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"
40 " cxl-fmw.0.targets.0=firsttarget,cxl-fmw.0.targets.1=secondtarget,cxl-fmw.0.size=size[,cxl-fmw.0.interleave-granularity=granularity]\n"
41 " zpcii-disable=on|off disables zPCI interpretation facilities (default=off)\n",
44 ``-machine [type=]name[,prop=value[,...]]``
45 Select the emulated machine by name. Use ``-machine help`` to list
48 For architectures which aim to support live migration compatibility
49 across releases, each release will introduce a new versioned machine
50 type. For example, the 2.8.0 release introduced machine types
51 "pc-i440fx-2.8" and "pc-q35-2.8" for the x86\_64/i686 architectures.
53 To allow live migration of guests from QEMU version 2.8.0, to QEMU
54 version 2.9.0, the 2.9.0 version must support the "pc-i440fx-2.8"
55 and "pc-q35-2.8" machines too. To allow users live migrating VMs to
56 skip multiple intermediate releases when upgrading, new releases of
57 QEMU will support machine types from many previous versions.
59 Supported machine properties are:
61 ``accel=accels1[:accels2[:...]]``
62 This is used to enable an accelerator. Depending on the target
63 architecture, kvm, xen, hax, hvf, nvmm, whpx or tcg can be available.
64 By default, tcg is used. If there is more than one accelerator
65 specified, the next one is used if the previous one fails to
68 ``vmport=on|off|auto``
69 Enables emulation of VMWare IO port, for vmmouse etc. auto says
70 to select the value based on accel. For accel=xen the default is
71 off otherwise the default is on.
73 ``dump-guest-core=on|off``
74 Include guest memory in a core dump. The default is on.
77 Enables or disables memory merge support. This feature, when
78 supported by the host, de-duplicates identical memory pages
79 among VMs instances (enabled by default).
81 ``aes-key-wrap=on|off``
82 Enables or disables AES key wrapping support on s390-ccw hosts.
83 This feature controls whether AES wrapping keys will be created
84 to allow execution of AES cryptographic functions. The default
87 ``dea-key-wrap=on|off``
88 Enables or disables DEA key wrapping support on s390-ccw hosts.
89 This feature controls whether DEA wrapping keys will be created
90 to allow execution of DEA cryptographic functions. The default
94 Enables or disables NVDIMM support. The default is off.
96 ``memory-encryption=``
97 Memory encryption object to use. The default is none.
100 Enables or disables ACPI Heterogeneous Memory Attribute Table
101 (HMAT) support. The default is off.
103 ``memory-backend='id'``
104 An alternative to legacy ``-mem-path`` and ``mem-prealloc`` options.
105 Allows to use a memory backend as main RAM.
110 -object memory-backend-file,id=pc.ram,size=512M,mem-path=/hugetlbfs,prealloc=on,share=on
111 -machine memory-backend=pc.ram
114 Migration compatibility note:
116 * as backend id one shall use value of 'default-ram-id', advertised by
117 machine type (available via ``query-machines`` QMP command), if migration
118 to/from old QEMU (<5.0) is expected.
119 * for machine types 4.0 and older, user shall
120 use ``x-use-canonical-path-for-ramblock-id=off`` backend option
121 if migration to/from old QEMU (<5.0) is expected.
126 -object memory-backend-ram,id=pc.ram,size=512M,x-use-canonical-path-for-ramblock-id=off
127 -machine memory-backend=pc.ram
130 ``cxl-fmw.0.targets.0=firsttarget,cxl-fmw.0.targets.1=secondtarget,cxl-fmw.0.size=size[,cxl-fmw.0.interleave-granularity=granularity]``
131 Define a CXL Fixed Memory Window (CFMW).
133 Described in the CXL 2.0 ECN: CEDT CFMWS & QTG _DSM.
135 They are regions of Host Physical Addresses (HPA) on a system which
136 may be interleaved across one or more CXL host bridges. The system
137 software will assign particular devices into these windows and
138 configure the downstream Host-managed Device Memory (HDM) decoders
139 in root ports, switch ports and devices appropriately to meet the
140 interleave requirements before enabling the memory devices.
142 ``targets.X=target`` provides the mapping to CXL host bridges
143 which may be identified by the id provied in the -device entry.
144 Multiple entries are needed to specify all the targets when
145 the fixed memory window represents interleaved memory. X is the
148 ``size=size`` sets the size of the CFMW. This must be a multiple of
149 256MiB. The region will be aligned to 256MiB but the location is
150 platform and configuration dependent.
152 ``interleave-granularity=granularity`` sets the granularity of
153 interleave. Default 256KiB. Only 256KiB, 512KiB, 1024KiB, 2048KiB
154 4096KiB, 8192KiB and 16384KiB granularities supported.
160 -machine cxl-fmw.0.targets.0=cxl.0,cxl-fmw.0.targets.1=cxl.1,cxl-fmw.0.size=128G,cxl-fmw.0.interleave-granularity=512k
162 ``zpcii-disable=on|off``
163 Disables zPCI interpretation facilties on s390-ccw hosts.
164 This feature can be used to disable hardware virtual assists
165 related to zPCI devices. The default is off.
168 DEF("M", HAS_ARG, QEMU_OPTION_M,
169 " sgx-epc.0.memdev=memid,sgx-epc.0.node=numaid\n",
173 ``sgx-epc.0.memdev=@var{memid},sgx-epc.0.node=@var{numaid}``
174 Define an SGX EPC section.
177 DEF("cpu", HAS_ARG, QEMU_OPTION_cpu,
178 "-cpu cpu select CPU ('-cpu help' for list)\n", QEMU_ARCH_ALL)
181 Select CPU model (``-cpu help`` for list and additional feature
185 DEF("accel", HAS_ARG, QEMU_OPTION_accel,
186 "-accel [accel=]accelerator[,prop[=value][,...]]\n"
187 " select accelerator (kvm, xen, hax, hvf, nvmm, whpx or tcg; use 'help' for a list)\n"
188 " igd-passthru=on|off (enable Xen integrated Intel graphics passthrough, default=off)\n"
189 " kernel-irqchip=on|off|split controls accelerated irqchip support (default=on)\n"
190 " kvm-shadow-mem=size of KVM shadow MMU in bytes\n"
191 " split-wx=on|off (enable TCG split w^x mapping)\n"
192 " tb-size=n (TCG translation block cache size)\n"
193 " dirty-ring-size=n (KVM dirty ring GFN count, default 0)\n"
194 " thread=single|multi (enable multi-threaded TCG)\n", QEMU_ARCH_ALL)
196 ``-accel name[,prop=value[,...]]``
197 This is used to enable an accelerator. Depending on the target
198 architecture, kvm, xen, hax, hvf, nvmm, whpx or tcg can be available. By
199 default, tcg is used. If there is more than one accelerator
200 specified, the next one is used if the previous one fails to
203 ``igd-passthru=on|off``
204 When Xen is in use, this option controls whether Intel
205 integrated graphics devices can be passed through to the guest
208 ``kernel-irqchip=on|off|split``
209 Controls KVM in-kernel irqchip support. The default is full
210 acceleration of the interrupt controllers. On x86, split irqchip
211 reduces the kernel attack surface, at a performance cost for
212 non-MSI interrupts. Disabling the in-kernel irqchip completely
213 is not recommended except for debugging purposes.
215 ``kvm-shadow-mem=size``
216 Defines the size of the KVM shadow MMU.
219 Controls the use of split w^x mapping for the TCG code generation
220 buffer. Some operating systems require this to be enabled, and in
221 such a case this will default on. On other operating systems, this
222 will default off, but one may enable this for testing or debugging.
225 Controls the size (in MiB) of the TCG translation block cache.
227 ``thread=single|multi``
228 Controls number of TCG threads. When the TCG is multi-threaded
229 there will be one thread per vCPU therefore taking advantage of
230 additional host cores. The default is to enable multi-threading
231 where both the back-end and front-ends support it and no
232 incompatible TCG features have been enabled (e.g.
235 ``dirty-ring-size=n``
236 When the KVM accelerator is used, it controls the size of the per-vCPU
237 dirty page ring buffer (number of entries for each vCPU). It should
238 be a value that is power of two, and it should be 1024 or bigger (but
239 still less than the maximum value that the kernel supports). 4096
240 could be a good initial value if you have no idea which is the best.
241 Set this value to 0 to disable the feature. By default, this feature
242 is disabled (dirty-ring-size=0). When enabled, KVM will instead
243 record dirty pages in a bitmap.
247 DEF("smp", HAS_ARG, QEMU_OPTION_smp,
248 "-smp [[cpus=]n][,maxcpus=maxcpus][,sockets=sockets][,dies=dies][,clusters=clusters][,cores=cores][,threads=threads]\n"
249 " set the number of initial CPUs to 'n' [default=1]\n"
250 " maxcpus= maximum number of total CPUs, including\n"
251 " offline CPUs for hotplug, etc\n"
252 " sockets= number of sockets on the machine board\n"
253 " dies= number of dies in one socket\n"
254 " clusters= number of clusters in one die\n"
255 " cores= number of cores in one cluster\n"
256 " threads= number of threads in one core\n"
257 "Note: Different machines may have different subsets of the CPU topology\n"
258 " parameters supported, so the actual meaning of the supported parameters\n"
259 " will vary accordingly. For example, for a machine type that supports a\n"
260 " three-level CPU hierarchy of sockets/cores/threads, the parameters will\n"
261 " sequentially mean as below:\n"
262 " sockets means the number of sockets on the machine board\n"
263 " cores means the number of cores in one socket\n"
264 " threads means the number of threads in one core\n"
265 " For a particular machine type board, an expected CPU topology hierarchy\n"
266 " can be defined through the supported sub-option. Unsupported parameters\n"
267 " can also be provided in addition to the sub-option, but their values\n"
268 " must be set as 1 in the purpose of correct parsing.\n",
271 ``-smp [[cpus=]n][,maxcpus=maxcpus][,sockets=sockets][,dies=dies][,clusters=clusters][,cores=cores][,threads=threads]``
272 Simulate a SMP system with '\ ``n``\ ' CPUs initially present on
273 the machine type board. On boards supporting CPU hotplug, the optional
274 '\ ``maxcpus``\ ' parameter can be set to enable further CPUs to be
275 added at runtime. When both parameters are omitted, the maximum number
276 of CPUs will be calculated from the provided topology members and the
277 initial CPU count will match the maximum number. When only one of them
278 is given then the omitted one will be set to its counterpart's value.
279 Both parameters may be specified, but the maximum number of CPUs must
280 be equal to or greater than the initial CPU count. Product of the
281 CPU topology hierarchy must be equal to the maximum number of CPUs.
282 Both parameters are subject to an upper limit that is determined by
283 the specific machine type chosen.
285 To control reporting of CPU topology information, values of the topology
286 parameters can be specified. Machines may only support a subset of the
287 parameters and different machines may have different subsets supported
288 which vary depending on capacity of the corresponding CPU targets. So
289 for a particular machine type board, an expected topology hierarchy can
290 be defined through the supported sub-option. Unsupported parameters can
291 also be provided in addition to the sub-option, but their values must be
292 set as 1 in the purpose of correct parsing.
294 Either the initial CPU count, or at least one of the topology parameters
295 must be specified. The specified parameters must be greater than zero,
296 explicit configuration like "cpus=0" is not allowed. Values for any
297 omitted parameters will be computed from those which are given.
299 For example, the following sub-option defines a CPU topology hierarchy
300 (2 sockets totally on the machine, 2 cores per socket, 2 threads per
301 core) for a machine that only supports sockets/cores/threads.
302 Some members of the option can be omitted but their values will be
303 automatically computed:
307 -smp 8,sockets=2,cores=2,threads=2,maxcpus=8
309 The following sub-option defines a CPU topology hierarchy (2 sockets
310 totally on the machine, 2 dies per socket, 2 cores per die, 2 threads
311 per core) for PC machines which support sockets/dies/cores/threads.
312 Some members of the option can be omitted but their values will be
313 automatically computed:
317 -smp 16,sockets=2,dies=2,cores=2,threads=2,maxcpus=16
319 The following sub-option defines a CPU topology hierarchy (2 sockets
320 totally on the machine, 2 clusters per socket, 2 cores per cluster,
321 2 threads per core) for ARM virt machines which support sockets/clusters
322 /cores/threads. Some members of the option can be omitted but their values
323 will be automatically computed:
327 -smp 16,sockets=2,clusters=2,cores=2,threads=2,maxcpus=16
329 Historically preference was given to the coarsest topology parameters
330 when computing missing values (ie sockets preferred over cores, which
331 were preferred over threads), however, this behaviour is considered
332 liable to change. Prior to 6.2 the preference was sockets over cores
333 over threads. Since 6.2 the preference is cores over sockets over threads.
335 For example, the following option defines a machine board with 2 sockets
336 of 1 core before 6.2 and 1 socket of 2 cores after 6.2:
343 DEF("numa", HAS_ARG, QEMU_OPTION_numa,
344 "-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
345 "-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
346 "-numa dist,src=source,dst=destination,val=distance\n"
347 "-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]\n"
348 "-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"
349 "-numa hmat-cache,node-id=node,size=size,level=level[,associativity=none|direct|complex][,policy=none|write-back|write-through][,line=size]\n",
352 ``-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
354 ``-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
356 ``-numa dist,src=source,dst=destination,val=distance``
358 ``-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]``
360 ``-numa hmat-lb,initiator=node,target=node,hierarchy=hierarchy,data-type=tpye[,latency=lat][,bandwidth=bw]``
362 ``-numa hmat-cache,node-id=node,size=size,level=level[,associativity=str][,policy=str][,line=size]``
363 Define a NUMA node and assign RAM and VCPUs to it. Set the NUMA
364 distance from a source node to a destination node. Set the ACPI
365 Heterogeneous Memory Attributes for the given nodes.
367 Legacy VCPU assignment uses '\ ``cpus``\ ' option where firstcpu and
368 lastcpu are CPU indexes. Each '\ ``cpus``\ ' option represent a
369 contiguous range of CPU indexes (or a single VCPU if lastcpu is
370 omitted). A non-contiguous set of VCPUs can be represented by
371 providing multiple '\ ``cpus``\ ' options. If '\ ``cpus``\ ' is
372 omitted on all nodes, VCPUs are automatically split between them.
374 For example, the following option assigns VCPUs 0, 1, 2 and 5 to a
379 -numa node,cpus=0-2,cpus=5
381 '\ ``cpu``\ ' option is a new alternative to '\ ``cpus``\ ' option
382 which uses '\ ``socket-id|core-id|thread-id``\ ' properties to
383 assign CPU objects to a node using topology layout properties of
384 CPU. The set of properties is machine specific, and depends on used
385 machine type/'\ ``smp``\ ' options. It could be queried with
386 '\ ``hotpluggable-cpus``\ ' monitor command. '\ ``node-id``\ '
387 property specifies node to which CPU object will be assigned, it's
388 required for node to be declared with '\ ``node``\ ' option before
389 it's used with '\ ``cpu``\ ' option.
396 -smp 1,sockets=2,maxcpus=2 \
397 -numa node,nodeid=0 -numa node,nodeid=1 \
398 -numa cpu,node-id=0,socket-id=0 -numa cpu,node-id=1,socket-id=1
400 Legacy '\ ``mem``\ ' assigns a given RAM amount to a node (not supported
401 for 5.1 and newer machine types). '\ ``memdev``\ ' assigns RAM from
402 a given memory backend device to a node. If '\ ``mem``\ ' and
403 '\ ``memdev``\ ' are omitted in all nodes, RAM is split equally between them.
406 '\ ``mem``\ ' and '\ ``memdev``\ ' are mutually exclusive.
407 Furthermore, if one node uses '\ ``memdev``\ ', all of them have to
410 '\ ``initiator``\ ' is an additional option that points to an
411 initiator NUMA node that has best performance (the lowest latency or
412 largest bandwidth) to this NUMA node. Note that this option can be
413 set only when the machine property 'hmat' is set to 'on'.
415 Following example creates a machine with 2 NUMA nodes, node 0 has
416 CPU. node 1 has only memory, and its initiator is node 0. Note that
417 because node 0 has CPU, by default the initiator of node 0 is itself
423 -m 2G,slots=2,maxmem=4G \
424 -object memory-backend-ram,size=1G,id=m0 \
425 -object memory-backend-ram,size=1G,id=m1 \
426 -numa node,nodeid=0,memdev=m0 \
427 -numa node,nodeid=1,memdev=m1,initiator=0 \
428 -smp 2,sockets=2,maxcpus=2 \
429 -numa cpu,node-id=0,socket-id=0 \
430 -numa cpu,node-id=0,socket-id=1
432 source and destination are NUMA node IDs. distance is the NUMA
433 distance from source to destination. The distance from a node to
434 itself is always 10. If any pair of nodes is given a distance, then
435 all pairs must be given distances. Although, when distances are only
436 given in one direction for each pair of nodes, then the distances in
437 the opposite directions are assumed to be the same. If, however, an
438 asymmetrical pair of distances is given for even one node pair, then
439 all node pairs must be provided distance values for both directions,
440 even when they are symmetrical. When a node is unreachable from
441 another node, set the pair's distance to 255.
443 Note that the -``numa`` option doesn't allocate any of the specified
444 resources, it just assigns existing resources to NUMA nodes. This
445 means that one still has to use the ``-m``, ``-smp`` options to
446 allocate RAM and VCPUs respectively.
448 Use '\ ``hmat-lb``\ ' to set System Locality Latency and Bandwidth
449 Information between initiator and target NUMA nodes in ACPI
450 Heterogeneous Attribute Memory Table (HMAT). Initiator NUMA node can
451 create memory requests, usually it has one or more processors.
452 Target NUMA node contains addressable memory.
454 In '\ ``hmat-lb``\ ' option, node are NUMA node IDs. hierarchy is
455 the memory hierarchy of the target NUMA node: if hierarchy is
456 'memory', the structure represents the memory performance; if
457 hierarchy is 'first-level\|second-level\|third-level', this
458 structure represents aggregated performance of memory side caches
459 for each domain. type of 'data-type' is type of data represented by
460 this structure instance: if 'hierarchy' is 'memory', 'data-type' is
461 'access\|read\|write' latency or 'access\|read\|write' bandwidth of
462 the target memory; if 'hierarchy' is
463 'first-level\|second-level\|third-level', 'data-type' is
464 'access\|read\|write' hit latency or 'access\|read\|write' hit
465 bandwidth of the target memory side cache.
467 lat is latency value in nanoseconds. bw is bandwidth value, the
468 possible value and units are NUM[M\|G\|T], mean that the bandwidth
469 value are NUM byte per second (or MB/s, GB/s or TB/s depending on
470 used suffix). Note that if latency or bandwidth value is 0, means
471 the corresponding latency or bandwidth information is not provided.
473 In '\ ``hmat-cache``\ ' option, node-id is the NUMA-id of the memory
474 belongs. size is the size of memory side cache in bytes. level is
475 the cache level described in this structure, note that the cache
476 level 0 should not be used with '\ ``hmat-cache``\ ' option.
477 associativity is the cache associativity, the possible value is
478 'none/direct(direct-mapped)/complex(complex cache indexing)'. policy
479 is the write policy. line is the cache Line size in bytes.
481 For example, the following options describe 2 NUMA nodes. Node 0 has
482 2 cpus and a ram, node 1 has only a ram. The processors in node 0
483 access memory in node 0 with access-latency 5 nanoseconds,
484 access-bandwidth is 200 MB/s; The processors in NUMA node 0 access
485 memory in NUMA node 1 with access-latency 10 nanoseconds,
486 access-bandwidth is 100 MB/s. And for memory side cache information,
487 NUMA node 0 and 1 both have 1 level memory cache, size is 10KB,
488 policy is write-back, the cache Line size is 8 bytes:
494 -object memory-backend-ram,size=1G,id=m0 \
495 -object memory-backend-ram,size=1G,id=m1 \
496 -smp 2,sockets=2,maxcpus=2 \
497 -numa node,nodeid=0,memdev=m0 \
498 -numa node,nodeid=1,memdev=m1,initiator=0 \
499 -numa cpu,node-id=0,socket-id=0 \
500 -numa cpu,node-id=0,socket-id=1 \
501 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-latency,latency=5 \
502 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-bandwidth,bandwidth=200M \
503 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-latency,latency=10 \
504 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-bandwidth,bandwidth=100M \
505 -numa hmat-cache,node-id=0,size=10K,level=1,associativity=direct,policy=write-back,line=8 \
506 -numa hmat-cache,node-id=1,size=10K,level=1,associativity=direct,policy=write-back,line=8
509 DEF("add-fd", HAS_ARG, QEMU_OPTION_add_fd,
510 "-add-fd fd=fd,set=set[,opaque=opaque]\n"
511 " Add 'fd' to fd 'set'\n", QEMU_ARCH_ALL)
513 ``-add-fd fd=fd,set=set[,opaque=opaque]``
514 Add a file descriptor to an fd set. Valid options are:
517 This option defines the file descriptor of which a duplicate is
518 added to fd set. The file descriptor cannot be stdin, stdout, or
522 This option defines the ID of the fd set to add the file
526 This option defines a free-form string that can be used to
529 You can open an image using pre-opened file descriptors from an fd
535 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
536 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
537 -drive file=/dev/fdset/2,index=0,media=disk
540 DEF("set", HAS_ARG, QEMU_OPTION_set,
541 "-set group.id.arg=value\n"
542 " set <arg> parameter for item <id> of type <group>\n"
543 " i.e. -set drive.$id.file=/path/to/image\n", QEMU_ARCH_ALL)
545 ``-set group.id.arg=value``
546 Set parameter arg for item id of type group
549 DEF("global", HAS_ARG, QEMU_OPTION_global,
550 "-global driver.property=value\n"
551 "-global driver=driver,property=property,value=value\n"
552 " set a global default for a driver property\n",
555 ``-global driver.prop=value``
557 ``-global driver=driver,property=property,value=value``
558 Set default value of driver's property prop to value, e.g.:
562 |qemu_system_x86| -global ide-hd.physical_block_size=4096 disk-image.img
564 In particular, you can use this to set driver properties for devices
565 which are created automatically by the machine model. To create a
566 device which is not created automatically and set properties on it,
569 -global driver.prop=value is shorthand for -global
570 driver=driver,property=prop,value=value. The longhand syntax works
571 even when driver contains a dot.
