, QEMU will generate a set of
4880 DH parameters at startup. This is a computationally expensive
4881 operation that consumes random pool entropy, so it is
4882 recommended that a persistent set of parameters be generated
4885 ``-object tls-creds-psk,id=id,endpoint=endpoint,dir=/path/to/keys/dir[,username=username]``
4886 Creates a TLS Pre-Shared Keys (PSK) credentials object, which
4887 can be used to provide TLS support on network backends. The
4888 ``id`` parameter is a unique ID which network backends will use
4889 to access the credentials. The ``endpoint`` is either ``server``
4890 or ``client`` depending on whether the QEMU network backend that
4891 uses the credentials will be acting as a client or as a server.
4892 For clients only, ``username`` is the username which will be
4893 sent to the server. If omitted it defaults to "qemu
".
4895 The dir parameter tells QEMU where to find the keys file. It is
4896 called "dir
/keys
.psk
" and contains "username
:key
" pairs. This
4897 file can most easily be created using the GnuTLS ``psktool``
4900 For server endpoints, dir may also contain a file dh-params.pem
4901 providing diffie-hellman parameters to use for the TLS server.
4902 If the file is missing, QEMU will generate a set of DH
4903 parameters at startup. This is a computationally expensive
4904 operation that consumes random pool entropy, so it is
4905 recommended that a persistent set of parameters be generated up
4908 ``-object tls-creds-x509,id=id,endpoint=endpoint,dir=/path/to/cred/dir,priority=priority,verify-peer=on|off,passwordid=id``
4909 Creates a TLS anonymous credentials object, which can be used to
4910 provide TLS support on network backends. The ``id`` parameter is
4911 a unique ID which network backends will use to access the
4912 credentials. The ``endpoint`` is either ``server`` or ``client``
4913 depending on whether the QEMU network backend that uses the
4914 credentials will be acting as a client or as a server. If
4915 ``verify-peer`` is enabled (the default) then once the handshake
4916 is completed, the peer credentials will be verified. With x509
4917 certificates, this implies that the clients must be provided
4918 with valid client certificates too.
4920 The dir parameter tells QEMU where to find the credential files.
4921 For server endpoints, this directory may contain a file
4922 dh-params.pem providing diffie-hellman parameters to use for the
4923 TLS server. If the file is missing, QEMU will generate a set of
4924 DH parameters at startup. This is a computationally expensive
4925 operation that consumes random pool entropy, so it is
4926 recommended that a persistent set of parameters be generated
4929 For x509 certificate credentials the directory will contain
4930 further files providing the x509 certificates. The certificates
4931 must be stored in PEM format, in filenames ca-cert.pem,
4932 ca-crl.pem (optional), server-cert.pem (only servers),
4933 server-key.pem (only servers), client-cert.pem (only clients),
4934 and client-key.pem (only clients).
4936 For the server-key.pem and client-key.pem files which contain
4937 sensitive private keys, it is possible to use an encrypted
4938 version by providing the passwordid parameter. This provides the
4939 ID of a previously created ``secret`` object containing the
4940 password for decryption.
4942 The priority parameter allows to override the global default
4943 priority used by gnutls. This can be useful if the system
4944 administrator needs to use a weaker set of crypto priorities for
4945 QEMU without potentially forcing the weakness onto all
4946 applications. Or conversely if one wants wants a stronger
4947 default for QEMU than for all other applications, they can do
4948 this through this parameter. Its format is a gnutls priority
4949 string as described at
4950 https://gnutls.org/manual/html_node/Priority-Strings.html.
4952 ``-object tls-cipher-suites,id=id,priority=priority``
4953 Creates a TLS cipher suites object, which can be used to control
4954 the TLS cipher/protocol algorithms that applications are permitted
4957 The ``id`` parameter is a unique ID which frontends will use to
4958 access the ordered list of permitted TLS cipher suites from the
4961 The ``priority`` parameter allows to override the global default
4962 priority used by gnutls. This can be useful if the system
4963 administrator needs to use a weaker set of crypto priorities for
4964 QEMU without potentially forcing the weakness onto all
4965 applications. Or conversely if one wants wants a stronger
4966 default for QEMU than for all other applications, they can do
4967 this through this parameter. Its format is a gnutls priority
4968 string as described at
4969 https://gnutls.org/manual/html_node/Priority-Strings.html.
4971 An example of use of this object is to control UEFI HTTPS Boot.
4972 The tls-cipher-suites object exposes the ordered list of permitted
4973 TLS cipher suites from the host side to the guest firmware, via
4974 fw_cfg. The list is represented as an array of IANA_TLS_CIPHER
4975 objects. The firmware uses the IANA_TLS_CIPHER array for configuring
4978 In the following example, the priority at which the host-side policy
4979 is retrieved is given by the ``priority`` property.