574 DEF("boot", HAS_ARG, QEMU_OPTION_boot,
575 "-boot [order=drives][,once=drives][,menu=on|off]\n"
576 " [,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_time][,strict=on|off]\n"
577 " 'drives': floppy (a), hard disk (c), CD-ROM (d), network (n)\n"
578 " 'sp_name': the file's name that would be passed to bios as logo picture, if menu=on\n"
579 " 'sp_time': the period that splash picture last if menu=on, unit is ms\n"
580 " 'rb_timeout': the timeout before guest reboot when boot failed, unit is ms\n",
583 ``-boot [order=drives][,once=drives][,menu=on|off][,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_timeout][,strict=on|off]``
584 Specify boot order drives as a string of drive letters. Valid drive
585 letters depend on the target architecture. The x86 PC uses: a, b
586 (floppy 1 and 2), c (first hard disk), d (first CD-ROM), n-p
587 (Etherboot from network adapter 1-4), hard disk boot is the default.
588 To apply a particular boot order only on the first startup, specify
589 it via ``once``. Note that the ``order`` or ``once`` parameter
590 should not be used together with the ``bootindex`` property of
591 devices, since the firmware implementations normally do not support
592 both at the same time.
594 Interactive boot menus/prompts can be enabled via ``menu=on`` as far
595 as firmware/BIOS supports them. The default is non-interactive boot.
597 A splash picture could be passed to bios, enabling user to show it
598 as logo, when option splash=sp\_name is given and menu=on, If
599 firmware/BIOS supports them. Currently Seabios for X86 system
600 support it. limitation: The splash file could be a jpeg file or a
601 BMP file in 24 BPP format(true color). The resolution should be
602 supported by the SVGA mode, so the recommended is 320x240, 640x480,
605 A timeout could be passed to bios, guest will pause for rb\_timeout
606 ms when boot failed, then reboot. If rb\_timeout is '-1', guest will
607 not reboot, qemu passes '-1' to bios by default. Currently Seabios
608 for X86 system support it.
610 Do strict boot via ``strict=on`` as far as firmware/BIOS supports
611 it. This only effects when boot priority is changed by bootindex
612 options. The default is non-strict boot.
616 # try to boot from network first, then from hard disk
617 |qemu_system_x86| -boot order=nc
618 # boot from CD-ROM first, switch back to default order after reboot
619 |qemu_system_x86| -boot once=d
620 # boot with a splash picture for 5 seconds.
621 |qemu_system_x86| -boot menu=on,splash=/root/boot.bmp,splash-time=5000
623 Note: The legacy format '-boot drives' is still supported but its
624 use is discouraged as it may be removed from future versions.
627 DEF("m", HAS_ARG, QEMU_OPTION_m,
628 "-m [size=]megs[,slots=n,maxmem=size]\n"
629 " configure guest RAM\n"
630 " size: initial amount of guest memory\n"
631 " slots: number of hotplug slots (default: none)\n"
632 " maxmem: maximum amount of guest memory (default: none)\n"
633 "NOTE: Some architectures might enforce a specific granularity\n",
636 ``-m [size=]megs[,slots=n,maxmem=size]``
637 Sets guest startup RAM size to megs megabytes. Default is 128 MiB.
638 Optionally, a suffix of "M" or "G" can be used to signify a value in
639 megabytes or gigabytes respectively. Optional pair slots, maxmem
640 could be used to set amount of hotpluggable memory slots and maximum
641 amount of memory. Note that maxmem must be aligned to the page size.
643 For example, the following command-line sets the guest startup RAM
644 size to 1GB, creates 3 slots to hotplug additional memory and sets
645 the maximum memory the guest can reach to 4GB:
649 |qemu_system| -m 1G,slots=3,maxmem=4G
651 If slots and maxmem are not specified, memory hotplug won't be
652 enabled and the guest startup RAM will never increase.
655 DEF("mem-path", HAS_ARG, QEMU_OPTION_mempath,
656 "-mem-path FILE provide backing storage for guest RAM\n", QEMU_ARCH_ALL)
659 Allocate guest RAM from a temporarily created file in path.
662 DEF("mem-prealloc", 0, QEMU_OPTION_mem_prealloc,
663 "-mem-prealloc preallocate guest memory (use with -mem-path)\n",
667 Preallocate memory when using -mem-path.
670 DEF("k", HAS_ARG, QEMU_OPTION_k,
671 "-k language use keyboard layout (for example 'fr' for French)\n",
675 Use keyboard layout language (for example ``fr`` for French). This
676 option is only needed where it is not easy to get raw PC keycodes
677 (e.g. on Macs, with some X11 servers or with a VNC or curses
678 display). You don't normally need to use it on PC/Linux or
681 The available layouts are:
685 ar de-ch es fo fr-ca hu ja mk no pt-br sv
686 da en-gb et fr fr-ch is lt nl pl ru th
687 de en-us fi fr-be hr it lv nl-be pt sl tr
689 The default is ``en-us``.
693 HXCOMM Deprecated by -audiodev
694 DEF("audio-help", 0, QEMU_OPTION_audio_help,
695 "-audio-help show -audiodev equivalent of the currently specified audio settings\n",
699 Will show the -audiodev equivalent of the currently specified
700 (deprecated) environment variables.
703 DEF("audio", HAS_ARG, QEMU_OPTION_audio,
704 "-audio [driver=]driver,model=value[,prop[=value][,...]]\n"
705 " specifies the audio backend and device to use;\n"
706 " apart from 'model', options are the same as for -audiodev.\n"
707 " use '-audio model=help' to show possible devices.\n",
710 ``-audio [driver=]driver,model=value[,prop[=value][,...]]``
711 This option is a shortcut for configuring both the guest audio
712 hardware and the host audio backend in one go.
713 The driver option is the same as with the corresponding ``-audiodev`` option below.
714 The guest hardware model can be set with ``model=modelname``.
716 Use ``driver=help`` to list the available drivers,
717 and ``model=help`` to list the available device types.
719 The following two example do exactly the same, to show how ``-audio``
720 can be used to shorten the command line length:
724 |qemu_system| -audiodev pa,id=pa -device sb16,audiodev=pa
725 |qemu_system| -audio pa,model=sb16
728 DEF("audiodev", HAS_ARG, QEMU_OPTION_audiodev,
729 "-audiodev [driver=]driver,id=id[,prop[=value][,...]]\n"
730 " specifies the audio backend to use\n"
731 " Use ``-audiodev help`` to list the available drivers\n"
732 " id= identifier of the backend\n"
733 " timer-period= timer period in microseconds\n"
734 " in|out.mixing-engine= use mixing engine to mix streams inside QEMU\n"
735 " in|out.fixed-settings= use fixed settings for host audio\n"
736 " in|out.frequency= frequency to use with fixed settings\n"
737 " in|out.channels= number of channels to use with fixed settings\n"
738 " in|out.format= sample format to use with fixed settings\n"
739 " valid values: s8, s16, s32, u8, u16, u32, f32\n"
740 " in|out.voices= number of voices to use\n"
741 " in|out.buffer-length= length of buffer in microseconds\n"
742 "-audiodev none,id=id,[,prop[=value][,...]]\n"
743 " dummy driver that discards all output\n"
744 #ifdef CONFIG_AUDIO_ALSA
745 "-audiodev alsa,id=id[,prop[=value][,...]]\n"
746 " in|out.dev= name of the audio device to use\n"
747 " in|out.period-length= length of period in microseconds\n"
748 " in|out.try-poll= attempt to use poll mode\n"
749 " threshold= threshold (in microseconds) when playback starts\n"
751 #ifdef CONFIG_AUDIO_COREAUDIO
752 "-audiodev coreaudio,id=id[,prop[=value][,...]]\n"
753 " in|out.buffer-count= number of buffers\n"
755 #ifdef CONFIG_AUDIO_DSOUND
756 "-audiodev dsound,id=id[,prop[=value][,...]]\n"
757 " latency= add extra latency to playback in microseconds\n"
759 #ifdef CONFIG_AUDIO_OSS
760 "-audiodev oss,id=id[,prop[=value][,...]]\n"
761 " in|out.dev= path of the audio device to use\n"
762 " in|out.buffer-count= number of buffers\n"
763 " in|out.try-poll= attempt to use poll mode\n"
764 " try-mmap= try using memory mapped access\n"
765 " exclusive= open device in exclusive mode\n"
766 " dsp-policy= set timing policy (0..10), -1 to use fragment mode\n"
768 #ifdef CONFIG_AUDIO_PA
769 "-audiodev pa,id=id[,prop[=value][,...]]\n"
770 " server= PulseAudio server address\n"
771 " in|out.name= source/sink device name\n"
772 " in|out.latency= desired latency in microseconds\n"
774 #ifdef CONFIG_AUDIO_SDL
775 "-audiodev sdl,id=id[,prop[=value][,...]]\n"
776 " in|out.buffer-count= number of buffers\n"
778 #ifdef CONFIG_AUDIO_SNDIO
779 "-audiodev sndio,id=id[,prop[=value][,...]]\n"
782 "-audiodev spice,id=id[,prop[=value][,...]]\n"
784 #ifdef CONFIG_DBUS_DISPLAY
785 "-audiodev dbus,id=id[,prop[=value][,...]]\n"
787 "-audiodev wav,id=id[,prop[=value][,...]]\n"
788 " path= path of wav file to record\n",
791 ``-audiodev [driver=]driver,id=id[,prop[=value][,...]]``
792 Adds a new audio backend driver identified by id. There are global
793 and driver specific properties. Some values can be set differently
794 for input and output, they're marked with ``in|out.``. You can set
795 the input's property with ``in.prop`` and the output's property with
796 ``out.prop``. For example:
800 -audiodev alsa,id=example,in.frequency=44110,out.frequency=8000
801 -audiodev alsa,id=example,out.channels=1 # leaves in.channels unspecified
803 NOTE: parameter validation is known to be incomplete, in many cases
804 specifying an invalid option causes QEMU to print an error message
805 and continue emulation without sound.
807 Valid global options are:
810 Identifies the audio backend.
812 ``timer-period=period``
813 Sets the timer period used by the audio subsystem in
814 microseconds. Default is 10000 (10 ms).
816 ``in|out.mixing-engine=on|off``
817 Use QEMU's mixing engine to mix all streams inside QEMU and
818 convert audio formats when not supported by the backend. When
819 off, fixed-settings must be off too. Note that disabling this
820 option means that the selected backend must support multiple
821 streams and the audio formats used by the virtual cards,
822 otherwise you'll get no sound. It's not recommended to disable
823 this option unless you want to use 5.1 or 7.1 audio, as mixing
824 engine only supports mono and stereo audio. Default is on.
826 ``in|out.fixed-settings=on|off``
827 Use fixed settings for host audio. When off, it will change
828 based on how the guest opens the sound card. In this case you
829 must not specify frequency, channels or format. Default is on.
831 ``in|out.frequency=frequency``
832 Specify the frequency to use when using fixed-settings. Default
835 ``in|out.channels=channels``
836 Specify the number of channels to use when using fixed-settings.
837 Default is 2 (stereo).
839 ``in|out.format=format``
840 Specify the sample format to use when using fixed-settings.
841 Valid values are: ``s8``, ``s16``, ``s32``, ``u8``, ``u16``,
842 ``u32``, ``f32``. Default is ``s16``.
844 ``in|out.voices=voices``
845 Specify the number of voices to use. Default is 1.
847 ``in|out.buffer-length=usecs``
848 Sets the size of the buffer in microseconds.
850 ``-audiodev none,id=id[,prop[=value][,...]]``
851 Creates a dummy backend that discards all outputs. This backend has
852 no backend specific properties.
854 ``-audiodev alsa,id=id[,prop[=value][,...]]``
855 Creates backend using the ALSA. This backend is only available on
858 ALSA specific options are:
860 ``in|out.dev=device``
861 Specify the ALSA device to use for input and/or output. Default
864 ``in|out.period-length=usecs``
865 Sets the period length in microseconds.
867 ``in|out.try-poll=on|off``
868 Attempt to use poll mode with the device. Default is on.
870 ``threshold=threshold``
871 Threshold (in microseconds) when playback starts. Default is 0.
873 ``-audiodev coreaudio,id=id[,prop[=value][,...]]``
874 Creates a backend using Apple's Core Audio. This backend is only
875 available on Mac OS and only supports playback.
877 Core Audio specific options are:
879 ``in|out.buffer-count=count``
880 Sets the count of the buffers.
882 ``-audiodev dsound,id=id[,prop[=value][,...]]``
883 Creates a backend using Microsoft's DirectSound. This backend is
884 only available on Windows and only supports playback.
886 DirectSound specific options are:
889 Add extra usecs microseconds latency to playback. Default is
892 ``-audiodev oss,id=id[,prop[=value][,...]]``
893 Creates a backend using OSS. This backend is available on most
896 OSS specific options are:
898 ``in|out.dev=device``
899 Specify the file name of the OSS device to use. Default is
902 ``in|out.buffer-count=count``
903 Sets the count of the buffers.
905 ``in|out.try-poll=on|of``
906 Attempt to use poll mode with the device. Default is on.
909 Try using memory mapped device access. Default is off.
912 Open the device in exclusive mode (vmix won't work in this
913 case). Default is off.
915 ``dsp-policy=policy``
916 Sets the timing policy (between 0 and 10, where smaller number
917 means smaller latency but higher CPU usage). Use -1 to use
918 buffer sizes specified by ``buffer`` and ``buffer-count``. This
919 option is ignored if you do not have OSS 4. Default is 5.
921 ``-audiodev pa,id=id[,prop[=value][,...]]``
922 Creates a backend using PulseAudio. This backend is available on
925 PulseAudio specific options are:
928 Sets the PulseAudio server to connect to.
931 Use the specified source/sink for recording/playback.
933 ``in|out.latency=usecs``
934 Desired latency in microseconds. The PulseAudio server will try
935 to honor this value but actual latencies may be lower or higher.
937 ``-audiodev sdl,id=id[,prop[=value][,...]]``
938 Creates a backend using SDL. This backend is available on most
939 systems, but you should use your platform's native backend if
942 SDL specific options are:
944 ``in|out.buffer-count=count``
945 Sets the count of the buffers.
947 ``-audiodev sndio,id=id[,prop[=value][,...]]``
948 Creates a backend using SNDIO. This backend is available on
949 OpenBSD and most other Unix-like systems.
951 Sndio specific options are:
953 ``in|out.dev=device``
954 Specify the sndio device to use for input and/or output. Default
957 ``in|out.latency=usecs``
958 Sets the desired period length in microseconds.
960 ``-audiodev spice,id=id[,prop[=value][,...]]``
961 Creates a backend that sends audio through SPICE. This backend
962 requires ``-spice`` and automatically selected in that case, so
963 usually you can ignore this option. This backend has no backend
966 ``-audiodev wav,id=id[,prop[=value][,...]]``
967 Creates a backend that writes audio to a WAV file.
969 Backend specific options are:
972 Write recorded audio into the specified file. Default is
976 DEF("device", HAS_ARG, QEMU_OPTION_device,
977 "-device driver[,prop[=value][,...]]\n"
978 " add device (based on driver)\n"
979 " prop=value,... sets driver properties\n"
980 " use '-device help' to print all possible drivers\n"
981 " use '-device driver,help' to print all possible properties\n",
984 ``-device driver[,prop[=value][,...]]``
985 Add device driver. prop=value sets driver properties. Valid
986 properties depend on the driver. To get help on possible drivers and
987 properties, use ``-device help`` and ``-device driver,help``.
991 ``-device ipmi-bmc-sim,id=id[,prop[=value][,...]]``
992 Add an IPMI BMC. This is a simulation of a hardware management
993 interface processor that normally sits on a system. It provides a
994 watchdog and the ability to reset and power control the system. You
995 need to connect this to an IPMI interface to make it useful
997 The IPMI slave address to use for the BMC. The default is 0x20. This
998 address is the BMC's address on the I2C network of management
999 controllers. If you don't know what this means, it is safe to ignore
1003 The BMC id for interfaces to use this device.
1006 Define slave address to use for the BMC. The default is 0x20.
1009 file containing raw Sensor Data Records (SDR) data. The default
1013 size of a Field Replaceable Unit (FRU) area. The default is
1016 ``frudatafile=file``
1017 file containing raw Field Replaceable Unit (FRU) inventory data.
1018 The default is none.
1021 value for the GUID for the BMC, in standard UUID format. If this
1022 is set, get "Get GUID" command to the BMC will return it.
1023 Otherwise "Get GUID" will return an error.
1025 ``-device ipmi-bmc-extern,id=id,chardev=id[,slave_addr=val]``
1026 Add a connection to an external IPMI BMC simulator. Instead of
1027 locally emulating the BMC like the above item, instead connect to an
1028 external entity that provides the IPMI services.
1030 A connection is made to an external BMC simulator. If you do this,
1031 it is strongly recommended that you use the "reconnect=" chardev
1032 option to reconnect to the simulator if the connection is lost. Note
1033 that if this is not used carefully, it can be a security issue, as
1034 the interface has the ability to send resets, NMIs, and power off
1035 the VM. It's best if QEMU makes a connection to an external
1036 simulator running on a secure port on localhost, so neither the
1037 simulator nor QEMU is exposed to any outside network.
1039 See the "lanserv/README.vm" file in the OpenIPMI library for more
1040 details on the external interface.
1042 ``-device isa-ipmi-kcs,bmc=id[,ioport=val][,irq=val]``
1043 Add a KCS IPMI interafce on the ISA bus. This also adds a
1044 corresponding ACPI and SMBIOS entries, if appropriate.
1047 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern
1051 Define the I/O address of the interface. The default is 0xca0
1055 Define the interrupt to use. The default is 5. To disable
1056 interrupts, set this to 0.
1058 ``-device isa-ipmi-bt,bmc=id[,ioport=val][,irq=val]``
1059 Like the KCS interface, but defines a BT interface. The default port
1060 is 0xe4 and the default interrupt is 5.
1062 ``-device pci-ipmi-kcs,bmc=id``
1063 Add a KCS IPMI interafce on the PCI bus.
1066 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern above.
1068 ``-device pci-ipmi-bt,bmc=id``
1069 Like the KCS interface, but defines a BT interface on the PCI bus.
1071 ``-device intel-iommu[,option=...]``
1072 This is only supported by ``-machine q35``, which will enable Intel VT-d
1073 emulation within the guest. It supports below options:
1075 ``intremap=on|off`` (default: auto)
1076 This enables interrupt remapping feature. It's required to enable
1077 complete x2apic. Currently it only supports kvm kernel-irqchip modes
1078 ``off`` or ``split``, while full kernel-irqchip is not yet supported.
1079 The default value is "auto", which will be decided by the mode of
1082 ``caching-mode=on|off`` (default: off)
1083 This enables caching mode for the VT-d emulated device. When
1084 caching-mode is enabled, each guest DMA buffer mapping will generate an
1085 IOTLB invalidation from the guest IOMMU driver to the vIOMMU device in
1086 a synchronous way. It is required for ``-device vfio-pci`` to work
1087 with the VT-d device, because host assigned devices requires to setup
1088 the DMA mapping on the host before guest DMA starts.
1090 ``device-iotlb=on|off`` (default: off)
1091 This enables device-iotlb capability for the emulated VT-d device. So
1092 far virtio/vhost should be the only real user for this parameter,
1093 paired with ats=on configured for the device.
1095 ``aw-bits=39|48`` (default: 39)
1096 This decides the address width of IOVA address space. The address
1097 space has 39 bits width for 3-level IOMMU page tables, and 48 bits for
1098 4-level IOMMU page tables.
1100 Please also refer to the wiki page for general scenarios of VT-d
1101 emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
1105 DEF("name", HAS_ARG, QEMU_OPTION_name,
1106 "-name string1[,process=string2][,debug-threads=on|off]\n"
1107 " set the name of the guest\n"
1108 " string1 sets the window title and string2 the process name\n"
1109 " When debug-threads is enabled, individual threads are given a separate name\n"
1110 " NOTE: The thread names are for debugging and not a stable API.\n",
1114 Sets the name of the guest. This name will be displayed in the SDL
1115 window caption. The name will also be used for the VNC server. Also
1116 optionally set the top visible process name in Linux. Naming of
1117 individual threads can also be enabled on Linux to aid debugging.
1120 DEF("uuid", HAS_ARG, QEMU_OPTION_uuid,
1121 "-uuid %08x-%04x-%04x-%04x-%012x\n"
1122 " specify machine UUID\n", QEMU_ARCH_ALL)
1130 DEFHEADING(Block device options:)
1133 The QEMU block device handling options have a long history and
1134 have gone through several iterations as the feature set and complexity
1135 of the block layer have grown. Many online guides to QEMU often
1136 reference older and deprecated options, which can lead to confusion.
1138 The recommended modern way to describe disks is to use a combination of
1139 ``-device`` to specify the hardware device and ``-blockdev`` to
1140 describe the backend. The device defines what the guest sees and the
1141 backend describes how QEMU handles the data.
1145 DEF("fda", HAS_ARG, QEMU_OPTION_fda,
1146 "-fda/-fdb file use 'file' as floppy disk 0/1 image\n", QEMU_ARCH_ALL)
1147 DEF("fdb", HAS_ARG, QEMU_OPTION_fdb, "", QEMU_ARCH_ALL)
1152 Use file as floppy disk 0/1 image (see the :ref:`disk images` chapter in
1153 the System Emulation Users Guide).
1156 DEF("hda", HAS_ARG, QEMU_OPTION_hda,
1157 "-hda/-hdb file use 'file' as IDE hard disk 0/1 image\n", QEMU_ARCH_ALL)
1158 DEF("hdb", HAS_ARG, QEMU_OPTION_hdb, "", QEMU_ARCH_ALL)
1159 DEF("hdc", HAS_ARG, QEMU_OPTION_hdc,
1160 "-hdc/-hdd file use 'file' as IDE hard disk 2/3 image\n", QEMU_ARCH_ALL)
1161 DEF("hdd", HAS_ARG, QEMU_OPTION_hdd, "", QEMU_ARCH_ALL)
1170 Use file as hard disk 0, 1, 2 or 3 image (see the :ref:`disk images`
1171 chapter in the System Emulation Users Guide).
1174 DEF("cdrom", HAS_ARG, QEMU_OPTION_cdrom,
1175 "-cdrom file use 'file' as IDE cdrom image (cdrom is ide1 master)\n",
1179 Use file as CD-ROM image (you cannot use ``-hdc`` and ``-cdrom`` at
1180 the same time). You can use the host CD-ROM by using ``/dev/cdrom``
1184 DEF("blockdev", HAS_ARG, QEMU_OPTION_blockdev,
1185 "-blockdev [driver=]driver[,node-name=N][,discard=ignore|unmap]\n"
1186 " [,cache.direct=on|off][,cache.no-flush=on|off]\n"
1187 " [,read-only=on|off][,auto-read-only=on|off]\n"
1188 " [,force-share=on|off][,detect-zeroes=on|off|unmap]\n"
1189 " [,driver specific parameters...]\n"
1190 " configure a block backend\n", QEMU_ARCH_ALL)
1192 ``-blockdev option[,option[,option[,...]]]``
1193 Define a new block driver node. Some of the options apply to all
1194 block drivers, other options are only accepted for a specific block
1195 driver. See below for a list of generic options and options for the
1196 most common block drivers.
1198 Options that expect a reference to another node (e.g. ``file``) can
1199 be given in two ways. Either you specify the node name of an already
1200 existing node (file=node-name), or you define a new node inline,
1201 adding options for the referenced node after a dot
1202 (file.filename=path,file.aio=native).
1204 A block driver node created with ``-blockdev`` can be used for a
1205 guest device by specifying its node name for the ``drive`` property
1206 in a ``-device`` argument that defines a block device.
1208 ``Valid options for any block driver node:``
1210 Specifies the block driver to use for the given node.
1213 This defines the name of the block driver node by which it
1214 will be referenced later. The name must be unique, i.e. it
1215 must not match the name of a different block driver node, or
1216 (if you use ``-drive`` as well) the ID of a drive.
1218 If no node name is specified, it is automatically generated.
1219 The generated node name is not intended to be predictable
1220 and changes between QEMU invocations. For the top level, an
1221 explicit node name must be specified.
1224 Open the node read-only. Guest write attempts will fail.
1226 Note that some block drivers support only read-only access,
1227 either generally or in certain configurations. In this case,
1228 the default value ``read-only=off`` does not work and the
1229 option must be specified explicitly.
1232 If ``auto-read-only=on`` is set, QEMU may fall back to
1233 read-only usage even when ``read-only=off`` is requested, or
1234 even switch between modes as needed, e.g. depending on
1235 whether the image file is writable or whether a writing user
1236 is attached to the node.
1239 Override the image locking system of QEMU by forcing the
1240 node to utilize weaker shared access for permissions where
1241 it would normally request exclusive access. When there is
1242 the potential for multiple instances to have the same file
1243 open (whether this invocation of QEMU is the first or the
1244 second instance), both instances must permit shared access
1245 for the second instance to succeed at opening the file.
1247 Enabling ``force-share=on`` requires ``read-only=on``.
1250 The host page cache can be avoided with ``cache.direct=on``.
1251 This will attempt to do disk IO directly to the guest's
1252 memory. QEMU may still perform an internal copy of the data.
1255 In case you don't care about data integrity over host
1256 failures, you can use ``cache.no-flush=on``. This option
1257 tells QEMU that it never needs to write any data to the disk
1258 but can instead keep things in cache. If anything goes
1259 wrong, like your host losing power, the disk storage getting
1260 disconnected accidentally, etc. your image will most
1261 probably be rendered unusable.
1264 discard is one of "ignore" (or "off") or "unmap" (or "on")
1265 and controls whether ``discard`` (also known as ``trim`` or
1266 ``unmap``) requests are ignored or passed to the filesystem.
1267 Some machine types may not support discard requests.
1269 ``detect-zeroes=detect-zeroes``
1270 detect-zeroes is "off", "on" or "unmap" and enables the
1271 automatic conversion of plain zero writes by the OS to
1272 driver specific optimized zero write commands. You may even
1273 choose "unmap" if discard is set to "unmap" to allow a zero
1274 write to be converted to an ``unmap`` operation.
1276 ``Driver-specific options for file``
1277 This is the protocol-level block driver for accessing regular
1281 The path to the image file in the local filesystem
1284 Specifies the AIO backend (threads/native/io_uring,
1288 Specifies whether the image file is protected with Linux OFD
1289 / POSIX locks. The default is to use the Linux Open File
1290 Descriptor API if available, otherwise no lock is applied.
1291 (auto/on/off, default: auto)
1297 -blockdev driver=file,node-name=disk,filename=disk.img
1299 ``Driver-specific options for raw``
1300 This is the image format block driver for raw images. It is
1301 usually stacked on top of a protocol level block driver such as
1305 Reference to or definition of the data source block driver
1306 node (e.g. a ``file`` driver node)
1312 -blockdev driver=file,node-name=disk_file,filename=disk.img
1313 -blockdev driver=raw,node-name=disk,file=disk_file
1319 -blockdev driver=raw,node-name=disk,file.driver=file,file.filename=disk.img
1321 ``Driver-specific options for qcow2``
1322 This is the image format block driver for qcow2 images. It is
1323 usually stacked on top of a protocol level block driver such as
1327 Reference to or definition of the data source block driver
1328 node (e.g. a ``file`` driver node)
1331 Reference to or definition of the backing file block device
1332 (default is taken from the image file). It is allowed to
1333 pass ``null`` here in order to disable the default backing
1337 Whether to enable the lazy refcounts feature (on/off;
1338 default is taken from the image file)
1341 The maximum total size of the L2 table and refcount block
1342 caches in bytes (default: the sum of l2-cache-size and
1343 refcount-cache-size)
1346 The maximum size of the L2 table cache in bytes (default: if
1347 cache-size is not specified - 32M on Linux platforms, and 8M
1348 on non-Linux platforms; otherwise, as large as possible
1349 within the cache-size, while permitting the requested or the
1350 minimal refcount cache size)
1352 ``refcount-cache-size``
1353 The maximum size of the refcount block cache in bytes
1354 (default: 4 times the cluster size; or if cache-size is
1355 specified, the part of it which is not used for the L2
1358 ``cache-clean-interval``
1359 Clean unused entries in the L2 and refcount caches. The
1360 interval is in seconds. The default value is 600 on
1361 supporting platforms, and 0 on other platforms. Setting it
1362 to 0 disables this feature.
1364 ``pass-discard-request``
1365 Whether discard requests to the qcow2 device should be
1366 forwarded to the data source (on/off; default: on if
1367 discard=unmap is specified, off otherwise)
1369 ``pass-discard-snapshot``
1370 Whether discard requests for the data source should be
1371 issued when a snapshot operation (e.g. deleting a snapshot)
1372 frees clusters in the qcow2 file (on/off; default: on)
1374 ``pass-discard-other``
1375 Whether discard requests for the data source should be
1376 issued on other occasions where a cluster gets freed
1377 (on/off; default: off)
1380 Which overlap checks to perform for writes to the image
1381 (none/constant/cached/all; default: cached). For details or
1382 finer granularity control refer to the QAPI documentation of
1389 -blockdev driver=file,node-name=my_file,filename=/tmp/disk.qcow2
1390 -blockdev driver=qcow2,node-name=hda,file=my_file,overlap-check=none,cache-size=16777216
1396 -blockdev driver=qcow2,node-name=disk,file.driver=http,file.filename=http://example.com/image.qcow2
1398 ``Driver-specific options for other drivers``
1399 Please refer to the QAPI documentation of the ``blockdev-add``
1403 DEF("drive", HAS_ARG, QEMU_OPTION_drive,
1404 "-drive [file=file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n"
1405 " [,cache=writethrough|writeback|none|directsync|unsafe][,format=f]\n"
1406 " [,snapshot=on|off][,rerror=ignore|stop|report]\n"
1407 " [,werror=ignore|stop|report|enospc][,id=name]\n"
1408 " [,aio=threads|native|io_uring]\n"
1409 " [,readonly=on|off][,copy-on-read=on|off]\n"
1410 " [,discard=ignore|unmap][,detect-zeroes=on|off|unmap]\n"
1411 " [[,bps=b]|[[,bps_rd=r][,bps_wr=w]]]\n"
1412 " [[,iops=i]|[[,iops_rd=r][,iops_wr=w]]]\n"
1413 " [[,bps_max=bm]|[[,bps_rd_max=rm][,bps_wr_max=wm]]]\n"
1414 " [[,iops_max=im]|[[,iops_rd_max=irm][,iops_wr_max=iwm]]]\n"
1415 " [[,iops_size=is]]\n"
1417 " use 'file' as a drive image\n", QEMU_ARCH_ALL)
1419 ``-drive option[,option[,option[,...]]]``
1420 Define a new drive. This includes creating a block driver node (the
1421 backend) as well as a guest device, and is mostly a shortcut for
1422 defining the corresponding ``-blockdev`` and ``-device`` options.
1424 ``-drive`` accepts all options that are accepted by ``-blockdev``.
1425 In addition, it knows the following options:
1428 This option defines which disk image (see the :ref:`disk images`
1429 chapter in the System Emulation Users Guide) to use with this drive.
1430 If the filename contains comma, you must double it (for instance,
1431 "file=my,,file" to use file "my,file").
1433 Special files such as iSCSI devices can be specified using
1434 protocol specific URLs. See the section for "Device URL Syntax"
1435 for more information.
1438 This option defines on which type on interface the drive is
1439 connected. Available types are: ide, scsi, sd, mtd, floppy,
1440 pflash, virtio, none.
1442 ``bus=bus,unit=unit``
1443 These options define where is connected the drive by defining
1444 the bus number and the unit id.
1447 This option defines where the drive is connected by using an
1448 index in the list of available connectors of a given interface
1452 This option defines the type of the media: disk or cdrom.
1454 ``snapshot=snapshot``
1455 snapshot is "on" or "off" and controls snapshot mode for the
1456 given drive (see ``-snapshot``).
1459 cache is "none", "writeback", "unsafe", "directsync" or
1460 "writethrough" and controls how the host cache is used to access
1461 block data. This is a shortcut that sets the ``cache.direct``
1462 and ``cache.no-flush`` options (as in ``-blockdev``), and
1463 additionally ``cache.writeback``, which provides a default for
1464 the ``write-cache`` option of block guest devices (as in
1465 ``-device``). The modes correspond to the following settings:
1467 ============= =============== ============ ==============
1468 \ cache.writeback cache.direct cache.no-flush
1469 ============= =============== ============ ==============
1470 writeback on off off
1472 writethrough off off off
1473 directsync off on off
1475 ============= =============== ============ ==============
1477 The default mode is ``cache=writeback``.
1480 aio is "threads", "native", or "io_uring" and selects between pthread
1481 based disk I/O, native Linux AIO, or Linux io_uring API.
1484 Specify which disk format will be used rather than detecting the
1485 format. Can be used to specify format=raw to avoid interpreting
1486 an untrusted format header.
1488 ``werror=action,rerror=action``
1489 Specify which action to take on write and read errors. Valid
1490 actions are: "ignore" (ignore the error and try to continue),
1491 "stop" (pause QEMU), "report" (report the error to the guest),
1492 "enospc" (pause QEMU only if the host disk is full; report the
1493 error to the guest otherwise). The default setting is
1494 ``werror=enospc`` and ``rerror=report``.
1496 ``copy-on-read=copy-on-read``
1497 copy-on-read is "on" or "off" and enables whether to copy read
1498 backing file sectors into the image file.
1500 ``bps=b,bps_rd=r,bps_wr=w``
1501 Specify bandwidth throttling limits in bytes per second, either
1502 for all request types or for reads or writes only. Small values
1503 can lead to timeouts or hangs inside the guest. A safe minimum
1504 for disks is 2 MB/s.
1506 ``bps_max=bm,bps_rd_max=rm,bps_wr_max=wm``
1507 Specify bursts in bytes per second, either for all request types
1508 or for reads or writes only. Bursts allow the guest I/O to spike
1509 above the limit temporarily.
1511 ``iops=i,iops_rd=r,iops_wr=w``
1512 Specify request rate limits in requests per second, either for
1513 all request types or for reads or writes only.
1515 ``iops_max=bm,iops_rd_max=rm,iops_wr_max=wm``
1516 Specify bursts in requests per second, either for all request
1517 types or for reads or writes only. Bursts allow the guest I/O to
1518 spike above the limit temporarily.
1521 Let every is bytes of a request count as a new request for iops
1522 throttling purposes. Use this option to prevent guests from
1523 circumventing iops limits by sending fewer but larger requests.
1526 Join a throttling quota group with given name g. All drives that
1527 are members of the same group are accounted for together. Use
1528 this option to prevent guests from circumventing throttling
1529 limits by using many small disks instead of a single larger
1532 By default, the ``cache.writeback=on`` mode is used. It will report
1533 data writes as completed as soon as the data is present in the host
1534 page cache. This is safe as long as your guest OS makes sure to
1535 correctly flush disk caches where needed. If your guest OS does not
1536 handle volatile disk write caches correctly and your host crashes or
1537 loses power, then the guest may experience data corruption.
1539 For such guests, you should consider using ``cache.writeback=off``.
1540 This means that the host page cache will be used to read and write
1541 data, but write notification will be sent to the guest only after
1542 QEMU has made sure to flush each write to the disk. Be aware that
1543 this has a major impact on performance.
1545 When using the ``-snapshot`` option, unsafe caching is always used.
1547 Copy-on-read avoids accessing the same backing file sectors
1548 repeatedly and is useful when the backing file is over a slow
1549 network. By default copy-on-read is off.
1551 Instead of ``-cdrom`` you can use:
1555 |qemu_system| -drive file=file,index=2,media=cdrom
1557 Instead of ``-hda``, ``-hdb``, ``-hdc``, ``-hdd``, you can use:
1561 |qemu_system| -drive file=file,index=0,media=disk
1562 |qemu_system| -drive file=file,index=1,media=disk
1563 |qemu_system| -drive file=file,index=2,media=disk
1564 |qemu_system| -drive file=file,index=3,media=disk
1566 You can open an image using pre-opened file descriptors from an fd
1572 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
1573 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
1574 -drive file=/dev/fdset/2,index=0,media=disk
1576 You can connect a CDROM to the slave of ide0:
1580 |qemu_system_x86| -drive file=file,if=ide,index=1,media=cdrom
1582 If you don't specify the "file=" argument, you define an empty
1587 |qemu_system_x86| -drive if=ide,index=1,media=cdrom
1589 Instead of ``-fda``, ``-fdb``, you can use:
1593 |qemu_system_x86| -drive file=file,index=0,if=floppy
1594 |qemu_system_x86| -drive file=file,index=1,if=floppy
1596 By default, interface is "ide" and index is automatically
1601 |qemu_system_x86| -drive file=a -drive file=b"
1603 is interpreted like:
1607 |qemu_system_x86| -hda a -hdb b
1610 DEF("mtdblock", HAS_ARG, QEMU_OPTION_mtdblock,
1611 "-mtdblock file use 'file' as on-board Flash memory image\n",
1615 Use file as on-board Flash memory image.
1618 DEF("sd", HAS_ARG, QEMU_OPTION_sd,
1619 "-sd file use 'file' as SecureDigital card image\n", QEMU_ARCH_ALL)
1622 Use file as SecureDigital card image.
1625 DEF("snapshot", 0, QEMU_OPTION_snapshot,
1626 "-snapshot write to temporary files instead of disk image files\n",
1630 Write to temporary files instead of disk image files. In this case,
1631 the raw disk image you use is not written back. You can however
1632 force the write back by pressing C-a s (see the :ref:`disk images`
1633 chapter in the System Emulation Users Guide).
1636 DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev,
1637 "-fsdev local,id=id,path=path,security_model=mapped-xattr|mapped-file|passthrough|none\n"
1638 " [,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode]\n"
1639 " [[,throttling.bps-total=b]|[[,throttling.bps-read=r][,throttling.bps-write=w]]]\n"
1640 " [[,throttling.iops-total=i]|[[,throttling.iops-read=r][,throttling.iops-write=w]]]\n"
1641 " [[,throttling.bps-total-max=bm]|[[,throttling.bps-read-max=rm][,throttling.bps-write-max=wm]]]\n"
1642 " [[,throttling.iops-total-max=im]|[[,throttling.iops-read-max=irm][,throttling.iops-write-max=iwm]]]\n"
1643 " [[,throttling.iops-size=is]]\n"
1644 "-fsdev proxy,id=id,socket=socket[,writeout=immediate][,readonly=on]\n"
1645 "-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=immediate][,readonly=on]\n"
1646 "-fsdev synth,id=id\n",
1650 ``-fsdev local,id=id,path=path,security_model=security_model [,writeout=writeout][,readonly=on][,fmode=fmode][,dmode=dmode] [,throttling.option=value[,throttling.option=value[,...]]]``
1652 ``-fsdev proxy,id=id,socket=socket[,writeout=writeout][,readonly=on]``
1654 ``-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=writeout][,readonly=on]``
1656 ``-fsdev synth,id=id[,readonly=on]``
1657 Define a new file system device. Valid options are:
1660 Accesses to the filesystem are done by QEMU.
1663 Accesses to the filesystem are done by virtfs-proxy-helper(1).
1666 Synthetic filesystem, only used by QTests.
1669 Specifies identifier for this device.
1672 Specifies the export path for the file system device. Files
1673 under this path will be available to the 9p client on the guest.
1675 ``security_model=security_model``
1676 Specifies the security model to be used for this export path.
1677 Supported security models are "passthrough", "mapped-xattr",
1678 "mapped-file" and "none". In "passthrough" security model, files
1679 are stored using the same credentials as they are created on the
1680 guest. This requires QEMU to run as root. In "mapped-xattr"
1681 security model, some of the file attributes like uid, gid, mode
1682 bits and link target are stored as file attributes. For
1683 "mapped-file" these attributes are stored in the hidden
1684 .virtfs\_metadata directory. Directories exported by this
1685 security model cannot interact with other unix tools. "none"
1686 security model is same as passthrough except the sever won't
1687 report failures if it fails to set file attributes like
1688 ownership. Security model is mandatory only for local fsdriver.
1689 Other fsdrivers (like proxy) don't take security model as a
1692 ``writeout=writeout``
1693 This is an optional argument. The only supported value is
1694 "immediate". This means that host page cache will be used to
1695 read and write data but write notification will be sent to the
1696 guest only when the data has been reported as written by the
1700 Enables exporting 9p share as a readonly mount for guests. By
1701 default read-write access is given.
1704 Enables proxy filesystem driver to use passed socket file for
1705 communicating with virtfs-proxy-helper(1).
1708 Enables proxy filesystem driver to use passed socket descriptor
1709 for communicating with virtfs-proxy-helper(1). Usually a helper
1710 like libvirt will create socketpair and pass one of the fds as
1714 Specifies the default mode for newly created files on the host.
1715 Works only with security models "mapped-xattr" and
1719 Specifies the default mode for newly created directories on the
1720 host. Works only with security models "mapped-xattr" and
1723 ``throttling.bps-total=b,throttling.bps-read=r,throttling.bps-write=w``
1724 Specify bandwidth throttling limits in bytes per second, either
1725 for all request types or for reads or writes only.
1727 ``throttling.bps-total-max=bm,bps-read-max=rm,bps-write-max=wm``
1728 Specify bursts in bytes per second, either for all request types
1729 or for reads or writes only. Bursts allow the guest I/O to spike
1730 above the limit temporarily.
1732 ``throttling.iops-total=i,throttling.iops-read=r, throttling.iops-write=w``
1733 Specify request rate limits in requests per second, either for
1734 all request types or for reads or writes only.
1736 ``throttling.iops-total-max=im,throttling.iops-read-max=irm, throttling.iops-write-max=iwm``
1737 Specify bursts in requests per second, either for all request
1738 types or for reads or writes only. Bursts allow the guest I/O to
1739 spike above the limit temporarily.
1741 ``throttling.iops-size=is``
1742 Let every is bytes of a request count as a new request for iops
1743 throttling purposes.
1745 -fsdev option is used along with -device driver "virtio-9p-...".
1747 ``-device virtio-9p-type,fsdev=id,mount_tag=mount_tag``
1748 Options for virtio-9p-... driver are:
1751 Specifies the variant to be used. Supported values are "pci",
1752 "ccw" or "device", depending on the machine type.
1755 Specifies the id value specified along with -fsdev option.
1757 ``mount_tag=mount_tag``
1758 Specifies the tag name to be used by the guest to mount this
1762 DEF("virtfs", HAS_ARG, QEMU_OPTION_virtfs,
1763 "-virtfs local,path=path,mount_tag=tag,security_model=mapped-xattr|mapped-file|passthrough|none\n"
1764 " [,id=id][,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode][,multidevs=remap|forbid|warn]\n"
1765 "-virtfs proxy,mount_tag=tag,socket=socket[,id=id][,writeout=immediate][,readonly=on]\n"
1766 "-virtfs proxy,mount_tag=tag,sock_fd=sock_fd[,id=id][,writeout=immediate][,readonly=on]\n"
1767 "-virtfs synth,mount_tag=tag[,id=id][,readonly=on]\n",
1771 ``-virtfs local,path=path,mount_tag=mount_tag ,security_model=security_model[,writeout=writeout][,readonly=on] [,fmode=fmode][,dmode=dmode][,multidevs=multidevs]``
1773 ``-virtfs proxy,socket=socket,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1775 ``-virtfs proxy,sock_fd=sock_fd,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1777 ``-virtfs synth,mount_tag=mount_tag``
1778 Define a new virtual filesystem device and expose it to the guest using
1779 a virtio-9p-device (a.k.a. 9pfs), which essentially means that a certain
1780 directory on host is made directly accessible by guest as a pass-through
1781 file system by using the 9P network protocol for communication between
1782 host and guests, if desired even accessible, shared by several guests
1785 Note that ``-virtfs`` is actually just a convenience shortcut for its
1786 generalized form ``-fsdev -device virtio-9p-pci``.
1788 The general form of pass-through file system options are:
1791 Accesses to the filesystem are done by QEMU.
1794 Accesses to the filesystem are done by virtfs-proxy-helper(1).
1797 Synthetic filesystem, only used by QTests.
1800 Specifies identifier for the filesystem device
1803 Specifies the export path for the file system device. Files
1804 under this path will be available to the 9p client on the guest.
1806 ``security_model=security_model``
1807 Specifies the security model to be used for this export path.
1808 Supported security models are "passthrough", "mapped-xattr",
1809 "mapped-file" and "none". In "passthrough" security model, files
1810 are stored using the same credentials as they are created on the
1811 guest. This requires QEMU to run as root. In "mapped-xattr"
1812 security model, some of the file attributes like uid, gid, mode
1813 bits and link target are stored as file attributes. For
1814 "mapped-file" these attributes are stored in the hidden
1815 .virtfs\_metadata directory. Directories exported by this
1816 security model cannot interact with other unix tools. "none"
1817 security model is same as passthrough except the sever won't
1818 report failures if it fails to set file attributes like
1819 ownership. Security model is mandatory only for local fsdriver.