4980 Given that QEMU uses GNUTLS, ``priority=@SYSTEM`` may be used to
4981 refer to /etc/crypto-policies/back-ends/gnutls.config.
4986 -object tls-cipher-suites,id=mysuite0,priority=@SYSTEM \\
4987 -fw_cfg name=etc/edk2/https/ciphers,gen_id=mysuite0
4989 ``-object filter-buffer,id=id,netdev=netdevid,interval=t[,queue=all|rx|tx][,status=on|off][,position=head|tail|id=<id>][,insert=behind|before]``
4990 Interval t can't be 0, this filter batches the packet delivery:
4991 all packets arriving in a given interval on netdev netdevid are
4992 delayed until the end of the interval. Interval is in
4993 microseconds. ``status`` is optional that indicate whether the
4994 netfilter is on (enabled) or off (disabled), the default status
4995 for netfilter will be 'on'.
4997 queue all\|rx\|tx is an option that can be applied to any
5000 ``all``: the filter is attached both to the receive and the
5001 transmit queue of the netdev (default).
5003 ``rx``: the filter is attached to the receive queue of the
5004 netdev, where it will receive packets sent to the netdev.
5006 ``tx``: the filter is attached to the transmit queue of the
5007 netdev, where it will receive packets sent by the netdev.
5009 position head\|tail\|id=<id> is an option to specify where the
5010 filter should be inserted in the filter list. It can be applied
5013 ``head``: the filter is inserted at the head of the filter list,
5014 before any existing filters.
5016 ``tail``: the filter is inserted at the tail of the filter list,
5017 behind any existing filters (default).
5019 ``id=<id>``: the filter is inserted before or behind the filter
5020 specified by <id>, see the insert option below.
5022 insert behind\|before is an option to specify where to insert
5023 the new filter relative to the one specified with
5024 position=id=<id>. It can be applied to any netfilter.
5026 ``before``: insert before the specified filter.
5028 ``behind``: insert behind the specified filter (default).
5030 ``-object filter-mirror,id=id,netdev=netdevid,outdev=chardevid,queue=all|rx|tx[,vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5031 filter-mirror on netdev netdevid,mirror net packet to
5032 chardevchardevid, if it has the vnet\_hdr\_support flag,
5033 filter-mirror will mirror packet with vnet\_hdr\_len.
5035 ``-object filter-redirector,id=id,netdev=netdevid,indev=chardevid,outdev=chardevid,queue=all|rx|tx[,vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5036 filter-redirector on netdev netdevid,redirect filter's net
5037 packet to chardev chardevid,and redirect indev's packet to
5038 filter.if it has the vnet\_hdr\_support flag, filter-redirector
5039 will redirect packet with vnet\_hdr\_len. Create a
5040 filter-redirector we need to differ outdev id from indev id, id
5041 can not be the same. we can just use indev or outdev, but at
5042 least one of indev or outdev need to be specified.
5044 ``-object filter-rewriter,id=id,netdev=netdevid,queue=all|rx|tx,[vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5045 Filter-rewriter is a part of COLO project.It will rewrite tcp
5046 packet to secondary from primary to keep secondary tcp
5047 connection,and rewrite tcp packet to primary from secondary make
5048 tcp packet can be handled by client.if it has the
5049 vnet\_hdr\_support flag, we can parse packet with vnet header.
5051 usage: colo secondary: -object
5052 filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 -object
5053 filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 -object
5054 filter-rewriter,id=rew0,netdev=hn0,queue=all
5056 ``-object filter-dump,id=id,netdev=dev[,file=filename][,maxlen=len][,position=head|tail|id=<id>][,insert=behind|before]``
5057 Dump the network traffic on netdev dev to the file specified by
5058 filename. At most len bytes (64k by default) per packet are
5059 stored. The file format is libpcap, so it can be analyzed with
5060 tools such as tcpdump or Wireshark.
5062 ``-object colo-compare,id=id,primary_in=chardevid,secondary_in=chardevid,outdev=chardevid,iothread=id[,vnet_hdr_support][,notify_dev=id][,compare_timeout=@var{ms}][,expired_scan_cycle=@var{ms}][,max_queue_size=@var{size}]``
5063 Colo-compare gets packet from primary\_in chardevid and
5064 secondary\_in, then compare whether the payload of primary packet
5065 and secondary packet are the same. If same, it will output
5066 primary packet to out\_dev, else it will notify COLO-framework to do
5067 checkpoint and send primary packet to out\_dev. In order to
5068 improve efficiency, we need to put the task of comparison in
5069 another iothread. If it has the vnet\_hdr\_support flag,
5070 colo compare will send/recv packet with vnet\_hdr\_len.