1820 Other fsdrivers (like proxy) don't take security model as a
1823 ``writeout=writeout``
1824 This is an optional argument. The only supported value is
1825 "immediate". This means that host page cache will be used to
1826 read and write data but write notification will be sent to the
1827 guest only when the data has been reported as written by the
1831 Enables exporting 9p share as a readonly mount for guests. By
1832 default read-write access is given.
1835 Enables proxy filesystem driver to use passed socket file for
1836 communicating with virtfs-proxy-helper(1). Usually a helper like
1837 libvirt will create socketpair and pass one of the fds as
1841 Enables proxy filesystem driver to use passed 'sock\_fd' as the
1842 socket descriptor for interfacing with virtfs-proxy-helper(1).
1845 Specifies the default mode for newly created files on the host.
1846 Works only with security models "mapped-xattr" and
1850 Specifies the default mode for newly created directories on the
1851 host. Works only with security models "mapped-xattr" and
1854 ``mount_tag=mount_tag``
1855 Specifies the tag name to be used by the guest to mount this
1858 ``multidevs=multidevs``
1859 Specifies how to deal with multiple devices being shared with a
1860 9p export. Supported behaviours are either "remap", "forbid" or
1861 "warn". The latter is the default behaviour on which virtfs 9p
1862 expects only one device to be shared with the same export, and
1863 if more than one device is shared and accessed via the same 9p
1864 export then only a warning message is logged (once) by qemu on
1865 host side. In order to avoid file ID collisions on guest you
1866 should either create a separate virtfs export for each device to
1867 be shared with guests (recommended way) or you might use "remap"
1868 instead which allows you to share multiple devices with only one
1869 export instead, which is achieved by remapping the original
1870 inode numbers from host to guest in a way that would prevent
1871 such collisions. Remapping inodes in such use cases is required
1872 because the original device IDs from host are never passed and
1873 exposed on guest. Instead all files of an export shared with
1874 virtfs always share the same device id on guest. So two files
1875 with identical inode numbers but from actually different devices
1876 on host would otherwise cause a file ID collision and hence
1877 potential misbehaviours on guest. "forbid" on the other hand
1878 assumes like "warn" that only one device is shared by the same
1879 export, however it will not only log a warning message but also
1880 deny access to additional devices on guest. Note though that
1881 "forbid" does currently not block all possible file access
1882 operations (e.g. readdir() would still return entries from other
1886 DEF("iscsi", HAS_ARG, QEMU_OPTION_iscsi,
1887 "-iscsi [user=user][,password=password]\n"
1888 " [,header-digest=CRC32C|CR32C-NONE|NONE-CRC32C|NONE\n"
1889 " [,initiator-name=initiator-iqn][,id=target-iqn]\n"
1890 " [,timeout=timeout]\n"
1891 " iSCSI session parameters\n", QEMU_ARCH_ALL)
1895 Configure iSCSI session parameters.
1900 DEFHEADING(USB convenience options:)
1902 DEF("usb", 0, QEMU_OPTION_usb,
1903 "-usb enable on-board USB host controller (if not enabled by default)\n",
1907 Enable USB emulation on machine types with an on-board USB host
1908 controller (if not enabled by default). Note that on-board USB host
1909 controllers may not support USB 3.0. In this case
1910 ``-device qemu-xhci`` can be used instead on machines with PCI.
1913 DEF("usbdevice", HAS_ARG, QEMU_OPTION_usbdevice,
1914 "-usbdevice name add the host or guest USB device 'name'\n",
1917 ``-usbdevice devname``
1918 Add the USB device devname, and enable an on-board USB controller
1919 if possible and necessary (just like it can be done via
1920 ``-machine usb=on``). Note that this option is mainly intended for
1921 the user's convenience only. More fine-grained control can be
1922 achieved by selecting a USB host controller (if necessary) and the
1923 desired USB device via the ``-device`` option instead. For example,
1924 instead of using ``-usbdevice mouse`` it is possible to use
1925 ``-device qemu-xhci -device usb-mouse`` to connect the USB mouse
1926 to a USB 3.0 controller instead (at least on machines that support
1927 PCI and do not have an USB controller enabled by default yet).
1928 For more details, see the chapter about
1929 :ref:`Connecting USB devices` in the System Emulation Users Guide.
1930 Possible devices for devname are:
1933 Braille device. This will use BrlAPI to display the braille
1934 output on a real or fake device (i.e. it also creates a
1935 corresponding ``braille`` chardev automatically beside the
1936 ``usb-braille`` USB device).
1939 Standard USB keyboard. Will override the PS/2 keyboard (if present).
1942 Virtual Mouse. This will override the PS/2 mouse emulation when
1946 Pointer device that uses absolute coordinates (like a
1947 touchscreen). This means QEMU is able to report the mouse
1948 position without having to grab the mouse. Also overrides the
1949 PS/2 mouse emulation when activated.
1952 Wacom PenPartner USB tablet.
1959 DEFHEADING(Display options:)
1961 DEF("display", HAS_ARG, QEMU_OPTION_display,
1962 #if defined(CONFIG_SPICE)
1963 "-display spice-app[,gl=on|off]\n"
1965 #if defined(CONFIG_SDL)
1966 "-display sdl[,gl=on|core|es|off][,grab-mod=<mod>][,show-cursor=on|off]\n"
1967 " [,window-close=on|off]\n"
1969 #if defined(CONFIG_GTK)
1970 "-display gtk[,full-screen=on|off][,gl=on|off][,grab-on-hover=on|off]\n"
1971 " [,show-tabs=on|off][,show-cursor=on|off][,window-close=on|off]\n"
1972 " [,show-menubar=on|off]\n"
1974 #if defined(CONFIG_VNC)
1975 "-display vnc=<display>[,<optargs>]\n"
1977 #if defined(CONFIG_CURSES)
1978 "-display curses[,charset=<encoding>]\n"
1980 #if defined(CONFIG_COCOA)
1981 "-display cocoa[,full-grab=on|off][,swap-opt-cmd=on|off]\n"
1983 #if defined(CONFIG_OPENGL)
1984 "-display egl-headless[,rendernode=<file>]\n"
1986 #if defined(CONFIG_DBUS_DISPLAY)
1987 "-display dbus[,addr=<dbusaddr>]\n"
1988 " [,gl=on|core|es|off][,rendernode=<file>]\n"
1990 #if defined(CONFIG_COCOA)
1991 "-display cocoa[,show-cursor=on|off][,left-command-key=on|off]\n"
1994 " select display backend type\n"
1995 " The default display is equivalent to\n "
1996 #if defined(CONFIG_GTK)
1997 "\"-display gtk\"\n"
1998 #elif defined(CONFIG_SDL)
1999 "\"-display sdl\"\n"
2000 #elif defined(CONFIG_COCOA)
2001 "\"-display cocoa\"\n"
2002 #elif defined(CONFIG_VNC)
2003 "\"-vnc localhost:0,to=99,id=default\"\n"
2005 "\"-display none\"\n"
2010 Select type of display to use. Use ``-display help`` to list the available
2011 display types. Valid values for type are
2013 ``spice-app[,gl=on|off]``
2014 Start QEMU as a Spice server and launch the default Spice client
2015 application. The Spice server will redirect the serial consoles
2016 and QEMU monitors. (Since 4.0)
2019 Export the display over D-Bus interfaces. (Since 7.0)
2021 The connection is registered with the "org.qemu" name (and queued when
2024 ``addr=<dbusaddr>`` : D-Bus bus address to connect to.
2026 ``p2p=yes|no`` : Use peer-to-peer connection, accepted via QMP ``add_client``.
2028 ``gl=on|off|core|es`` : Use OpenGL for rendering (the D-Bus interface
2029 will share framebuffers with DMABUF file descriptors).
2032 Display video output via SDL (usually in a separate graphics
2033 window; see the SDL documentation for other possibilities).
2034 Valid parameters are:
2036 ``grab-mod=<mods>`` : Used to select the modifier keys for toggling
2037 the mouse grabbing in conjunction with the "g" key. ``<mods>`` can be
2038 either ``lshift-lctrl-lalt`` or ``rctrl``.
2040 ``gl=on|off|core|es`` : Use OpenGL for displaying
2042 ``show-cursor=on|off`` : Force showing the mouse cursor
2044 ``window-close=on|off`` : Allow to quit qemu with window close button
2047 Display video output in a GTK window. This interface provides
2048 drop-down menus and other UI elements to configure and control
2049 the VM during runtime. Valid parameters are:
2051 ``full-screen=on|off`` : Start in fullscreen mode
2053 ``gl=on|off`` : Use OpenGL for displaying
2055 ``grab-on-hover=on|off`` : Grab keyboard input on mouse hover
2057 ``show-tabs=on|off`` : Display the tab bar for switching between the
2058 various graphical interfaces (e.g. VGA and
2059 virtual console character devices) by default.
2061 ``show-cursor=on|off`` : Force showing the mouse cursor
2063 ``window-close=on|off`` : Allow to quit qemu with window close button
2065 ``show-menubar=on|off`` : Display the main window menubar, defaults to "on"
2067 ``curses[,charset=<encoding>]``
2068 Display video output via curses. For graphics device models
2069 which support a text mode, QEMU can display this output using a
2070 curses/ncurses interface. Nothing is displayed when the graphics
2071 device is in graphical mode or if the graphics device does not
2072 support a text mode. Generally only the VGA device models
2073 support text mode. The font charset used by the guest can be
2074 specified with the ``charset`` option, for example
2075 ``charset=CP850`` for IBM CP850 encoding. The default is
2079 Display video output in a Cocoa window. Mac only. This interface
2080 provides drop-down menus and other UI elements to configure and
2081 control the VM during runtime. Valid parameters are:
2083 ``show-cursor=on|off`` : Force showing the mouse cursor
2085 ``left-command-key=on|off`` : Disable forwarding left command key to host
2087 ``egl-headless[,rendernode=<file>]``
2088 Offload all OpenGL operations to a local DRI device. For any
2089 graphical display, this display needs to be paired with either
2090 VNC or SPICE displays.
2093 Start a VNC server on display <display>
2096 Do not display video output. The guest will still see an
2097 emulated graphics card, but its output will not be displayed to
2098 the QEMU user. This option differs from the -nographic option in
2099 that it only affects what is done with video output; -nographic
2100 also changes the destination of the serial and parallel port
2104 DEF("nographic", 0, QEMU_OPTION_nographic,
2105 "-nographic disable graphical output and redirect serial I/Os to console\n",
2109 Normally, if QEMU is compiled with graphical window support, it
2110 displays output such as guest graphics, guest console, and the QEMU
2111 monitor in a window. With this option, you can totally disable
2112 graphical output so that QEMU is a simple command line application.
2113 The emulated serial port is redirected on the console and muxed with
2114 the monitor (unless redirected elsewhere explicitly). Therefore, you
2115 can still use QEMU to debug a Linux kernel with a serial console.
2116 Use C-a h for help on switching between the console and monitor.
2120 DEF("spice", HAS_ARG, QEMU_OPTION_spice,
2121 "-spice [port=port][,tls-port=secured-port][,x509-dir=<dir>]\n"
2122 " [,x509-key-file=<file>][,x509-key-password=<file>]\n"
2123 " [,x509-cert-file=<file>][,x509-cacert-file=<file>]\n"
2124 " [,x509-dh-key-file=<file>][,addr=addr]\n"
2125 " [,ipv4=on|off][,ipv6=on|off][,unix=on|off]\n"
2126 " [,tls-ciphers=<list>]\n"
2127 " [,tls-channel=[main|display|cursor|inputs|record|playback]]\n"
2128 " [,plaintext-channel=[main|display|cursor|inputs|record|playback]]\n"
2129 " [,sasl=on|off][,disable-ticketing=on|off]\n"
2130 " [,password=<string>][,password-secret=<secret-id>]\n"
2131 " [,image-compression=[auto_glz|auto_lz|quic|glz|lz|off]]\n"
2132 " [,jpeg-wan-compression=[auto|never|always]]\n"
2133 " [,zlib-glz-wan-compression=[auto|never|always]]\n"
2134 " [,streaming-video=[off|all|filter]][,disable-copy-paste=on|off]\n"
2135 " [,disable-agent-file-xfer=on|off][,agent-mouse=[on|off]]\n"
2136 " [,playback-compression=[on|off]][,seamless-migration=[on|off]]\n"
2137 " [,gl=[on|off]][,rendernode=<file>]\n"
2139 " at least one of {port, tls-port} is mandatory\n",
2143 ``-spice option[,option[,...]]``
2144 Enable the spice remote desktop protocol. Valid options are
2147 Set the TCP port spice is listening on for plaintext channels.
2150 Set the IP address spice is listening on. Default is any
2153 ``ipv4=on|off``; \ ``ipv6=on|off``; \ ``unix=on|off``
2154 Force using the specified IP version.
2156 ``password=<string>``
2157 Set the password you need to authenticate.
2159 This option is deprecated and insecure because it leaves the
2160 password visible in the process listing. Use ``password-secret``
2163 ``password-secret=<secret-id>``
2164 Set the ID of the ``secret`` object containing the password
2165 you need to authenticate.
2168 Require that the client use SASL to authenticate with the spice.
2169 The exact choice of authentication method used is controlled
2170 from the system / user's SASL configuration file for the 'qemu'
2171 service. This is typically found in /etc/sasl2/qemu.conf. If
2172 running QEMU as an unprivileged user, an environment variable
2173 SASL\_CONF\_PATH can be used to make it search alternate
2174 locations for the service config. While some SASL auth methods
2175 can also provide data encryption (eg GSSAPI), it is recommended
2176 that SASL always be combined with the 'tls' and 'x509' settings
2177 to enable use of SSL and server certificates. This ensures a
2178 data encryption preventing compromise of authentication
2181 ``disable-ticketing=on|off``
2182 Allow client connects without authentication.
2184 ``disable-copy-paste=on|off``
2185 Disable copy paste between the client and the guest.
2187 ``disable-agent-file-xfer=on|off``
2188 Disable spice-vdagent based file-xfer between the client and the
2192 Set the TCP port spice is listening on for encrypted channels.
2195 Set the x509 file directory. Expects same filenames as -vnc
2198 ``x509-key-file=<file>``; \ ``x509-key-password=<file>``; \ ``x509-cert-file=<file>``; \ ``x509-cacert-file=<file>``; \ ``x509-dh-key-file=<file>``
2199 The x509 file names can also be configured individually.
2201 ``tls-ciphers=<list>``
2202 Specify which ciphers to use.
2204 ``tls-channel=[main|display|cursor|inputs|record|playback]``; \ ``plaintext-channel=[main|display|cursor|inputs|record|playback]``
2205 Force specific channel to be used with or without TLS
2206 encryption. The options can be specified multiple times to
2207 configure multiple channels. The special name "default" can be
2208 used to set the default mode. For channels which are not
2209 explicitly forced into one mode the spice client is allowed to
2210 pick tls/plaintext as he pleases.
2212 ``image-compression=[auto_glz|auto_lz|quic|glz|lz|off]``
2213 Configure image compression (lossless). Default is auto\_glz.
2215 ``jpeg-wan-compression=[auto|never|always]``; \ ``zlib-glz-wan-compression=[auto|never|always]``
2216 Configure wan image compression (lossy for slow links). Default
2219 ``streaming-video=[off|all|filter]``
2220 Configure video stream detection. Default is off.
2222 ``agent-mouse=[on|off]``
2223 Enable/disable passing mouse events via vdagent. Default is on.
2225 ``playback-compression=[on|off]``
2226 Enable/disable audio stream compression (using celt 0.5.1).
2229 ``seamless-migration=[on|off]``
2230 Enable/disable spice seamless migration. Default is off.
2233 Enable/disable OpenGL context. Default is off.
2235 ``rendernode=<file>``
2236 DRM render node for OpenGL rendering. If not specified, it will
2237 pick the first available. (Since 2.9)
2240 DEF("portrait", 0, QEMU_OPTION_portrait,
2241 "-portrait rotate graphical output 90 deg left (only PXA LCD)\n",
2245 Rotate graphical output 90 deg left (only PXA LCD).
2248 DEF("rotate", HAS_ARG, QEMU_OPTION_rotate,
2249 "-rotate <deg> rotate graphical output some deg left (only PXA LCD)\n",
2253 Rotate graphical output some deg left (only PXA LCD).
2256 DEF("vga", HAS_ARG, QEMU_OPTION_vga,
2257 "-vga [std|cirrus|vmware|qxl|xenfb|tcx|cg3|virtio|none]\n"
2258 " select video card type\n", QEMU_ARCH_ALL)
2261 Select type of VGA card to emulate. Valid values for type are
2264 Cirrus Logic GD5446 Video card. All Windows versions starting
2265 from Windows 95 should recognize and use this graphic card. For
2266 optimal performances, use 16 bit color depth in the guest and
2267 the host OS. (This card was the default before QEMU 2.2)
2270 Standard VGA card with Bochs VBE extensions. If your guest OS
2271 supports the VESA 2.0 VBE extensions (e.g. Windows XP) and if
2272 you want to use high resolution modes (>= 1280x1024x16) then you
2273 should use this option. (This card is the default since QEMU
2277 VMWare SVGA-II compatible adapter. Use it if you have
2278 sufficiently recent XFree86/XOrg server or Windows guest with a
2279 driver for this card.
2282 QXL paravirtual graphic card. It is VGA compatible (including
2283 VESA 2.0 VBE support). Works best with qxl guest drivers
2284 installed though. Recommended choice when using the spice
2288 (sun4m only) Sun TCX framebuffer. This is the default
2289 framebuffer for sun4m machines and offers both 8-bit and 24-bit
2290 colour depths at a fixed resolution of 1024x768.
2293 (sun4m only) Sun cgthree framebuffer. This is a simple 8-bit
2294 framebuffer for sun4m machines available in both 1024x768
2295 (OpenBIOS) and 1152x900 (OBP) resolutions aimed at people
2296 wishing to run older Solaris versions.
2305 DEF("full-screen", 0, QEMU_OPTION_full_screen,
2306 "-full-screen start in full screen\n", QEMU_ARCH_ALL)
2309 Start in full screen.
2312 DEF("g", HAS_ARG, QEMU_OPTION_g ,
2313 "-g WxH[xDEPTH] Set the initial graphical resolution and depth\n",
2314 QEMU_ARCH_PPC | QEMU_ARCH_SPARC | QEMU_ARCH_M68K)
2316 ``-g`` *width*\ ``x``\ *height*\ ``[x``\ *depth*\ ``]``
2317 Set the initial graphical resolution and depth (PPC, SPARC only).
2319 For PPC the default is 800x600x32.
2321 For SPARC with the TCX graphics device, the default is 1024x768x8
2322 with the option of 1024x768x24. For cgthree, the default is
2323 1024x768x8 with the option of 1152x900x8 for people who wish to use
2327 DEF("vnc", HAS_ARG, QEMU_OPTION_vnc ,
2328 "-vnc <display> shorthand for -display vnc=<display>\n", QEMU_ARCH_ALL)
2330 ``-vnc display[,option[,option[,...]]]``
2331 Normally, if QEMU is compiled with graphical window support, it
2332 displays output such as guest graphics, guest console, and the QEMU
2333 monitor in a window. With this option, you can have QEMU listen on
2334 VNC display display and redirect the VGA display over the VNC
2335 session. It is very useful to enable the usb tablet device when
2336 using this option (option ``-device usb-tablet``). When using the
2337 VNC display, you must use the ``-k`` parameter to set the keyboard
2338 layout if you are not using en-us. Valid syntax for the display is
2341 With this option, QEMU will try next available VNC displays,
2342 until the number L, if the origianlly defined "-vnc display" is
2343 not available, e.g. port 5900+display is already used by another
2344 application. By default, to=0.
2347 TCP connections will only be allowed from host on display d. By
2348 convention the TCP port is 5900+d. Optionally, host can be
2349 omitted in which case the server will accept connections from
2353 Connections will be allowed over UNIX domain sockets where path
2354 is the location of a unix socket to listen for connections on.
2357 VNC is initialized but not started. The monitor ``change``
2358 command can be used to later start the VNC server.
2360 Following the display value there may be one or more option flags
2361 separated by commas. Valid options are
2364 Connect to a listening VNC client via a "reverse" connection.
2365 The client is specified by the display. For reverse network
2366 connections (host:d,``reverse``), the d argument is a TCP port
2367 number, not a display number.
2369 ``websocket=on|off``
2370 Opens an additional TCP listening port dedicated to VNC
2371 Websocket connections. If a bare websocket option is given, the
2372 Websocket port is 5700+display. An alternative port can be
2373 specified with the syntax ``websocket``\ =port.
2375 If host is specified connections will only be allowed from this
2376 host. It is possible to control the websocket listen address
2377 independently, using the syntax ``websocket``\ =host:port.
2379 If no TLS credentials are provided, the websocket connection
2380 runs in unencrypted mode. If TLS credentials are provided, the
2381 websocket connection requires encrypted client connections.
2384 Require that password based authentication is used for client
2387 The password must be set separately using the ``set_password``
2388 command in the :ref:`QEMU monitor`. The
2389 syntax to change your password is:
2390 ``set_password <protocol> <password>`` where <protocol> could be
2391 either "vnc" or "spice".
2393 If you would like to change <protocol> password expiration, you
2394 should use ``expire_password <protocol> <expiration-time>``
2395 where expiration time could be one of the following options:
2396 now, never, +seconds or UNIX time of expiration, e.g. +60 to
2397 make password expire in 60 seconds, or 1335196800 to make
2398 password expire on "Mon Apr 23 12:00:00 EDT 2012" (UNIX time for
2399 this date and time).
2401 You can also use keywords "now" or "never" for the expiration
2402 time to allow <protocol> password to expire immediately or never
2405 ``password-secret=<secret-id>``
2406 Require that password based authentication is used for client
2407 connections, using the password provided by the ``secret``
2408 object identified by ``secret-id``.
2411 Provides the ID of a set of TLS credentials to use to secure the
2412 VNC server. They will apply to both the normal VNC server socket
2413 and the websocket socket (if enabled). Setting TLS credentials
2414 will cause the VNC server socket to enable the VeNCrypt auth
2415 mechanism. The credentials should have been previously created
2416 using the ``-object tls-creds`` argument.
2419 Provides the ID of the QAuthZ authorization object against which
2420 the client's x509 distinguished name will validated. This object
2421 is only resolved at time of use, so can be deleted and recreated
2422 on the fly while the VNC server is active. If missing, it will
2423 default to denying access.
2426 Require that the client use SASL to authenticate with the VNC
2427 server. The exact choice of authentication method used is
2428 controlled from the system / user's SASL configuration file for
2429 the 'qemu' service. This is typically found in
2430 /etc/sasl2/qemu.conf. If running QEMU as an unprivileged user,
2431 an environment variable SASL\_CONF\_PATH can be used to make it
2432 search alternate locations for the service config. While some
2433 SASL auth methods can also provide data encryption (eg GSSAPI),
2434 it is recommended that SASL always be combined with the 'tls'
2435 and 'x509' settings to enable use of SSL and server
2436 certificates. This ensures a data encryption preventing
2437 compromise of authentication credentials. See the
2438 :ref:`VNC security` section in the System Emulation Users Guide
2439 for details on using SASL authentication.
2442 Provides the ID of the QAuthZ authorization object against which
2443 the client's SASL username will validated. This object is only
2444 resolved at time of use, so can be deleted and recreated on the
2445 fly while the VNC server is active. If missing, it will default
2449 Legacy method for enabling authorization of clients against the
2450 x509 distinguished name and SASL username. It results in the
2451 creation of two ``authz-list`` objects with IDs of
2452 ``vnc.username`` and ``vnc.x509dname``. The rules for these
2453 objects must be configured with the HMP ACL commands.
2455 This option is deprecated and should no longer be used. The new
2456 ``sasl-authz`` and ``tls-authz`` options are a replacement.
2459 Enable lossy compression methods (gradient, JPEG, ...). If this
2460 option is set, VNC client may receive lossy framebuffer updates
2461 depending on its encoding settings. Enabling this option can
2462 save a lot of bandwidth at the expense of quality.
2464 ``non-adaptive=on|off``
2465 Disable adaptive encodings. Adaptive encodings are enabled by
2466 default. An adaptive encoding will try to detect frequently
2467 updated screen regions, and send updates in these regions using
2468 a lossy encoding (like JPEG). This can be really helpful to save
2469 bandwidth when playing videos. Disabling adaptive encodings
2470 restores the original static behavior of encodings like Tight.
2472 ``share=[allow-exclusive|force-shared|ignore]``
2473 Set display sharing policy. 'allow-exclusive' allows clients to
2474 ask for exclusive access. As suggested by the rfb spec this is
2475 implemented by dropping other connections. Connecting multiple
2476 clients in parallel requires all clients asking for a shared
2477 session (vncviewer: -shared switch). This is the default.
2478 'force-shared' disables exclusive client access. Useful for
2479 shared desktop sessions, where you don't want someone forgetting
2480 specify -shared disconnect everybody else. 'ignore' completely
2481 ignores the shared flag and allows everybody connect
2482 unconditionally. Doesn't conform to the rfb spec but is
2483 traditional QEMU behavior.
2486 Set keyboard delay, for key down and key up events, in
2487 milliseconds. Default is 10. Keyboards are low-bandwidth
2488 devices, so this slowdown can help the device and guest to keep
2489 up and not lose events in case events are arriving in bulk.
2490 Possible causes for the latter are flaky network connections, or
2491 scripts for automated testing.
2493 ``audiodev=audiodev``
2494 Use the specified audiodev when the VNC client requests audio
2495 transmission. When not using an -audiodev argument, this option
2496 must be omitted, otherwise is must be present and specify a
2499 ``power-control=on|off``
2500 Permit the remote client to issue shutdown, reboot or reset power
2504 ARCHHEADING(, QEMU_ARCH_I386)
2506 ARCHHEADING(i386 target only:, QEMU_ARCH_I386)
2508 DEF("win2k-hack", 0, QEMU_OPTION_win2k_hack,
2509 "-win2k-hack use it when installing Windows 2000 to avoid a disk full bug\n",
2513 Use it when installing Windows 2000 to avoid a disk full bug. After
2514 Windows 2000 is installed, you no longer need this option (this
2515 option slows down the IDE transfers).
2518 DEF("no-fd-bootchk", 0, QEMU_OPTION_no_fd_bootchk,
2519 "-no-fd-bootchk disable boot signature checking for floppy disks\n",
2523 Disable boot signature checking for floppy disks in BIOS. May be
2524 needed to boot from old floppy disks.
2527 DEF("no-acpi", 0, QEMU_OPTION_no_acpi,
2528 "-no-acpi disable ACPI\n", QEMU_ARCH_I386 | QEMU_ARCH_ARM)
2531 Disable ACPI (Advanced Configuration and Power Interface) support.
2532 Use it if your guest OS complains about ACPI problems (PC target
2536 DEF("no-hpet", 0, QEMU_OPTION_no_hpet,
2537 "-no-hpet disable HPET\n", QEMU_ARCH_I386)
2540 Disable HPET support.
2543 DEF("acpitable", HAS_ARG, QEMU_OPTION_acpitable,
2544 "-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"
2545 " ACPI table description\n", QEMU_ARCH_I386)
2547 ``-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]...]``
2548 Add ACPI table with specified header fields and context from
2549 specified files. For file=, take whole ACPI table from the specified
2550 files, including all ACPI headers (possible overridden by other
2551 options). For data=, only data portion of the table is used, all
2552 header information is specified in the command line. If a SLIC table
2553 is supplied to QEMU, then the SLIC's oem\_id and oem\_table\_id
2554 fields will override the same in the RSDT and the FADT (a.k.a.
2555 FACP), in order to ensure the field matches required by the
2556 Microsoft SLIC spec and the ACPI spec.
2559 DEF("smbios", HAS_ARG, QEMU_OPTION_smbios,
2560 "-smbios file=binary\n"
2561 " load SMBIOS entry from binary file\n"
2562 "-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d]\n"
2564 " specify SMBIOS type 0 fields\n"
2565 "-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str]\n"
2566 " [,uuid=uuid][,sku=str][,family=str]\n"
2567 " specify SMBIOS type 1 fields\n"
2568 "-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str]\n"
2569 " [,asset=str][,location=str]\n"
2570 " specify SMBIOS type 2 fields\n"
2571 "-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str]\n"
2573 " specify SMBIOS type 3 fields\n"
2574 "-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str]\n"
2575 " [,asset=str][,part=str][,max-speed=%d][,current-speed=%d]\n"
2576 " [,processor-id=%d]\n"
2577 " specify SMBIOS type 4 fields\n"
2578 "-smbios type=11[,value=str][,path=filename]\n"
2579 " specify SMBIOS type 11 fields\n"
2580 "-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str]\n"
2581 " [,asset=str][,part=str][,speed=%d]\n"
2582 " specify SMBIOS type 17 fields\n"
2583 "-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]\n"
2584 " specify SMBIOS type 41 fields\n",
2585 QEMU_ARCH_I386 | QEMU_ARCH_ARM)
2587 ``-smbios file=binary``
2588 Load SMBIOS entry from binary file.
2590 ``-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d][,uefi=on|off]``
2591 Specify SMBIOS type 0 fields
2593 ``-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str][,uuid=uuid][,sku=str][,family=str]``
2594 Specify SMBIOS type 1 fields
2596 ``-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str][,asset=str][,location=str]``
2597 Specify SMBIOS type 2 fields
2599 ``-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str][,sku=str]``
2600 Specify SMBIOS type 3 fields
2602 ``-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str][,asset=str][,part=str][,processor-id=%d]``
2603 Specify SMBIOS type 4 fields
2605 ``-smbios type=11[,value=str][,path=filename]``
2606 Specify SMBIOS type 11 fields
2608 This argument can be repeated multiple times, and values are added in the order they are parsed.
2609 Applications intending to use OEM strings data are encouraged to use their application name as
2610 a prefix for the value string. This facilitates passing information for multiple applications
2613 The ``value=str`` syntax provides the string data inline, while the ``path=filename`` syntax
2614 loads data from a file on disk. Note that the file is not permitted to contain any NUL bytes.
2616 Both the ``value`` and ``path`` options can be repeated multiple times and will be added to
2617 the SMBIOS table in the order in which they appear.
2619 Note that on the x86 architecture, the total size of all SMBIOS tables is limited to 65535
2620 bytes. Thus the OEM strings data is not suitable for passing large amounts of data into the
2621 guest. Instead it should be used as a indicator to inform the guest where to locate the real
2622 data set, for example, by specifying the serial ID of a block device.
2624 An example passing three strings is
2628 -smbios type=11,value=cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/,\\
2629 value=anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os,\\
2630 path=/some/file/with/oemstringsdata.txt
2632 In the guest OS this is visible with the ``dmidecode`` command
2637 Handle 0x0E00, DMI type 11, 5 bytes
2639 String 1: cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/
2640 String 2: anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os
2641 String 3: myapp:some extra data
2644 ``-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str][,asset=str][,part=str][,speed=%d]``
2645 Specify SMBIOS type 17 fields
2647 ``-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]``
2648 Specify SMBIOS type 41 fields
2650 This argument can be repeated multiple times. Its main use is to allow network interfaces be created
2651 as ``enoX`` on Linux, with X being the instance number, instead of the name depending on the interface
2652 position on the PCI bus.
2654 Here is an example of use:
2658 -netdev user,id=internet \\
2659 -device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \\
2660 -smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev
2662 In the guest OS, the device should then appear as ``eno1``:
2667 lo UNKNOWN 00:00:00:00:00:00 <LOOPBACK,UP,LOWER_UP>
2668 eno1 UP 50:54:00:00:00:42 <BROADCAST,MULTICAST,UP,LOWER_UP>
2670 Currently, the PCI device has to be attached to the root bus.
2676 DEFHEADING(Network options:)
2678 DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
2680 "-netdev user,id=str[,ipv4=on|off][,net=addr[/mask]][,host=addr]\n"
2681 " [,ipv6=on|off][,ipv6-net=addr[/int]][,ipv6-host=addr]\n"
2682 " [,restrict=on|off][,hostname=host][,dhcpstart=addr]\n"
2683 " [,dns=addr][,ipv6-dns=addr][,dnssearch=domain][,domainname=domain]\n"
2684 " [,tftp=dir][,tftp-server-name=name][,bootfile=f][,hostfwd=rule][,guestfwd=rule]"
2686 "[,smb=dir[,smbserver=addr]]\n"
2688 " configure a user mode network backend with ID 'str',\n"
2689 " its DHCP server and optional services\n"
2692 "-netdev tap,id=str,ifname=name\n"
2693 " configure a host TAP network backend with ID 'str'\n"
2695 "-netdev tap,id=str[,fd=h][,fds=x:y:...:z][,ifname=name][,script=file][,downscript=dfile]\n"
2696 " [,br=bridge][,helper=helper][,sndbuf=nbytes][,vnet_hdr=on|off][,vhost=on|off]\n"
2697 " [,vhostfd=h][,vhostfds=x:y:...:z][,vhostforce=on|off][,queues=n]\n"
2699 " configure a host TAP network backend with ID 'str'\n"
2700 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2701 " use network scripts 'file' (default=" DEFAULT_NETWORK_SCRIPT ")\n"
2702 " to configure it and 'dfile' (default=" DEFAULT_NETWORK_DOWN_SCRIPT ")\n"
2703 " to deconfigure it\n"
2704 " use '[down]script=no' to disable script execution\n"
2705 " use network helper 'helper' (default=" DEFAULT_BRIDGE_HELPER ") to\n"
2707 " use 'fd=h' to connect to an already opened TAP interface\n"
2708 " use 'fds=x:y:...:z' to connect to already opened multiqueue capable TAP interfaces\n"
2709 " use 'sndbuf=nbytes' to limit the size of the send buffer (the\n"
2710 " default is disabled 'sndbuf=0' to enable flow control set 'sndbuf=1048576')\n"
2711 " use vnet_hdr=off to avoid enabling the IFF_VNET_HDR tap flag\n"
2712 " use vnet_hdr=on to make the lack of IFF_VNET_HDR support an error condition\n"
2713 " use vhost=on to enable experimental in kernel accelerator\n"
2714 " (only has effect for virtio guests which use MSIX)\n"
2715 " use vhostforce=on to force vhost on for non-MSIX virtio guests\n"
2716 " use 'vhostfd=h' to connect to an already opened vhost net device\n"
2717 " use 'vhostfds=x:y:...:z to connect to multiple already opened vhost net devices\n"
2718 " use 'queues=n' to specify the number of queues to be created for multiqueue TAP\n"
2719 " use 'poll-us=n' to specify the maximum number of microseconds that could be\n"
2720 " spent on busy polling for vhost net\n"
2721 "-netdev bridge,id=str[,br=bridge][,helper=helper]\n"
2722 " configure a host TAP network backend with ID 'str' that is\n"
2723 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2724 " using the program 'helper (default=" DEFAULT_BRIDGE_HELPER ")\n"
2727 "-netdev l2tpv3,id=str,src=srcaddr,dst=dstaddr[,srcport=srcport][,dstport=dstport]\n"
2728 " [,rxsession=rxsession],txsession=txsession[,ipv6=on|off][,udp=on|off]\n"
2729 " [,cookie64=on|off][,counter][,pincounter][,txcookie=txcookie]\n"
2730 " [,rxcookie=rxcookie][,offset=offset]\n"
2731 " configure a network backend with ID 'str' connected to\n"
2732 " an Ethernet over L2TPv3 pseudowire.\n"
2733 " Linux kernel 3.3+ as well as most routers can talk\n"
2734 " L2TPv3. This transport allows connecting a VM to a VM,\n"
2735 " VM to a router and even VM to Host. It is a nearly-universal\n"
2736 " standard (RFC3931). Note - this implementation uses static\n"
2737 " pre-configured tunnels (same as the Linux kernel).\n"
2738 " use 'src=' to specify source address\n"
2739 " use 'dst=' to specify destination address\n"
2740 " use 'udp=on' to specify udp encapsulation\n"
2741 " use 'srcport=' to specify source udp port\n"
2742 " use 'dstport=' to specify destination udp port\n"
2743 " use 'ipv6=on' to force v6\n"
2744 " L2TPv3 uses cookies to prevent misconfiguration as\n"
2745 " well as a weak security measure\n"
2746 " use 'rxcookie=0x012345678' to specify a rxcookie\n"
2747 " use 'txcookie=0x012345678' to specify a txcookie\n"
2748 " use 'cookie64=on' to set cookie size to 64 bit, otherwise 32\n"
2749 " use 'counter=off' to force a 'cut-down' L2TPv3 with no counter\n"
2750 " use 'pincounter=on' to work around broken counter handling in peer\n"
2751 " use 'offset=X' to add an extra offset between header and data\n"
2753 "-netdev socket,id=str[,fd=h][,listen=[host]:port][,connect=host:port]\n"
2754 " configure a network backend to connect to another network\n"
2755 " using a socket connection\n"
2756 "-netdev socket,id=str[,fd=h][,mcast=maddr:port[,localaddr=addr]]\n"
2757 " configure a network backend to connect to a multicast maddr and port\n"
2758 " use 'localaddr=addr' to specify the host address to send packets from\n"
2759 "-netdev socket,id=str[,fd=h][,udp=host:port][,localaddr=host:port]\n"
2760 " configure a network backend to connect to another network\n"
2761 " using an UDP tunnel\n"
2763 "-netdev vde,id=str[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]\n"
2764 " configure a network backend to connect to port 'n' of a vde switch\n"
2765 " running on host and listening for incoming connections on 'socketpath'.\n"
2766 " Use group 'groupname' and mode 'octalmode' to change default\n"
2767 " ownership and permissions for communication port.\n"
2769 #ifdef CONFIG_NETMAP
2770 "-netdev netmap,id=str,ifname=name[,devname=nmname]\n"
2771 " attach to the existing netmap-enabled network interface 'name', or to a\n"
2772 " VALE port (created on the fly) called 'name' ('nmname' is name of the \n"
2773 " netmap device, defaults to '/dev/netmap')\n"
2776 "-netdev vhost-user,id=str,chardev=dev[,vhostforce=on|off]\n"
2777 " configure a vhost-user network, backed by a chardev 'dev'\n"
2780 "-netdev vhost-vdpa,id=str,vhostdev=/path/to/dev\n"
2781 " configure a vhost-vdpa network,Establish a vhost-vdpa netdev\n"
2784 "-netdev vmnet-host,id=str[,isolated=on|off][,net-uuid=uuid]\n"
2785 " [,start-address=addr,end-address=addr,subnet-mask=mask]\n"
2786 " configure a vmnet network backend in host mode with ID 'str',\n"
2787 " isolate this interface from others with 'isolated',\n"
2788 " configure the address range and choose a subnet mask,\n"
2789 " specify network UUID 'uuid' to disable DHCP and interact with\n"
2790 " vmnet-host interfaces within this isolated network\n"
2791 "-netdev vmnet-shared,id=str[,isolated=on|off][,nat66-prefix=addr]\n"
2792 " [,start-address=addr,end-address=addr,subnet-mask=mask]\n"
2793 " configure a vmnet network backend in shared mode with ID 'str',\n"
2794 " configure the address range and choose a subnet mask,\n"
2795 " set IPv6 ULA prefix (of length 64) to use for internal network,\n"
2796 " isolate this interface from others with 'isolated'\n"
2797 "-netdev vmnet-bridged,id=str,ifname=name[,isolated=on|off]\n"
2798 " configure a vmnet network backend in bridged mode with ID 'str',\n"
2799 " use 'ifname=name' to select a physical network interface to be bridged,\n"
2800 " isolate this interface from others with 'isolated'\n"
2802 "-netdev hubport,id=str,hubid=n[,netdev=nd]\n"
2803 " configure a hub port on the hub with ID 'n'\n", QEMU_ARCH_ALL)
2804 DEF("nic", HAS_ARG, QEMU_OPTION_nic,
2815 #ifdef CONFIG_NETMAP
2822 "vmnet-host|vmnet-shared|vmnet-bridged|"
2824 "socket][,option][,...][mac=macaddr]\n"
2825 " initialize an on-board / default host NIC (using MAC address\n"
2826 " macaddr) and connect it to the given host network backend\n"
2827 "-nic none use it alone to have zero network devices (the default is to\n"
2828 " provided a 'user' network connection)\n",
2830 DEF("net", HAS_ARG, QEMU_OPTION_net,
2831 "-net nic[,macaddr=mac][,model=type][,name=str][,addr=str][,vectors=v]\n"
2832 " configure or create an on-board (or machine default) NIC and\n"
2833 " connect it to hub 0 (please use -nic unless you need a hub)\n"
2843 #ifdef CONFIG_NETMAP
2847 "vmnet-host|vmnet-shared|vmnet-bridged|"
2849 "socket][,option][,option][,...]\n"
2850 " old way to initialize a host network interface\n"
2851 " (use the -netdev option if possible instead)\n", QEMU_ARCH_ALL)
2853 ``-nic [tap|bridge|user|l2tpv3|vde|netmap|vhost-user|socket][,...][,mac=macaddr][,model=mn]``
2854 This option is a shortcut for configuring both the on-board
2855 (default) guest NIC hardware and the host network backend in one go.
2856 The host backend options are the same as with the corresponding
2857 ``-netdev`` options below. The guest NIC model can be set with
2858 ``model=modelname``. Use ``model=help`` to list the available device
2859 types. The hardware MAC address can be set with ``mac=macaddr``.
2861 The following two example do exactly the same, to show how ``-nic``
2862 can be used to shorten the command line length:
2866 |qemu_system| -netdev user,id=n1,ipv6=off -device e1000,netdev=n1,mac=52:54:98:76:54:32
2867 |qemu_system| -nic user,ipv6=off,model=e1000,mac=52:54:98:76:54:32
2870 Indicate that no network devices should be configured. It is used to
2871 override the default configuration (default NIC with "user" host
2872 network backend) which is activated if no other networking options
2875 ``-netdev user,id=id[,option][,option][,...]``
2876 Configure user mode host network backend which requires no
2877 administrator privilege to run. Valid options are:
2880 Assign symbolic name for use in monitor commands.
2882 ``ipv4=on|off and ipv6=on|off``
2883 Specify that either IPv4 or IPv6 must be enabled. If neither is
2884 specified both protocols are enabled.
2887 Set IP network address the guest will see. Optionally specify
2888 the netmask, either in the form a.b.c.d or as number of valid
2889 top-most bits. Default is 10.0.2.0/24.
2892 Specify the guest-visible address of the host. Default is the
2893 2nd IP in the guest network, i.e. x.x.x.2.
2895 ``ipv6-net=addr[/int]``
2896 Set IPv6 network address the guest will see (default is
2897 fec0::/64). The network prefix is given in the usual hexadecimal
2898 IPv6 address notation. The prefix size is optional, and is given
2899 as the number of valid top-most bits (default is 64).
2902 Specify the guest-visible IPv6 address of the host. Default is
2903 the 2nd IPv6 in the guest network, i.e. xxxx::2.
2906 If this option is enabled, the guest will be isolated, i.e. it
2907 will not be able to contact the host and no guest IP packets
2908 will be routed over the host to the outside. This option does
2909 not affect any explicitly set forwarding rules.
2912 Specifies the client hostname reported by the built-in DHCP
2916 Specify the first of the 16 IPs the built-in DHCP server can
2917 assign. Default is the 15th to 31st IP in the guest network,
2918 i.e. x.x.x.15 to x.x.x.31.
2921 Specify the guest-visible address of the virtual nameserver. The
2922 address must be different from the host address. Default is the
2923 3rd IP in the guest network, i.e. x.x.x.3.
2926 Specify the guest-visible address of the IPv6 virtual
2927 nameserver. The address must be different from the host address.
2928 Default is the 3rd IP in the guest network, i.e. xxxx::3.
2930 ``dnssearch=domain``
2931 Provides an entry for the domain-search list sent by the
2932 built-in DHCP server. More than one domain suffix can be
2933 transmitted by specifying this option multiple times. If
2934 supported, this will cause the guest to automatically try to
2935 append the given domain suffix(es) in case a domain name can not
2942 |qemu_system| -nic user,dnssearch=mgmt.example.org,dnssearch=example.org
2944 ``domainname=domain``
2945 Specifies the client domain name reported by the built-in DHCP
2949 When using the user mode network stack, activate a built-in TFTP
2950 server. The files in dir will be exposed as the root of a TFTP
2951 server. The TFTP client on the guest must be configured in
2952 binary mode (use the command ``bin`` of the Unix TFTP client).