5071 The compare\_timeout=@var{ms} determines the maximum time of the
5072 colo-compare hold the packet. The expired\_scan\_cycle=@var{ms}
5073 is to set the period of scanning expired primary node network packets.
5074 The max\_queue\_size=@var{size} is to set the max compare queue
5075 size depend on user environment.
5076 If user want to use Xen COLO, need to add the notify\_dev to
5077 notify Xen colo-frame to do checkpoint.
5079 COLO-compare must be used with the help of filter-mirror,
5080 filter-redirector and filter-rewriter.
5087 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5088 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5089 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5090 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5091 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5092 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5093 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5094 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5095 -object iothread,id=iothread1
5096 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5097 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5098 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5099 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,iothread=iothread1
5102 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5103 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5104 -chardev socket,id=red0,host=3.3.3.3,port=9003
5105 -chardev socket,id=red1,host=3.3.3.3,port=9004
5106 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5107 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5113 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5114 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5115 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5116 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5117 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5118 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5119 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5120 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5121 -chardev socket,id=notify_way,host=3.3.3.3,port=9009,server=on,wait=off
5122 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5123 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5124 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5125 -object iothread,id=iothread1
5126 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,notify_dev=nofity_way,iothread=iothread1
5129 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5130 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5131 -chardev socket,id=red0,host=3.3.3.3,port=9003
5132 -chardev socket,id=red1,host=3.3.3.3,port=9004
5133 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5134 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5136 If you want to know the detail of above command line, you can
5137 read the colo-compare git log.
5139 ``-object cryptodev-backend-builtin,id=id[,queues=queues]``
5140 Creates a cryptodev backend which executes crypto opreation from
5141 the QEMU cipher APIS. The id parameter is a unique ID that will
5142 be used to reference this cryptodev backend from the
5143 ``virtio-crypto`` device. The queues parameter is optional,
5144 which specify the queue number of cryptodev backend, the default
5151 -object cryptodev-backend-builtin,id=cryptodev0 \\
5152 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5155 ``-object cryptodev-vhost-user,id=id,chardev=chardevid[,queues=queues]``
5156 Creates a vhost-user cryptodev backend, backed by a chardev
5157 chardevid. The id parameter is a unique ID that will be used to
5158 reference this cryptodev backend from the ``virtio-crypto``
5159 device. The chardev should be a unix domain socket backed one.
5160 The vhost-user uses a specifically defined protocol to pass
5161 vhost ioctl replacement messages to an application on the other
5162 end of the socket. The queues parameter is optional, which
5163 specify the queue number of cryptodev backend for multiqueue
5164 vhost-user, the default of queues is 1.
5170 -chardev socket,id=chardev0,path=/path/to/socket \\
5171 -object cryptodev-vhost-user,id=cryptodev0,chardev=chardev0 \\
5172 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5175 ``-object secret,id=id,data=string,format=raw|base64[,keyid=secretid,iv=string]``
5177 ``-object secret,id=id,file=filename,format=raw|base64[,keyid=secretid,iv=string]``
5178 Defines a secret to store a password, encryption key, or some
5179 other sensitive data. The sensitive data can either be passed
5180 directly via the data parameter, or indirectly via the file
5181 parameter. Using the data parameter is insecure unless the
5182 sensitive data is encrypted.
5184 The sensitive data can be provided in raw format (the default),
5185 or base64. When encoded as JSON, the raw format only supports
5186 valid UTF-8 characters, so base64 is recommended for sending
5187 binary data. QEMU will convert from which ever format is
5188 provided to the format it needs internally. eg, an RBD password
5189 can be provided in raw format, even though it will be base64
5190 encoded when passed onto the RBD sever.
5192 For added protection, it is possible to encrypt the data
5193 associated with a secret using the AES-256-CBC cipher. Use of
5194 encryption is indicated by providing the keyid and iv
5195 parameters. The keyid parameter provides the ID of a previously
5196 defined secret that contains the AES-256 decryption key. This
5197 key should be 32-bytes long and be base64 encoded. The iv
5198 parameter provides the random initialization vector used for
5199 encryption of this particular secret and should be a base64
5200 encrypted string of the 16-byte IV.