2954 ``tftp-server-name=name``
2955 In BOOTP reply, broadcast name as the "TFTP server name"
2956 (RFC2132 option 66). This can be used to advise the guest to
2957 load boot files or configurations from a different server than
2961 When using the user mode network stack, broadcast file as the
2962 BOOTP filename. In conjunction with ``tftp``, this can be used
2963 to network boot a guest from a local directory.
2965 Example (using pxelinux):
2969 |qemu_system| -hda linux.img -boot n -device e1000,netdev=n1 \\
2970 -netdev user,id=n1,tftp=/path/to/tftp/files,bootfile=/pxelinux.0
2972 ``smb=dir[,smbserver=addr]``
2973 When using the user mode network stack, activate a built-in SMB
2974 server so that Windows OSes can access to the host files in
2975 ``dir`` transparently. The IP address of the SMB server can be
2976 set to addr. By default the 4th IP in the guest network is used,
2979 In the guest Windows OS, the line:
2985 must be added in the file ``C:\WINDOWS\LMHOSTS`` (for windows
2986 9x/Me) or ``C:\WINNT\SYSTEM32\DRIVERS\ETC\LMHOSTS`` (Windows
2989 Then ``dir`` can be accessed in ``\\smbserver\qemu``.
2991 Note that a SAMBA server must be installed on the host OS.
2993 ``hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport``
2994 Redirect incoming TCP or UDP connections to the host port
2995 hostport to the guest IP address guestaddr on guest port
2996 guestport. If guestaddr is not specified, its value is x.x.x.15
2997 (default first address given by the built-in DHCP server). By
2998 specifying hostaddr, the rule can be bound to a specific host
2999 interface. If no connection type is set, TCP is used. This
3000 option can be given multiple times.
3002 For example, to redirect host X11 connection from screen 1 to
3003 guest screen 0, use the following:
3008 |qemu_system| -nic user,hostfwd=tcp:127.0.0.1:6001-:6000
3009 # this host xterm should open in the guest X11 server
3012 To redirect telnet connections from host port 5555 to telnet
3013 port on the guest, use the following:
3018 |qemu_system| -nic user,hostfwd=tcp::5555-:23
3019 telnet localhost 5555
3021 Then when you use on the host ``telnet localhost 5555``, you
3022 connect to the guest telnet server.
3024 ``guestfwd=[tcp]:server:port-dev``; \ ``guestfwd=[tcp]:server:port-cmd:command``
3025 Forward guest TCP connections to the IP address server on port
3026 port to the character device dev or to a program executed by
3027 cmd:command which gets spawned for each connection. This option
3028 can be given multiple times.
3030 You can either use a chardev directly and have that one used
3031 throughout QEMU's lifetime, like in the following example:
3035 # open 10.10.1.1:4321 on bootup, connect 10.0.2.100:1234 to it whenever
3036 # the guest accesses it
3037 |qemu_system| -nic user,guestfwd=tcp:10.0.2.100:1234-tcp:10.10.1.1:4321
3039 Or you can execute a command on every TCP connection established
3040 by the guest, so that QEMU behaves similar to an inetd process
3041 for that virtual server:
3045 # call "netcat 10.10.1.1 4321" on every TCP connection to 10.0.2.100:1234
3046 # and connect the TCP stream to its stdin/stdout
3047 |qemu_system| -nic 'user,id=n1,guestfwd=tcp:10.0.2.100:1234-cmd:netcat 10.10.1.1 4321'
3049 ``-netdev tap,id=id[,fd=h][,ifname=name][,script=file][,downscript=dfile][,br=bridge][,helper=helper]``
3050 Configure a host TAP network backend with ID id.
3052 Use the network script file to configure it and the network script
3053 dfile to deconfigure it. If name is not provided, the OS
3054 automatically provides one. The default network configure script is
3055 ``/etc/qemu-ifup`` and the default network deconfigure script is
3056 ``/etc/qemu-ifdown``. Use ``script=no`` or ``downscript=no`` to
3057 disable script execution.
3059 If running QEMU as an unprivileged user, use the network helper
3060 to configure the TAP interface and attach it to the bridge.
3061 The default network helper executable is
3062 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3065 ``fd``\ =h can be used to specify the handle of an already opened
3072 #launch a QEMU instance with the default network script
3073 |qemu_system| linux.img -nic tap
3077 #launch a QEMU instance with two NICs, each one connected
3079 |qemu_system| linux.img \\
3080 -netdev tap,id=nd0,ifname=tap0 -device e1000,netdev=nd0 \\
3081 -netdev tap,id=nd1,ifname=tap1 -device rtl8139,netdev=nd1
3085 #launch a QEMU instance with the default network helper to
3086 #connect a TAP device to bridge br0
3087 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3088 -netdev tap,id=n1,"helper=/path/to/qemu-bridge-helper"
3090 ``-netdev bridge,id=id[,br=bridge][,helper=helper]``
3091 Connect a host TAP network interface to a host bridge device.
3093 Use the network helper helper to configure the TAP interface and
3094 attach it to the bridge. The default network helper executable is
3095 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3102 #launch a QEMU instance with the default network helper to
3103 #connect a TAP device to bridge br0
3104 |qemu_system| linux.img -netdev bridge,id=n1 -device virtio-net,netdev=n1
3108 #launch a QEMU instance with the default network helper to
3109 #connect a TAP device to bridge qemubr0
3110 |qemu_system| linux.img -netdev bridge,br=qemubr0,id=n1 -device virtio-net,netdev=n1
3112 ``-netdev socket,id=id[,fd=h][,listen=[host]:port][,connect=host:port]``
3113 This host network backend can be used to connect the guest's network
3114 to another QEMU virtual machine using a TCP socket connection. If
3115 ``listen`` is specified, QEMU waits for incoming connections on port
3116 (host is optional). ``connect`` is used to connect to another QEMU
3117 instance using the ``listen`` option. ``fd``\ =h specifies an
3118 already opened TCP socket.
3124 # launch a first QEMU instance
3125 |qemu_system| linux.img \\
3126 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3127 -netdev socket,id=n1,listen=:1234
3128 # connect the network of this instance to the network of the first instance
3129 |qemu_system| linux.img \\
3130 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3131 -netdev socket,id=n2,connect=127.0.0.1:1234
3133 ``-netdev socket,id=id[,fd=h][,mcast=maddr:port[,localaddr=addr]]``
3134 Configure a socket host network backend to share the guest's network
3135 traffic with another QEMU virtual machines using a UDP multicast
3136 socket, effectively making a bus for every QEMU with same multicast
3137 address maddr and port. NOTES:
3139 1. Several QEMU can be running on different hosts and share same bus
3140 (assuming correct multicast setup for these hosts).
3142 2. mcast support is compatible with User Mode Linux (argument
3143 ``ethN=mcast``), see http://user-mode-linux.sf.net.
3145 3. Use ``fd=h`` to specify an already opened UDP multicast socket.
3151 # launch one QEMU instance
3152 |qemu_system| linux.img \\
3153 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3154 -netdev socket,id=n1,mcast=230.0.0.1:1234
3155 # launch another QEMU instance on same "bus"
3156 |qemu_system| linux.img \\
3157 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3158 -netdev socket,id=n2,mcast=230.0.0.1:1234
3159 # launch yet another QEMU instance on same "bus"
3160 |qemu_system| linux.img \\
3161 -device e1000,netdev=n3,mac=52:54:00:12:34:58 \\
3162 -netdev socket,id=n3,mcast=230.0.0.1:1234
3164 Example (User Mode Linux compat.):
3168 # launch QEMU instance (note mcast address selected is UML's default)
3169 |qemu_system| linux.img \\
3170 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3171 -netdev socket,id=n1,mcast=239.192.168.1:1102
3173 /path/to/linux ubd0=/path/to/root_fs eth0=mcast
3175 Example (send packets from host's 1.2.3.4):
3179 |qemu_system| linux.img \\
3180 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3181 -netdev socket,id=n1,mcast=239.192.168.1:1102,localaddr=1.2.3.4
3183 ``-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]``
3184 Configure a L2TPv3 pseudowire host network backend. L2TPv3 (RFC3931)
3185 is a popular protocol to transport Ethernet (and other Layer 2) data
3186 frames between two systems. It is present in routers, firewalls and
3187 the Linux kernel (from version 3.3 onwards).
3189 This transport allows a VM to communicate to another VM, router or
3193 source address (mandatory)
3196 destination address (mandatory)
3199 select udp encapsulation (default is ip).
3205 destination udp port.
3208 force v6, otherwise defaults to v4.
3210 ``rxcookie=rxcookie``; \ ``txcookie=txcookie``
3211 Cookies are a weak form of security in the l2tpv3 specification.
3212 Their function is mostly to prevent misconfiguration. By default
3216 Set cookie size to 64 bit instead of the default 32
3219 Force a 'cut-down' L2TPv3 with no counter as in
3220 draft-mkonstan-l2tpext-keyed-ipv6-tunnel-00
3223 Work around broken counter handling in peer. This may also help
3224 on networks which have packet reorder.
3227 Add an extra offset between header and data
3229 For example, to attach a VM running on host 4.3.2.1 via L2TPv3 to
3230 the bridge br-lan on the remote Linux host 1.2.3.4:
3234 # Setup tunnel on linux host using raw ip as encapsulation
3236 ip l2tp add tunnel remote 4.3.2.1 local 1.2.3.4 tunnel_id 1 peer_tunnel_id 1 \\
3237 encap udp udp_sport 16384 udp_dport 16384
3238 ip l2tp add session tunnel_id 1 name vmtunnel0 session_id \\
3239 0xFFFFFFFF peer_session_id 0xFFFFFFFF
3240 ifconfig vmtunnel0 mtu 1500
3241 ifconfig vmtunnel0 up
3242 brctl addif br-lan vmtunnel0
3246 # launch QEMU instance - if your network has reorder or is very lossy add ,pincounter
3248 |qemu_system| linux.img -device e1000,netdev=n1 \\
3249 -netdev l2tpv3,id=n1,src=4.2.3.1,dst=1.2.3.4,udp,srcport=16384,dstport=16384,rxsession=0xffffffff,txsession=0xffffffff,counter
3251 ``-netdev vde,id=id[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]``
3252 Configure VDE backend to connect to PORT n of a vde switch running
3253 on host and listening for incoming connections on socketpath. Use
3254 GROUP groupname and MODE octalmode to change default ownership and
3255 permissions for communication port. This option is only available if
3256 QEMU has been compiled with vde support enabled.
3263 vde_switch -F -sock /tmp/myswitch
3264 # launch QEMU instance
3265 |qemu_system| linux.img -nic vde,sock=/tmp/myswitch
3267 ``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]``
3268 Establish a vhost-user netdev, backed by a chardev id. The chardev
3269 should be a unix domain socket backed one. The vhost-user uses a
3270 specifically defined protocol to pass vhost ioctl replacement
3271 messages to an application on the other end of the socket. On
3272 non-MSIX guests, the feature can be forced with vhostforce. Use
3273 'queues=n' to specify the number of queues to be created for
3274 multiqueue vhost-user.
3280 qemu -m 512 -object memory-backend-file,id=mem,size=512M,mem-path=/hugetlbfs,share=on \
3281 -numa node,memdev=mem \
3282 -chardev socket,id=chr0,path=/path/to/socket \
3283 -netdev type=vhost-user,id=net0,chardev=chr0 \
3284 -device virtio-net-pci,netdev=net0
3286 ``-netdev vhost-vdpa,vhostdev=/path/to/dev``
3287 Establish a vhost-vdpa netdev.
3289 vDPA device is a device that uses a datapath which complies with
3290 the virtio specifications with a vendor specific control path.
3291 vDPA devices can be both physically located on the hardware or
3292 emulated by software.
3294 ``-netdev hubport,id=id,hubid=hubid[,netdev=nd]``
3295 Create a hub port on the emulated hub with ID hubid.
3297 The hubport netdev lets you connect a NIC to a QEMU emulated hub
3298 instead of a single netdev. Alternatively, you can also connect the
3299 hubport to another netdev with ID nd by using the ``netdev=nd``
3302 ``-net nic[,netdev=nd][,macaddr=mac][,model=type] [,name=name][,addr=addr][,vectors=v]``
3303 Legacy option to configure or create an on-board (or machine
3304 default) Network Interface Card(NIC) and connect it either to the
3305 emulated hub with ID 0 (i.e. the default hub), or to the netdev nd.
3306 If model is omitted, then the default NIC model associated with the
3307 machine type is used. Note that the default NIC model may change in
3308 future QEMU releases, so it is highly recommended to always specify
3309 a model. Optionally, the MAC address can be changed to mac, the
3310 device address set to addr (PCI cards only), and a name can be
3311 assigned for use in monitor commands. Optionally, for PCI cards, you
3312 can specify the number v of MSI-X vectors that the card should have;
3313 this option currently only affects virtio cards; set v = 0 to
3314 disable MSI-X. If no ``-net`` option is specified, a single NIC is
3315 created. QEMU can emulate several different models of network card.
3316 Use ``-net nic,model=help`` for a list of available devices for your
3319 ``-net user|tap|bridge|socket|l2tpv3|vde[,...][,name=name]``
3320 Configure a host network backend (with the options corresponding to
3321 the same ``-netdev`` option) and connect it to the emulated hub 0
3322 (the default hub). Use name to specify the name of the hub port.
3327 DEFHEADING(Character device options:)
3329 DEF("chardev", HAS_ARG, QEMU_OPTION_chardev,
3331 "-chardev null,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3332 "-chardev socket,id=id[,host=host],port=port[,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]\n"
3333 " [,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds][,mux=on|off]\n"
3334 " [,logfile=PATH][,logappend=on|off][,tls-creds=ID][,tls-authz=ID] (tcp)\n"
3335 "-chardev socket,id=id,path=path[,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds]\n"
3336 " [,mux=on|off][,logfile=PATH][,logappend=on|off][,abstract=on|off][,tight=on|off] (unix)\n"
3337 "-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr]\n"
3338 " [,localport=localport][,ipv4=on|off][,ipv6=on|off][,mux=on|off]\n"
3339 " [,logfile=PATH][,logappend=on|off]\n"
3340 "-chardev msmouse,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3341 "-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]\n"
3342 " [,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3343 "-chardev ringbuf,id=id[,size=size][,logfile=PATH][,logappend=on|off]\n"
3344 "-chardev file,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3345 "-chardev pipe,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3347 "-chardev console,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3348 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3350 "-chardev pty,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3351 "-chardev stdio,id=id[,mux=on|off][,signal=on|off][,logfile=PATH][,logappend=on|off]\n"
3353 #ifdef CONFIG_BRLAPI
3354 "-chardev braille,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3356 #if defined(__linux__) || defined(__sun__) || defined(__FreeBSD__) \
3357 || defined(__NetBSD__) || defined(__OpenBSD__) || defined(__DragonFly__)
3358 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3359 "-chardev tty,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3361 #if defined(__linux__) || defined(__FreeBSD__) || defined(__DragonFly__)
3362 "-chardev parallel,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3363 "-chardev parport,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3365 #if defined(CONFIG_SPICE)
3366 "-chardev spicevmc,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n"
3367 "-chardev spiceport,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n"
3373 The general form of a character device option is:
3375 ``-chardev backend,id=id[,mux=on|off][,options]``
3376 Backend is one of: ``null``, ``socket``, ``udp``, ``msmouse``,
3377 ``vc``, ``ringbuf``, ``file``, ``pipe``, ``console``, ``serial``,
3378 ``pty``, ``stdio``, ``braille``, ``tty``, ``parallel``, ``parport``,
3379 ``spicevmc``, ``spiceport``. The specific backend will determine the
3382 Use ``-chardev help`` to print all available chardev backend types.
3384 All devices must have an id, which can be any string up to 127
3385 characters long. It is used to uniquely identify this device in
3386 other command line directives.
3388 A character device may be used in multiplexing mode by multiple
3389 front-ends. Specify ``mux=on`` to enable this mode. A multiplexer is
3390 a "1:N" device, and here the "1" end is your specified chardev
3391 backend, and the "N" end is the various parts of QEMU that can talk
3392 to a chardev. If you create a chardev with ``id=myid`` and
3393 ``mux=on``, QEMU will create a multiplexer with your specified ID,
3394 and you can then configure multiple front ends to use that chardev
3395 ID for their input/output. Up to four different front ends can be
3396 connected to a single multiplexed chardev. (Without multiplexing
3397 enabled, a chardev can only be used by a single front end.) For
3398 instance you could use this to allow a single stdio chardev to be
3399 used by two serial ports and the QEMU monitor:
3403 -chardev stdio,mux=on,id=char0 \
3404 -mon chardev=char0,mode=readline \
3405 -serial chardev:char0 \
3406 -serial chardev:char0
3408 You can have more than one multiplexer in a system configuration;
3409 for instance you could have a TCP port multiplexed between UART 0
3410 and UART 1, and stdio multiplexed between the QEMU monitor and a
3415 -chardev stdio,mux=on,id=char0 \
3416 -mon chardev=char0,mode=readline \
3417 -parallel chardev:char0 \
3418 -chardev tcp,...,mux=on,id=char1 \
3419 -serial chardev:char1 \
3420 -serial chardev:char1
3422 When you're using a multiplexed character device, some escape
3423 sequences are interpreted in the input. See the chapter about
3424 :ref:`keys in the character backend multiplexer` in the
3425 System Emulation Users Guide for more details.
3427 Note that some other command line options may implicitly create
3428 multiplexed character backends; for instance ``-serial mon:stdio``
3429 creates a multiplexed stdio backend connected to the serial port and
3430 the QEMU monitor, and ``-nographic`` also multiplexes the console
3431 and the monitor to stdio.
3433 There is currently no support for multiplexing in the other
3434 direction (where a single QEMU front end takes input and output from
3437 Every backend supports the ``logfile`` option, which supplies the
3438 path to a file to record all data transmitted via the backend. The
3439 ``logappend`` option controls whether the log file will be truncated
3440 or appended to when opened.
3442 The available backends are:
3444 ``-chardev null,id=id``
3445 A void device. This device will not emit any data, and will drop any
3446 data it receives. The null backend does not take any options.
3448 ``-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]``
3449 Create a two-way stream socket, which can be either a TCP or a unix
3450 socket. A unix socket will be created if ``path`` is specified.
3451 Behaviour is undefined if TCP options are specified for a unix
3454 ``server=on|off`` specifies that the socket shall be a listening socket.
3456 ``wait=on|off`` specifies that QEMU should not block waiting for a client
3457 to connect to a listening socket.
3459 ``telnet=on|off`` specifies that traffic on the socket should interpret
3460 telnet escape sequences.
3462 ``websocket=on|off`` specifies that the socket uses WebSocket protocol for
3465 ``reconnect`` sets the timeout for reconnecting on non-server
3466 sockets when the remote end goes away. qemu will delay this many
3467 seconds and then attempt to reconnect. Zero disables reconnecting,
3470 ``tls-creds`` requests enablement of the TLS protocol for
3471 encryption, and specifies the id of the TLS credentials to use for
3472 the handshake. The credentials must be previously created with the
3473 ``-object tls-creds`` argument.
3475 ``tls-auth`` provides the ID of the QAuthZ authorization object
3476 against which the client's x509 distinguished name will be
3477 validated. This object is only resolved at time of use, so can be
3478 deleted and recreated on the fly while the chardev server is active.
3479 If missing, it will default to denying access.
3481 TCP and unix socket options are given below:
3483 ``TCP options: port=port[,host=host][,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]``
3484 ``host`` for a listening socket specifies the local address to
3485 be bound. For a connecting socket species the remote host to
3486 connect to. ``host`` is optional for listening sockets. If not
3487 specified it defaults to ``0.0.0.0``.
3489 ``port`` for a listening socket specifies the local port to be
3490 bound. For a connecting socket specifies the port on the remote
3491 host to connect to. ``port`` can be given as either a port
3492 number or a service name. ``port`` is required.
3494 ``to`` is only relevant to listening sockets. If it is
3495 specified, and ``port`` cannot be bound, QEMU will attempt to
3496 bind to subsequent ports up to and including ``to`` until it
3497 succeeds. ``to`` must be specified as a port number.
3499 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4
3500 or IPv6 must be used. If neither is specified the socket may
3501 use either protocol.
3503 ``nodelay=on|off`` disables the Nagle algorithm.
3505 ``unix options: path=path[,abstract=on|off][,tight=on|off]``
3506 ``path`` specifies the local path of the unix socket. ``path``
3508 ``abstract=on|off`` specifies the use of the abstract socket namespace,
3509 rather than the filesystem. Optional, defaults to false.
3510 ``tight=on|off`` sets the socket length of abstract sockets to their minimum,
3511 rather than the full sun_path length. Optional, defaults to true.
3513 ``-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr][,localport=localport][,ipv4=on|off][,ipv6=on|off]``
3514 Sends all traffic from the guest to a remote host over UDP.
3516 ``host`` specifies the remote host to connect to. If not specified
3517 it defaults to ``localhost``.
3519 ``port`` specifies the port on the remote host to connect to.
3520 ``port`` is required.
3522 ``localaddr`` specifies the local address to bind to. If not
3523 specified it defaults to ``0.0.0.0``.
3525 ``localport`` specifies the local port to bind to. If not specified
3526 any available local port will be used.
3528 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4 or IPv6 must be used.
3529 If neither is specified the device may use either protocol.
3531 ``-chardev msmouse,id=id``
3532 Forward QEMU's emulated msmouse events to the guest. ``msmouse``
3533 does not take any options.
3535 ``-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]``
3536 Connect to a QEMU text console. ``vc`` may optionally be given a
3539 ``width`` and ``height`` specify the width and height respectively
3540 of the console, in pixels.
3542 ``cols`` and ``rows`` specify that the console be sized to fit a
3543 text console with the given dimensions.
3545 ``-chardev ringbuf,id=id[,size=size]``
3546 Create a ring buffer with fixed size ``size``. size must be a power
3547 of two and defaults to ``64K``.
3549 ``-chardev file,id=id,path=path``
3550 Log all traffic received from the guest to a file.
3552 ``path`` specifies the path of the file to be opened. This file will
3553 be created if it does not already exist, and overwritten if it does.
3554 ``path`` is required.
3556 ``-chardev pipe,id=id,path=path``
3557 Create a two-way connection to the guest. The behaviour differs
3558 slightly between Windows hosts and other hosts:
3560 On Windows, a single duplex pipe will be created at
3563 On other hosts, 2 pipes will be created called ``path.in`` and
3564 ``path.out``. Data written to ``path.in`` will be received by the
3565 guest. Data written by the guest can be read from ``path.out``. QEMU
3566 will not create these fifos, and requires them to be present.
3568 ``path`` forms part of the pipe path as described above. ``path`` is
3571 ``-chardev console,id=id``
3572 Send traffic from the guest to QEMU's standard output. ``console``
3573 does not take any options.
3575 ``console`` is only available on Windows hosts.
3577 ``-chardev serial,id=id,path=path``
3578 Send traffic from the guest to a serial device on the host.
3580 On Unix hosts serial will actually accept any tty device, not only
3583 ``path`` specifies the name of the serial device to open.
3585 ``-chardev pty,id=id``
3586 Create a new pseudo-terminal on the host and connect to it. ``pty``
3587 does not take any options.
3589 ``pty`` is not available on Windows hosts.
3591 ``-chardev stdio,id=id[,signal=on|off]``
3592 Connect to standard input and standard output of the QEMU process.
3594 ``signal`` controls if signals are enabled on the terminal, that
3595 includes exiting QEMU with the key sequence Control-c. This option
3596 is enabled by default, use ``signal=off`` to disable it.
3598 ``-chardev braille,id=id``
3599 Connect to a local BrlAPI server. ``braille`` does not take any
3602 ``-chardev tty,id=id,path=path``
3603 ``tty`` is only available on Linux, Sun, FreeBSD, NetBSD, OpenBSD
3604 and DragonFlyBSD hosts. It is an alias for ``serial``.
3606 ``path`` specifies the path to the tty. ``path`` is required.
3608 ``-chardev parallel,id=id,path=path``
3610 ``-chardev parport,id=id,path=path``
3611 ``parallel`` is only available on Linux, FreeBSD and DragonFlyBSD
3614 Connect to a local parallel port.
3616 ``path`` specifies the path to the parallel port device. ``path`` is
3619 ``-chardev spicevmc,id=id,debug=debug,name=name``
3620 ``spicevmc`` is only available when spice support is built in.
3622 ``debug`` debug level for spicevmc
3624 ``name`` name of spice channel to connect to
3626 Connect to a spice virtual machine channel, such as vdiport.
3628 ``-chardev spiceport,id=id,debug=debug,name=name``
3629 ``spiceport`` is only available when spice support is built in.
3631 ``debug`` debug level for spicevmc
3633 ``name`` name of spice port to connect to
3635 Connect to a spice port, allowing a Spice client to handle the
3636 traffic identified by a name (preferably a fqdn).
3642 DEFHEADING(TPM device options:)
3644 DEF("tpmdev", HAS_ARG, QEMU_OPTION_tpmdev, \
3645 "-tpmdev passthrough,id=id[,path=path][,cancel-path=path]\n"
3646 " use path to provide path to a character device; default is /dev/tpm0\n"
3647 " use cancel-path to provide path to TPM's cancel sysfs entry; if\n"
3648 " not provided it will be searched for in /sys/class/misc/tpm?/device\n"
3649 "-tpmdev emulator,id=id,chardev=dev\n"
3650 " configure the TPM device using chardev backend\n",
3653 The general form of a TPM device option is:
3655 ``-tpmdev backend,id=id[,options]``
3656 The specific backend type will determine the applicable options. The
3657 ``-tpmdev`` option creates the TPM backend and requires a
3658 ``-device`` option that specifies the TPM frontend interface model.
3660 Use ``-tpmdev help`` to print all available TPM backend types.
3662 The available backends are:
3664 ``-tpmdev passthrough,id=id,path=path,cancel-path=cancel-path``
3665 (Linux-host only) Enable access to the host's TPM using the
3668 ``path`` specifies the path to the host's TPM device, i.e., on a
3669 Linux host this would be ``/dev/tpm0``. ``path`` is optional and by
3670 default ``/dev/tpm0`` is used.
3672 ``cancel-path`` specifies the path to the host TPM device's sysfs
3673 entry allowing for cancellation of an ongoing TPM command.
3674 ``cancel-path`` is optional and by default QEMU will search for the
3677 Some notes about using the host's TPM with the passthrough driver:
3679 The TPM device accessed by the passthrough driver must not be used
3680 by any other application on the host.
3682 Since the host's firmware (BIOS/UEFI) has already initialized the
3683 TPM, the VM's firmware (BIOS/UEFI) will not be able to initialize
3684 the TPM again and may therefore not show a TPM-specific menu that
3685 would otherwise allow the user to configure the TPM, e.g., allow the
3686 user to enable/disable or activate/deactivate the TPM. Further, if
3687 TPM ownership is released from within a VM then the host's TPM will
3688 get disabled and deactivated. To enable and activate the TPM again
3689 afterwards, the host has to be rebooted and the user is required to
3690 enter the firmware's menu to enable and activate the TPM. If the TPM
3691 is left disabled and/or deactivated most TPM commands will fail.
3693 To create a passthrough TPM use the following two options:
3697 -tpmdev passthrough,id=tpm0 -device tpm-tis,tpmdev=tpm0
3699 Note that the ``-tpmdev`` id is ``tpm0`` and is referenced by
3700 ``tpmdev=tpm0`` in the device option.
3702 ``-tpmdev emulator,id=id,chardev=dev``
3703 (Linux-host only) Enable access to a TPM emulator using Unix domain
3704 socket based chardev backend.
3706 ``chardev`` specifies the unique ID of a character device backend
3707 that provides connection to the software TPM server.
3709 To create a TPM emulator backend device with chardev socket backend:
3713 -chardev socket,id=chrtpm,path=/tmp/swtpm-sock -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0
3720 DEFHEADING(Boot Image or Kernel specific:)
3722 There are broadly 4 ways you can boot a system with QEMU.
3724 - specify a firmware and let it control finding a kernel
3725 - specify a firmware and pass a hint to the kernel to boot
3726 - direct kernel image boot
3727 - manually load files into the guest's address space
3729 The third method is useful for quickly testing kernels but as there is
3730 no firmware to pass configuration information to the kernel the
3731 hardware must either be probeable, the kernel built for the exact
3732 configuration or passed some configuration data (e.g. a DTB blob)
3733 which tells the kernel what drivers it needs. This exact details are
3734 often hardware specific.
3736 The final method is the most generic way of loading images into the
3737 guest address space and used mostly for ``bare metal`` type
3738 development where the reset vectors of the processor are taken into
3745 For x86 machines and some other architectures ``-bios`` will generally
3746 do the right thing with whatever it is given. For other machines the
3747 more strict ``-pflash`` option needs an image that is sized for the
3748 flash device for the given machine type.
3750 Please see the :ref:`system-targets-ref` section of the manual for
3751 more detailed documentation.
3755 DEF("bios", HAS_ARG, QEMU_OPTION_bios, \
3756 "-bios file set the filename for the BIOS\n", QEMU_ARCH_ALL)
3759 Set the filename for the BIOS.
3762 DEF("pflash", HAS_ARG, QEMU_OPTION_pflash,
3763 "-pflash file use 'file' as a parallel flash image\n", QEMU_ARCH_ALL)
3766 Use file as a parallel flash image.
3771 The kernel options were designed to work with Linux kernels although
3772 other things (like hypervisors) can be packaged up as a kernel
3773 executable image. The exact format of a executable image is usually
3774 architecture specific.
3776 The way in which the kernel is started (what address it is loaded at,
3777 what if any information is passed to it via CPU registers, the state
3778 of the hardware when it is started, and so on) is also architecture
3779 specific. Typically it follows the specification laid down by the
3780 Linux kernel for how kernels for that architecture must be started.
3784 DEF("kernel", HAS_ARG, QEMU_OPTION_kernel, \
3785 "-kernel bzImage use 'bzImage' as kernel image\n", QEMU_ARCH_ALL)
3788 Use bzImage as kernel image. The kernel can be either a Linux kernel
3789 or in multiboot format.
3792 DEF("append", HAS_ARG, QEMU_OPTION_append, \
3793 "-append cmdline use 'cmdline' as kernel command line\n", QEMU_ARCH_ALL)
3796 Use cmdline as kernel command line
3799 DEF("initrd", HAS_ARG, QEMU_OPTION_initrd, \
3800 "-initrd file use 'file' as initial ram disk\n", QEMU_ARCH_ALL)
3803 Use file as initial ram disk.
3805 ``-initrd "file1 arg=foo,file2"``
3806 This syntax is only available with multiboot.
3808 Use file1 and file2 as modules and pass arg=foo as parameter to the
3812 DEF("dtb", HAS_ARG, QEMU_OPTION_dtb, \
3813 "-dtb file use 'file' as device tree image\n", QEMU_ARCH_ALL)
3816 Use file as a device tree binary (dtb) image and pass it to the
3822 Finally you can also manually load images directly into the address
3823 space of the guest. This is most useful for developers who already
3824 know the layout of their guest and take care to ensure something sane
3825 will happen when the reset vector executes.
3827 The generic loader can be invoked by using the loader device:
3829 ``-device loader,addr=<addr>,data=<data>,data-len=<data-len>[,data-be=<data-be>][,cpu-num=<cpu-num>]``
3831 there is also the guest loader which operates in a similar way but
3832 tweaks the DTB so a hypervisor loaded via ``-kernel`` can find where
3835 ``-device guest-loader,addr=<addr>[,kernel=<path>,[bootargs=<arguments>]][,initrd=<path>]``
3841 DEFHEADING(Debug/Expert options:)
3843 DEF("compat", HAS_ARG, QEMU_OPTION_compat,
3844 "-compat [deprecated-input=accept|reject|crash][,deprecated-output=accept|hide]\n"
3845 " Policy for handling deprecated management interfaces\n"
3846 "-compat [unstable-input=accept|reject|crash][,unstable-output=accept|hide]\n"
3847 " Policy for handling unstable management interfaces\n",
3850 ``-compat [deprecated-input=@var{input-policy}][,deprecated-output=@var{output-policy}]``
3851 Set policy for handling deprecated management interfaces (experimental):
3853 ``deprecated-input=accept`` (default)
3854 Accept deprecated commands and arguments
3855 ``deprecated-input=reject``
3856 Reject deprecated commands and arguments
3857 ``deprecated-input=crash``
3858 Crash on deprecated commands and arguments
3859 ``deprecated-output=accept`` (default)
3860 Emit deprecated command results and events
3861 ``deprecated-output=hide``
3862 Suppress deprecated command results and events
3864 Limitation: covers only syntactic aspects of QMP.
3866 ``-compat [unstable-input=@var{input-policy}][,unstable-output=@var{output-policy}]``
3867 Set policy for handling unstable management interfaces (experimental):
3869 ``unstable-input=accept`` (default)
3870 Accept unstable commands and arguments
3871 ``unstable-input=reject``
3872 Reject unstable commands and arguments
3873 ``unstable-input=crash``
3874 Crash on unstable commands and arguments
3875 ``unstable-output=accept`` (default)
3876 Emit unstable command results and events
3877 ``unstable-output=hide``
3878 Suppress unstable command results and events
3880 Limitation: covers only syntactic aspects of QMP.
3883 DEF("fw_cfg", HAS_ARG, QEMU_OPTION_fwcfg,
3884 "-fw_cfg [name=]<name>,file=<file>\n"
3885 " add named fw_cfg entry with contents from file\n"
3886 "-fw_cfg [name=]<name>,string=<str>\n"
3887 " add named fw_cfg entry with contents from string\n",
3890 ``-fw_cfg [name=]name,file=file``
3891 Add named fw\_cfg entry with contents from file file.
3893 ``-fw_cfg [name=]name,string=str``
3894 Add named fw\_cfg entry with contents from string str.
3896 The terminating NUL character of the contents of str will not be
3897 included as part of the fw\_cfg item data. To insert contents with
3898 embedded NUL characters, you have to use the file parameter.
3900 The fw\_cfg entries are passed by QEMU through to the guest.
3906 -fw_cfg name=opt/com.mycompany/blob,file=./my_blob.bin
3908 creates an fw\_cfg entry named opt/com.mycompany/blob with contents
3909 from ./my\_blob.bin.
3912 DEF("serial", HAS_ARG, QEMU_OPTION_serial, \
3913 "-serial dev redirect the serial port to char device 'dev'\n",
3917 Redirect the virtual serial port to host character device dev. The
3918 default device is ``vc`` in graphical mode and ``stdio`` in non
3921 This option can be used several times to simulate up to 4 serial
3924 Use ``-serial none`` to disable all serial ports.
3926 Available character devices are:
3929 Virtual console. Optionally, a width and height can be given in
3936 It is also possible to specify width or height in characters:
3943 [Linux only] Pseudo TTY (a new PTY is automatically allocated)
3946 No device is allocated.
3952 Use a named character device defined with the ``-chardev``
3956 [Linux only] Use host tty, e.g. ``/dev/ttyS0``. The host serial
3957 port parameters are set according to the emulated ones.
3960 [Linux only, parallel port only] Use host parallel port N.
3961 Currently SPP and EPP parallel port features can be used.
3964 Write output to filename. No character can be read.
3967 [Unix only] standard input/output
3973 [Windows only] Use host serial port n
3975 ``udp:[remote_host]:remote_port[@[src_ip]:src_port]``
3976 This implements UDP Net Console. When remote\_host or src\_ip
3977 are not specified they default to ``0.0.0.0``. When not using a
3978 specified src\_port a random port is automatically chosen.
3980 If you just want a simple readonly console you can use
3981 ``netcat`` or ``nc``, by starting QEMU with:
3982 ``-serial udp::4555`` and nc as: ``nc -u -l -p 4555``. Any time
3983 QEMU writes something to that port it will appear in the
3986 If you plan to send characters back via netconsole or you want
3987 to stop and start QEMU a lot of times, you should have QEMU use
3988 the same source port each time by using something like ``-serial
3989 udp::4555@:4556`` to QEMU. Another approach is to use a patched
3990 version of netcat which can listen to a TCP port and send and
3991 receive characters via udp. If you have a patched version of
3992 netcat which activates telnet remote echo and single char
3993 transfer, then you can use the following options to set up a
3994 netcat redirector to allow telnet on port 5555 to access the
3998 -serial udp::4555@:4556
4001 -u -P 4555 -L 0.0.0.0:4556 -t -p 5555 -I -T
4006 ``tcp:[host]:port[,server=on|off][,wait=on|off][,nodelay=on|off][,reconnect=seconds]``
4007 The TCP Net Console has two modes of operation. It can send the
4008 serial I/O to a location or wait for a connection from a
4009 location. By default the TCP Net Console is sent to host at the
4010 port. If you use the ``server=on`` option QEMU will wait for a client
4011 socket application to connect to the port before continuing,
4012 unless the ``wait=on|off`` option was specified. The ``nodelay=on|off``
4013 option disables the Nagle buffering algorithm. The ``reconnect=on``
4014 option only applies if ``server=no`` is set, if the connection goes
4015 down it will attempt to reconnect at the given interval. If host
4016 is omitted, 0.0.0.0 is assumed. Only one TCP connection at a
4017 time is accepted. You can use ``telnet=on`` to connect to the
4018 corresponding character device.
4020 ``Example to send tcp console to 192.168.0.2 port 4444``
4021 -serial tcp:192.168.0.2:4444
4023 ``Example to listen and wait on port 4444 for connection``
4024 -serial tcp::4444,server=on
4026 ``Example to not wait and listen on ip 192.168.0.100 port 4444``
4027 -serial tcp:192.168.0.100:4444,server=on,wait=off
4029 ``telnet:host:port[,server=on|off][,wait=on|off][,nodelay=on|off]``
4030 The telnet protocol is used instead of raw tcp sockets. The
4031 options work the same as if you had specified ``-serial tcp``.
4032 The difference is that the port acts like a telnet server or
4033 client using telnet option negotiation. This will also allow you
4034 to send the MAGIC\_SYSRQ sequence if you use a telnet that
4035 supports sending the break sequence. Typically in unix telnet
4036 you do it with Control-] and then type "send break" followed by
4037 pressing the enter key.
4039 ``websocket:host:port,server=on[,wait=on|off][,nodelay=on|off]``
4040 The WebSocket protocol is used instead of raw tcp socket. The
4041 port acts as a WebSocket server. Client mode is not supported.
4043 ``unix:path[,server=on|off][,wait=on|off][,reconnect=seconds]``
4044 A unix domain socket is used instead of a tcp socket. The option
4045 works the same as if you had specified ``-serial tcp`` except
4046 the unix domain socket path is used for connections.
4049 This is a special option to allow the monitor to be multiplexed
4050 onto another serial port. The monitor is accessed with key
4051 sequence of Control-a and then pressing c. dev\_string should be
4052 any one of the serial devices specified above. An example to
4053 multiplex the monitor onto a telnet server listening on port
4056 ``-serial mon:telnet::4444,server=on,wait=off``
4058 When the monitor is multiplexed to stdio in this way, Ctrl+C
4059 will not terminate QEMU any more but will be passed to the guest
4063 Braille device. This will use BrlAPI to display the braille
4064 output on a real or fake device.
4067 Three button serial mouse. Configure the guest to use Microsoft
4071 DEF("parallel", HAS_ARG, QEMU_OPTION_parallel, \
4072 "-parallel dev redirect the parallel port to char device 'dev'\n",
4076 Redirect the virtual parallel port to host device dev (same devices
4077 as the serial port). On Linux hosts, ``/dev/parportN`` can be used
4078 to use hardware devices connected on the corresponding host parallel
4081 This option can be used several times to simulate up to 3 parallel
4084 Use ``-parallel none`` to disable all parallel ports.
4087 DEF("monitor", HAS_ARG, QEMU_OPTION_monitor, \
4088 "-monitor dev redirect the monitor to char device 'dev'\n",
4092 Redirect the monitor to host device dev (same devices as the serial
4093 port). The default device is ``vc`` in graphical mode and ``stdio``
4094 in non graphical mode. Use ``-monitor none`` to disable the default
4097 DEF("qmp", HAS_ARG, QEMU_OPTION_qmp, \
4098 "-qmp dev like -monitor but opens in 'control' mode\n",
4102 Like -monitor but opens in 'control' mode.
4104 DEF("qmp-pretty", HAS_ARG, QEMU_OPTION_qmp_pretty, \
4105 "-qmp-pretty dev like -qmp but uses pretty JSON formatting\n",
4109 Like -qmp but uses pretty JSON formatting.
4112 DEF("mon", HAS_ARG, QEMU_OPTION_mon, \
4113 "-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]\n", QEMU_ARCH_ALL)
4115 ``-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]``
4116 Setup monitor on chardev name. ``mode=control`` configures
4117 a QMP monitor (a JSON RPC-style protocol) and it is not the
4118 same as HMP, the human monitor that has a "(qemu)" prompt.
4119 ``pretty`` is only valid when ``mode=control``,
4120 turning on JSON pretty printing to ease
4121 human reading and debugging.
4124 DEF("debugcon", HAS_ARG, QEMU_OPTION_debugcon, \
4125 "-debugcon dev redirect the debug console to char device 'dev'\n",
4129 Redirect the debug console to host device dev (same devices as the
4130 serial port). The debug console is an I/O port which is typically
4131 port 0xe9; writing to that I/O port sends output to this device. The
4132 default device is ``vc`` in graphical mode and ``stdio`` in non
4136 DEF("pidfile", HAS_ARG, QEMU_OPTION_pidfile, \
4137 "-pidfile file write PID to 'file'\n", QEMU_ARCH_ALL)
4140 Store the QEMU process PID in file. It is useful if you launch QEMU
4144 DEF("singlestep", 0, QEMU_OPTION_singlestep, \
4145 "-singlestep always run in singlestep mode\n", QEMU_ARCH_ALL)
4148 Run the emulation in single step mode.
4151 DEF("preconfig", 0, QEMU_OPTION_preconfig, \
4152 "--preconfig pause QEMU before machine is initialized (experimental)\n",
4156 Pause QEMU for interactive configuration before the machine is
4157 created, which allows querying and configuring properties that will
4158 affect machine initialization. Use QMP command 'x-exit-preconfig' to
4159 exit the preconfig state and move to the next state (i.e. run guest
4160 if -S isn't used or pause the second time if -S is used). This
4161 option is experimental.
4164 DEF("S", 0, QEMU_OPTION_S, \
4165 "-S freeze CPU at startup (use 'c' to start execution)\n",
4169 Do not start CPU at startup (you must type 'c' in the monitor).
4172 DEF("overcommit", HAS_ARG, QEMU_OPTION_overcommit,
4173 "-overcommit [mem-lock=on|off][cpu-pm=on|off]\n"
4174 " run qemu with overcommit hints\n"
4175 " mem-lock=on|off controls memory lock support (default: off)\n"
4176 " cpu-pm=on|off controls cpu power management (default: off)\n",
4179 ``-overcommit mem-lock=on|off``
4181 ``-overcommit cpu-pm=on|off``
4182 Run qemu with hints about host resource overcommit. The default is
4183 to assume that host overcommits all resources.
4185 Locking qemu and guest memory can be enabled via ``mem-lock=on``
4186 (disabled by default). This works when host memory is not
4187 overcommitted and reduces the worst-case latency for guest.
4189 Guest ability to manage power state of host cpus (increasing latency
4190 for other processes on the same host cpu, but decreasing latency for
4191 guest) can be enabled via ``cpu-pm=on`` (disabled by default). This
4192 works best when host CPU is not overcommitted. When used, host
4193 estimates of CPU cycle and power utilization will be incorrect, not
4194 taking into account guest idle time.
4197 DEF("gdb", HAS_ARG, QEMU_OPTION_gdb, \
4198 "-gdb dev accept gdb connection on 'dev'. (QEMU defaults to starting\n"
4199 " the guest without waiting for gdb to connect; use -S too\n"
4200 " if you want it to not start execution.)\n",
4204 Accept a gdb connection on device dev (see the :ref:`GDB usage` chapter
4205 in the System Emulation Users Guide). Note that this option does not pause QEMU
4206 execution -- if you want QEMU to not start the guest until you
4207 connect with gdb and issue a ``continue`` command, you will need to
4208 also pass the ``-S`` option to QEMU.
4210 The most usual configuration is to listen on a local TCP socket::
4214 but you can specify other backends; UDP, pseudo TTY, or even stdio
4215 are all reasonable use cases. For example, a stdio connection
4216 allows you to start QEMU from within gdb and establish the
4217 connection via a pipe:
4221 (gdb) target remote | exec |qemu_system| -gdb stdio ...