5202 The simplest (insecure) usage is to provide the secret inline
5206 # |qemu_system| -object secret,id=sec0,data=letmein,format=raw
5208 The simplest secure usage is to provide the secret via a file
5210 # printf "letmein
" > mypasswd.txt # QEMU\_SYSTEM\_MACRO -object
5211 secret,id=sec0,file=mypasswd.txt,format=raw
5213 For greater security, AES-256-CBC should be used. To illustrate
5214 usage, consider the openssl command line tool which can encrypt
5215 the data. Note that when encrypting, the plaintext must be
5216 padded to the cipher block size (32 bytes) using the standard
5217 PKCS#5/6 compatible padding algorithm.
5219 First a master key needs to be created in base64 encoding:
5223 # openssl rand -base64 32 > key.b64
5224 # KEY=$(base64 -d key.b64 | hexdump -v -e '/1 "%02X
"')
5226 Each secret to be encrypted needs to have a random
5227 initialization vector generated. These do not need to be kept
5232 # openssl rand -base64 16 > iv.b64
5233 # IV=$(base64 -d iv.b64 | hexdump -v -e '/1 "%02X
"')
5235 The secret to be defined can now be encrypted, in this case
5236 we're telling openssl to base64 encode the result, but it could
5237 be left as raw bytes if desired.
5241 # SECRET=$(printf "letmein
" |
5242 openssl enc -aes-256-cbc -a -K $KEY -iv $IV)
5244 When launching QEMU, create a master secret pointing to
5245 ``key.b64`` and specify that to be used to decrypt the user
5246 password. Pass the contents of ``iv.b64`` to the second secret
5251 -object secret,id=secmaster0,format=base64,file=key.b64 \\
5252 -object secret,id=sec0,keyid=secmaster0,format=base64,\\
5253 data=$SECRET,iv=$(<iv.b64)
5255 ``-object sev-guest,id=id,cbitpos=cbitpos,reduced-phys-bits=val,[sev-device=string,policy=policy,handle=handle,dh-cert-file=file,session-file=file,kernel-hashes=on|off]``
5256 Create a Secure Encrypted Virtualization (SEV) guest object,
5257 which can be used to provide the guest memory encryption support
5260 When memory encryption is enabled, one of the physical address
5261 bit (aka the C-bit) is utilized to mark if a memory page is
5262 protected. The ``cbitpos`` is used to provide the C-bit
5263 position. The C-bit position is Host family dependent hence user
5264 must provide this value. On EPYC, the value should be 47.
5266 When memory encryption is enabled, we loose certain bits in
5267 physical address space. The ``reduced-phys-bits`` is used to
5268 provide the number of bits we loose in physical address space.
5269 Similar to C-bit, the value is Host family dependent. On EPYC,
5270 the value should be 5.
5272 The ``sev-device`` provides the device file to use for
5273 communicating with the SEV firmware running inside AMD Secure
5274 Processor. The default device is '/dev/sev'. If hardware
5275 supports memory encryption then /dev/sev devices are created by
5278 The ``policy`` provides the guest policy to be enforced by the
5279 SEV firmware and restrict what configuration and operational
5280 commands can be performed on this guest by the hypervisor. The
5281 policy should be provided by the guest owner and is bound to the
5282 guest and cannot be changed throughout the lifetime of the
5283 guest. The default is 0.
5285 If guest ``policy`` allows sharing the key with another SEV
5286 guest then ``handle`` can be use to provide handle of the guest
5287 from which to share the key.
5289 The ``dh-cert-file`` and ``session-file`` provides the guest
5290 owner's Public Diffie-Hillman key defined in SEV spec. The PDH
5291 and session parameters are used for establishing a cryptographic
5292 session with the guest owner to negotiate keys used for
5293 attestation. The file must be encoded in base64.
5295 The ``kernel-hashes`` adds the hashes of given kernel/initrd/
5296 cmdline to a designated guest firmware page for measured Linux
5297 boot with -kernel. The default is off. (Since 6.2)
5299 e.g to launch a SEV guest
5303 # |qemu_system_x86| \\
5305 -object sev-guest,id=sev0,cbitpos=47,reduced-phys-bits=5 \\
5306 -machine ...,memory-encryption=sev0 \\
5309 ``-object authz-simple,id=id,identity=string``
5310 Create an authorization object that will control access to
5313 The ``identity`` parameter is identifies the user and its format
5314 depends on the network service that authorization object is
5315 associated with. For authorizing based on TLS x509 certificates,
5316 the identity must be the x509 distinguished name. Note that care
5317 must be taken to escape any commas in the distinguished name.
5319 An example authorization object to validate a x509 distinguished
5320 name would look like:
5326 -object 'authz-simple,id=auth0,identity=CN=laptop.example.com,,O=Example Org,,L=London,,ST=London,,C=GB' \\
5329 Note the use of quotes due to the x509 distinguished name
5330 containing whitespace, and escaping of ','.
5332 ``-object authz-listfile,id=id,filename=path,refresh=on|off``
5333 Create an authorization object that will control access to
5336 The ``filename`` parameter is the fully qualified path to a file
5337 containing the access control list rules in JSON format.
5339 An example set of rules that match against SASL usernames might
5346 { "match
": "fred
", "policy
": "allow
", "format
": "exact
" },
5347 { "match
": "bob
", "policy
": "allow
", "format
": "exact
" },
5348 { "match
": "danb
", "policy
": "deny
", "format
": "glob
" },
5349 { "match
": "dan
*", "policy
": "allow
", "format
": "exact
" },
5354 When checking access the object will iterate over all the rules
5355 and the first rule to match will have its ``policy`` value
5356 returned as the result. If no rules match, then the default
5357 ``policy`` value is returned.
5359 The rules can either be an exact string match, or they can use
5360 the simple UNIX glob pattern matching to allow wildcards to be
5363 If ``refresh`` is set to true the file will be monitored and
5364 automatically reloaded whenever its content changes.
5366 As with the ``authz-simple`` object, the format of the identity
5367 strings being matched depends on the network service, but is
5368 usually a TLS x509 distinguished name, or a SASL username.
5370 An example authorization object to validate a SASL username
5377 -object authz-simple,id=auth0,filename=/etc/qemu/vnc-sasl.acl,refresh=on \\
5380 ``-object authz-pam,id=id,service=string``
5381 Create an authorization object that will control access to
5384 The ``service`` parameter provides the name of a PAM service to
5385 use for authorization. It requires that a file
5386 ``/etc/pam.d/service`` exist to provide the configuration for
5387 the ``account`` subsystem.
5389 An example authorization object to validate a TLS x509
5390 distinguished name would look like:
5396 -object authz-pam,id=auth0,service=qemu-vnc \\
5399 There would then be a corresponding config file for PAM at
5400 ``/etc/pam.d/qemu-vnc`` that contains:
5404 account requisite pam_listfile.so item=user sense=allow \
5405 file=/etc/qemu/vnc.allow
5407 Finally the ``/etc/qemu/vnc.allow`` file would contain the list
5408 of x509 distingished names that are permitted access
5412 CN=laptop.example.com,O=Example Home,L=London,ST=London,C=GB
5414 ``-object iothread,id=id,poll-max-ns=poll-max-ns,poll-grow=poll-grow,poll-shrink=poll-shrink,aio-max-batch=aio-max-batch``
5415 Creates a dedicated event loop thread that devices can be
5416 assigned to. This is known as an IOThread. By default device
5417 emulation happens in vCPU threads or the main event loop thread.
5418 This can become a scalability bottleneck. IOThreads allow device
5419 emulation and I/O to run on other host CPUs.
5421 The ``id`` parameter is a unique ID that will be used to
5422 reference this IOThread from ``-device ...,iothread=id``.
5423 Multiple devices can be assigned to an IOThread. Note that not
5424 all devices support an ``iothread`` parameter.
5426 The ``query-iothreads`` QMP command lists IOThreads and reports
5427 their thread IDs so that the user can configure host CPU
5430 IOThreads use an adaptive polling algorithm to reduce event loop
5431 latency. Instead of entering a blocking system call to monitor
5432 file descriptors and then pay the cost of being woken up when an
5433 event occurs, the polling algorithm spins waiting for events for
5434 a short time. The algorithm's default parameters are suitable
5435 for many cases but can be adjusted based on knowledge of the
5436 workload and/or host device latency.
5438 The ``poll-max-ns`` parameter is the maximum number of
5439 nanoseconds to busy wait for events. Polling can be disabled by
5440 setting this value to 0.
5442 The ``poll-grow`` parameter is the multiplier used to increase
5443 the polling time when the algorithm detects it is missing events
5444 due to not polling long enough.
5446 The ``poll-shrink`` parameter is the divisor used to decrease
5447 the polling time when the algorithm detects it is spending too
5448 long polling without encountering events.
5450 The ``aio-max-batch`` parameter is the maximum number of requests
5451 in a batch for the AIO engine, 0 means that the engine will use
5454 The IOThread parameters can be modified at run-time using the
5455 ``qom-set`` command (where ``iothread1`` is the IOThread's
5460 (qemu) qom-set /objects/iothread1 poll-max-ns 100000
5464 HXCOMM This is the last statement. Insert new options before this line!