4224 DEF("s", 0, QEMU_OPTION_s, \
4225 "-s shorthand for -gdb tcp::" DEFAULT_GDBSTUB_PORT "\n",
4229 Shorthand for -gdb tcp::1234, i.e. open a gdbserver on TCP port 1234
4230 (see the :ref:`GDB usage` chapter in the System Emulation Users Guide).
4233 DEF("d", HAS_ARG, QEMU_OPTION_d, \
4234 "-d item1,... enable logging of specified items (use '-d help' for a list of log items)\n",
4238 Enable logging of specified items. Use '-d help' for a list of log
4242 DEF("D", HAS_ARG, QEMU_OPTION_D, \
4243 "-D logfile output log to logfile (default stderr)\n",
4247 Output log in logfile instead of to stderr
4250 DEF("dfilter", HAS_ARG, QEMU_OPTION_DFILTER, \
4251 "-dfilter range,.. filter debug output to range of addresses (useful for -d cpu,exec,etc..)\n",
4254 ``-dfilter range1[,...]``
4255 Filter debug output to that relevant to a range of target addresses.
4256 The filter spec can be either start+size, start-size or start..end
4257 where start end and size are the addresses and sizes required. For
4262 -dfilter 0x8000..0x8fff,0xffffffc000080000+0x200,0xffffffc000060000-0x1000
4264 Will dump output for any code in the 0x1000 sized block starting at
4265 0x8000 and the 0x200 sized block starting at 0xffffffc000080000 and
4266 another 0x1000 sized block starting at 0xffffffc00005f000.
4269 DEF("seed", HAS_ARG, QEMU_OPTION_seed, \
4270 "-seed number seed the pseudo-random number generator\n",
4274 Force the guest to use a deterministic pseudo-random number
4275 generator, seeded with number. This does not affect crypto routines
4279 DEF("L", HAS_ARG, QEMU_OPTION_L, \
4280 "-L path set the directory for the BIOS, VGA BIOS and keymaps\n",
4284 Set the directory for the BIOS, VGA BIOS and keymaps.
4286 To list all the data directories, use ``-L help``.
4289 DEF("enable-kvm", 0, QEMU_OPTION_enable_kvm, \
4290 "-enable-kvm enable KVM full virtualization support\n",
4291 QEMU_ARCH_ARM | QEMU_ARCH_I386 | QEMU_ARCH_MIPS | QEMU_ARCH_PPC |
4292 QEMU_ARCH_RISCV | QEMU_ARCH_S390X)
4295 Enable KVM full virtualization support. This option is only
4296 available if KVM support is enabled when compiling.
4299 DEF("xen-domid", HAS_ARG, QEMU_OPTION_xen_domid,
4300 "-xen-domid id specify xen guest domain id\n",
4301 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4302 DEF("xen-attach", 0, QEMU_OPTION_xen_attach,
4303 "-xen-attach attach to existing xen domain\n"
4304 " libxl will use this when starting QEMU\n",
4305 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4306 DEF("xen-domid-restrict", 0, QEMU_OPTION_xen_domid_restrict,
4307 "-xen-domid-restrict restrict set of available xen operations\n"
4308 " to specified domain id. (Does not affect\n"
4309 " xenpv machine type).\n",
4310 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4313 Specify xen guest domain id (XEN only).
4316 Attach to existing xen domain. libxl will use this when starting
4317 QEMU (XEN only). Restrict set of available xen operations to
4318 specified domain id (XEN only).
4321 DEF("no-reboot", 0, QEMU_OPTION_no_reboot, \
4322 "-no-reboot exit instead of rebooting\n", QEMU_ARCH_ALL)
4325 Exit instead of rebooting.
4328 DEF("no-shutdown", 0, QEMU_OPTION_no_shutdown, \
4329 "-no-shutdown stop before shutdown\n", QEMU_ARCH_ALL)
4332 Don't exit QEMU on guest shutdown, but instead only stop the
4333 emulation. This allows for instance switching to monitor to commit
4334 changes to the disk image.
4337 DEF("action", HAS_ARG, QEMU_OPTION_action,
4338 "-action reboot=reset|shutdown\n"
4339 " action when guest reboots [default=reset]\n"
4340 "-action shutdown=poweroff|pause\n"
4341 " action when guest shuts down [default=poweroff]\n"
4342 "-action panic=pause|shutdown|exit-failure|none\n"
4343 " action when guest panics [default=shutdown]\n"
4344 "-action watchdog=reset|shutdown|poweroff|inject-nmi|pause|debug|none\n"
4345 " action when watchdog fires [default=reset]\n",
4348 ``-action event=action``
4349 The action parameter serves to modify QEMU's default behavior when
4350 certain guest events occur. It provides a generic method for specifying the
4351 same behaviors that are modified by the ``-no-reboot`` and ``-no-shutdown``
4356 ``-action panic=none``
4357 ``-action reboot=shutdown,shutdown=pause``
4358 ``-device i6300esb -action watchdog=pause``
4362 DEF("loadvm", HAS_ARG, QEMU_OPTION_loadvm, \
4363 "-loadvm [tag|id]\n" \
4364 " start right away with a saved state (loadvm in monitor)\n",
4368 Start right away with a saved state (``loadvm`` in monitor)
4372 DEF("daemonize", 0, QEMU_OPTION_daemonize, \
4373 "-daemonize daemonize QEMU after initializing\n", QEMU_ARCH_ALL)
4377 Daemonize the QEMU process after initialization. QEMU will not
4378 detach from standard IO until it is ready to receive connections on
4379 any of its devices. This option is a useful way for external
4380 programs to launch QEMU without having to cope with initialization
4384 DEF("option-rom", HAS_ARG, QEMU_OPTION_option_rom, \
4385 "-option-rom rom load a file, rom, into the option ROM space\n",
4388 ``-option-rom file``
4389 Load the contents of file as an option ROM. This option is useful to
4390 load things like EtherBoot.
4393 DEF("rtc", HAS_ARG, QEMU_OPTION_rtc, \
4394 "-rtc [base=utc|localtime|<datetime>][,clock=host|rt|vm][,driftfix=none|slew]\n" \
4395 " set the RTC base and clock, enable drift fix for clock ticks (x86 only)\n",
4399 ``-rtc [base=utc|localtime|datetime][,clock=host|rt|vm][,driftfix=none|slew]``
4400 Specify ``base`` as ``utc`` or ``localtime`` to let the RTC start at
4401 the current UTC or local time, respectively. ``localtime`` is
4402 required for correct date in MS-DOS or Windows. To start at a
4403 specific point in time, provide datetime in the format
4404 ``2006-06-17T16:01:21`` or ``2006-06-17``. The default base is UTC.
4406 By default the RTC is driven by the host system time. This allows
4407 using of the RTC as accurate reference clock inside the guest,
4408 specifically if the host time is smoothly following an accurate
4409 external reference clock, e.g. via NTP. If you want to isolate the
4410 guest time from the host, you can set ``clock`` to ``rt`` instead,
4411 which provides a host monotonic clock if host support it. To even
4412 prevent the RTC from progressing during suspension, you can set
4413 ``clock`` to ``vm`` (virtual clock). '\ ``clock=vm``\ ' is
4414 recommended especially in icount mode in order to preserve
4415 determinism; however, note that in icount mode the speed of the
4416 virtual clock is variable and can in general differ from the host
4419 Enable ``driftfix`` (i386 targets only) if you experience time drift
4420 problems, specifically with Windows' ACPI HAL. This option will try
4421 to figure out how many timer interrupts were not processed by the
4422 Windows guest and will re-inject them.
4425 DEF("icount", HAS_ARG, QEMU_OPTION_icount, \
4426 "-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=<filename>[,rrsnapshot=<snapshot>]]\n" \
4427 " enable virtual instruction counter with 2^N clock ticks per\n" \
4428 " instruction, enable aligning the host and virtual clocks\n" \
4429 " or disable real time cpu sleeping, and optionally enable\n" \
4430 " record-and-replay mode\n", QEMU_ARCH_ALL)
4432 ``-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=filename[,rrsnapshot=snapshot]]``
4433 Enable virtual instruction counter. The virtual cpu will execute one
4434 instruction every 2^N ns of virtual time. If ``auto`` is specified
4435 then the virtual cpu speed will be automatically adjusted to keep
4436 virtual time within a few seconds of real time.
4438 Note that while this option can give deterministic behavior, it does
4439 not provide cycle accurate emulation. Modern CPUs contain
4440 superscalar out of order cores with complex cache hierarchies. The
4441 number of instructions executed often has little or no correlation
4442 with actual performance.
4444 When the virtual cpu is sleeping, the virtual time will advance at
4445 default speed unless ``sleep=on`` is specified. With
4446 ``sleep=on``, the virtual time will jump to the next timer
4447 deadline instantly whenever the virtual cpu goes to sleep mode and
4448 will not advance if no timer is enabled. This behavior gives
4449 deterministic execution times from the guest point of view.
4450 The default if icount is enabled is ``sleep=off``.
4451 ``sleep=on`` cannot be used together with either ``shift=auto``
4454 ``align=on`` will activate the delay algorithm which will try to
4455 synchronise the host clock and the virtual clock. The goal is to
4456 have a guest running at the real frequency imposed by the shift
4457 option. Whenever the guest clock is behind the host clock and if
4458 ``align=on`` is specified then we print a message to the user to
4459 inform about the delay. Currently this option does not work when
4460 ``shift`` is ``auto``. Note: The sync algorithm will work for those
4461 shift values for which the guest clock runs ahead of the host clock.
4462 Typically this happens when the shift value is high (how high
4463 depends on the host machine). The default if icount is enabled
4466 When the ``rr`` option is specified deterministic record/replay is
4467 enabled. The ``rrfile=`` option must also be provided to
4468 specify the path to the replay log. In record mode data is written
4469 to this file, and in replay mode it is read back.
4470 If the ``rrsnapshot`` option is given then it specifies a VM snapshot
4471 name. In record mode, a new VM snapshot with the given name is created
4472 at the start of execution recording. In replay mode this option
4473 specifies the snapshot name used to load the initial VM state.
4476 DEF("watchdog-action", HAS_ARG, QEMU_OPTION_watchdog_action, \
4477 "-watchdog-action reset|shutdown|poweroff|inject-nmi|pause|debug|none\n" \
4478 " action when watchdog fires [default=reset]\n",
4481 ``-watchdog-action action``
4482 The action controls what QEMU will do when the watchdog timer
4483 expires. The default is ``reset`` (forcefully reset the guest).
4484 Other possible actions are: ``shutdown`` (attempt to gracefully
4485 shutdown the guest), ``poweroff`` (forcefully poweroff the guest),
4486 ``inject-nmi`` (inject a NMI into the guest), ``pause`` (pause the
4487 guest), ``debug`` (print a debug message and continue), or ``none``
4490 Note that the ``shutdown`` action requires that the guest responds
4491 to ACPI signals, which it may not be able to do in the sort of
4492 situations where the watchdog would have expired, and thus
4493 ``-watchdog-action shutdown`` is not recommended for production use.
4497 ``-device i6300esb -watchdog-action pause``
4501 DEF("echr", HAS_ARG, QEMU_OPTION_echr, \
4502 "-echr chr set terminal escape character instead of ctrl-a\n",
4505 ``-echr numeric_ascii_value``
4506 Change the escape character used for switching to the monitor when
4507 using monitor and serial sharing. The default is ``0x01`` when using
4508 the ``-nographic`` option. ``0x01`` is equal to pressing
4509 ``Control-a``. You can select a different character from the ascii
4510 control keys where 1 through 26 map to Control-a through Control-z.
4511 For instance you could use the either of the following to change the
4512 escape character to Control-t.
4514 ``-echr 0x14``; \ ``-echr 20``
4518 DEF("incoming", HAS_ARG, QEMU_OPTION_incoming, \
4519 "-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]\n" \
4520 "-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]\n" \
4521 "-incoming unix:socketpath\n" \
4522 " prepare for incoming migration, listen on\n" \
4523 " specified protocol and socket address\n" \
4524 "-incoming fd:fd\n" \
4525 "-incoming exec:cmdline\n" \
4526 " accept incoming migration on given file descriptor\n" \
4527 " or from given external command\n" \
4528 "-incoming defer\n" \
4529 " wait for the URI to be specified via migrate_incoming\n",
4532 ``-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]``
4534 ``-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]``
4535 Prepare for incoming migration, listen on a given tcp port.
4537 ``-incoming unix:socketpath``
4538 Prepare for incoming migration, listen on a given unix socket.
4541 Accept incoming migration from a given filedescriptor.
4543 ``-incoming exec:cmdline``
4544 Accept incoming migration as an output from specified external
4548 Wait for the URI to be specified via migrate\_incoming. The monitor
4549 can be used to change settings (such as migration parameters) prior
4550 to issuing the migrate\_incoming to allow the migration to begin.
4553 DEF("only-migratable", 0, QEMU_OPTION_only_migratable, \
4554 "-only-migratable allow only migratable devices\n", QEMU_ARCH_ALL)
4556 ``-only-migratable``
4557 Only allow migratable devices. Devices will not be allowed to enter
4558 an unmigratable state.
4561 DEF("nodefaults", 0, QEMU_OPTION_nodefaults, \
4562 "-nodefaults don't create default devices\n", QEMU_ARCH_ALL)
4565 Don't create default devices. Normally, QEMU sets the default
4566 devices like serial port, parallel port, virtual console, monitor
4567 device, VGA adapter, floppy and CD-ROM drive and others. The
4568 ``-nodefaults`` option will disable all those default devices.
4572 DEF("chroot", HAS_ARG, QEMU_OPTION_chroot, \
4573 "-chroot dir chroot to dir just before starting the VM\n",
4578 Immediately before starting guest execution, chroot to the specified
4579 directory. Especially useful in combination with -runas.
4583 DEF("runas", HAS_ARG, QEMU_OPTION_runas, \
4584 "-runas user change to user id user just before starting the VM\n" \
4585 " user can be numeric uid:gid instead\n",
4590 Immediately before starting guest execution, drop root privileges,
4591 switching to the specified user.
4594 DEF("prom-env", HAS_ARG, QEMU_OPTION_prom_env,
4595 "-prom-env variable=value\n"
4596 " set OpenBIOS nvram variables\n",
4597 QEMU_ARCH_PPC | QEMU_ARCH_SPARC)
4599 ``-prom-env variable=value``
4600 Set OpenBIOS nvram variable to given value (PPC, SPARC only).
4604 qemu-system-sparc -prom-env 'auto-boot?=false' \
4605 -prom-env 'boot-device=sd(0,2,0):d' -prom-env 'boot-args=linux single'
4609 qemu-system-ppc -prom-env 'auto-boot?=false' \
4610 -prom-env 'boot-device=hd:2,\yaboot' \
4611 -prom-env 'boot-args=conf=hd:2,\yaboot.conf'
4613 DEF("semihosting", 0, QEMU_OPTION_semihosting,
4614 "-semihosting semihosting mode\n",
4615 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4616 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4619 Enable semihosting mode (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V only).
4621 Note that this allows guest direct access to the host filesystem, so
4622 should only be used with a trusted guest OS.
4624 See the -semihosting-config option documentation for further
4625 information about the facilities this enables.
4627 DEF("semihosting-config", HAS_ARG, QEMU_OPTION_semihosting_config,
4628 "-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,userspace=on|off][,arg=str[,...]]\n" \
4629 " semihosting configuration\n",
4630 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4631 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4633 ``-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,userspace=on|off][,arg=str[,...]]``
4634 Enable and configure semihosting (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V
4637 Note that this allows guest direct access to the host filesystem, so
4638 should only be used with a trusted guest OS.
4640 On Arm this implements the standard semihosting API, version 2.0.
4642 On M68K this implements the "ColdFire GDB" interface used by
4645 Xtensa semihosting provides basic file IO calls, such as
4646 open/read/write/seek/select. Tensilica baremetal libc for ISS and
4647 linux platform "sim" use this interface.
4649 On RISC-V this implements the standard semihosting API, version 0.2.
4651 ``target=native|gdb|auto``
4652 Defines where the semihosting calls will be addressed, to QEMU
4653 (``native``) or to GDB (``gdb``). The default is ``auto``, which
4654 means ``gdb`` during debug sessions and ``native`` otherwise.
4657 Send the output to a chardev backend output for native or auto
4658 output when not in gdb
4660 ``userspace=on|off``
4661 Allows code running in guest userspace to access the semihosting
4662 interface. The default is that only privileged guest code can
4663 make semihosting calls. Note that setting ``userspace=on`` should
4664 only be used if all guest code is trusted (for example, in
4665 bare-metal test case code).
4667 ``arg=str1,arg=str2,...``
4668 Allows the user to pass input arguments, and can be used
4669 multiple times to build up a list. The old-style
4670 ``-kernel``/``-append`` method of passing a command line is
4671 still supported for backward compatibility. If both the
4672 ``--semihosting-config arg`` and the ``-kernel``/``-append`` are
4673 specified, the former is passed to semihosting as it always
4676 DEF("old-param", 0, QEMU_OPTION_old_param,
4677 "-old-param old param mode\n", QEMU_ARCH_ARM)
4680 Old param mode (ARM only).
4683 DEF("sandbox", HAS_ARG, QEMU_OPTION_sandbox, \
4684 "-sandbox on[,obsolete=allow|deny][,elevateprivileges=allow|deny|children]\n" \
4685 " [,spawn=allow|deny][,resourcecontrol=allow|deny]\n" \
4686 " Enable seccomp mode 2 system call filter (default 'off').\n" \
4687 " use 'obsolete' to allow obsolete system calls that are provided\n" \
4688 " by the kernel, but typically no longer used by modern\n" \
4689 " C library implementations.\n" \
4690 " use 'elevateprivileges' to allow or deny the QEMU process ability\n" \
4691 " to elevate privileges using set*uid|gid system calls.\n" \
4692 " The value 'children' will deny set*uid|gid system calls for\n" \
4693 " main QEMU process but will allow forks and execves to run unprivileged\n" \
4694 " use 'spawn' to avoid QEMU to spawn new threads or processes by\n" \
4695 " blocking *fork and execve\n" \
4696 " use 'resourcecontrol' to disable process affinity and schedular priority\n",
4699 ``-sandbox arg[,obsolete=string][,elevateprivileges=string][,spawn=string][,resourcecontrol=string]``
4700 Enable Seccomp mode 2 system call filter. 'on' will enable syscall
4701 filtering and 'off' will disable it. The default is 'off'.
4704 Enable Obsolete system calls
4706 ``elevateprivileges=string``
4707 Disable set\*uid\|gid system calls
4710 Disable \*fork and execve
4712 ``resourcecontrol=string``
4713 Disable process affinity and schedular priority
4716 DEF("readconfig", HAS_ARG, QEMU_OPTION_readconfig,
4717 "-readconfig <file>\n"
4718 " read config file\n", QEMU_ARCH_ALL)
4720 ``-readconfig file``
4721 Read device configuration from file. This approach is useful when
4722 you want to spawn QEMU process with many command line options but
4723 you don't want to exceed the command line character limit.
4726 DEF("no-user-config", 0, QEMU_OPTION_nouserconfig,
4728 " do not load default user-provided config files at startup\n",
4732 The ``-no-user-config`` option makes QEMU not load any of the
4733 user-provided config files on sysconfdir.
4736 DEF("trace", HAS_ARG, QEMU_OPTION_trace,
4737 "-trace [[enable=]<pattern>][,events=<file>][,file=<file>]\n"
4738 " specify tracing options\n",
4741 ``-trace [[enable=]pattern][,events=file][,file=file]``
4742 .. include:: ../qemu-option-trace.rst.inc
4745 DEF("plugin", HAS_ARG, QEMU_OPTION_plugin,
4746 "-plugin [file=]<file>[,<argname>=<argvalue>]\n"
4750 ``-plugin file=file[,argname=argvalue]``
4754 Load the given plugin from a shared library file.
4756 ``argname=argvalue``
4757 Argument passed to the plugin. (Can be given multiple times.)
4761 DEF("qtest", HAS_ARG, QEMU_OPTION_qtest, "", QEMU_ARCH_ALL)
4762 DEF("qtest-log", HAS_ARG, QEMU_OPTION_qtest_log, "", QEMU_ARCH_ALL)
4764 DEF("msg", HAS_ARG, QEMU_OPTION_msg,
4765 "-msg [timestamp[=on|off]][,guest-name=[on|off]]\n"
4766 " control error message format\n"
4767 " timestamp=on enables timestamps (default: off)\n"
4768 " guest-name=on enables guest name prefix but only if\n"
4769 " -name guest option is set (default: off)\n",
4772 ``-msg [timestamp[=on|off]][,guest-name[=on|off]]``
4773 Control error message format.
4775 ``timestamp=on|off``
4776 Prefix messages with a timestamp. Default is off.
4778 ``guest-name=on|off``
4779 Prefix messages with guest name but only if -name guest option is set
4780 otherwise the option is ignored. Default is off.
4783 DEF("dump-vmstate", HAS_ARG, QEMU_OPTION_dump_vmstate,
4784 "-dump-vmstate <file>\n"
4785 " Output vmstate information in JSON format to file.\n"
4786 " Use the scripts/vmstate-static-checker.py file to\n"
4787 " check for possible regressions in migration code\n"
4788 " by comparing two such vmstate dumps.\n",
4791 ``-dump-vmstate file``
4792 Dump json-encoded vmstate information for current machine type to
4796 DEF("enable-sync-profile", 0, QEMU_OPTION_enable_sync_profile,
4797 "-enable-sync-profile\n"
4798 " enable synchronization profiling\n",
4801 ``-enable-sync-profile``
4802 Enable synchronization profiling.
4807 DEFHEADING(Generic object creation:)
4809 DEF("object", HAS_ARG, QEMU_OPTION_object,
4810 "-object TYPENAME[,PROP1=VALUE1,...]\n"
4811 " create a new object of type TYPENAME setting properties\n"
4812 " in the order they are specified. Note that the 'id'\n"
4813 " property must be set. These objects are placed in the\n"
4814 " '/objects' path.\n",
4817 ``-object typename[,prop1=value1,...]``
4818 Create a new object of type typename setting properties in the order
4819 they are specified. Note that the 'id' property must be set. These
4820 objects are placed in the '/objects' path.
4822 ``-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``
4823 Creates a memory file backend object, which can be used to back
4824 the guest RAM with huge pages.
4826 The ``id`` parameter is a unique ID that will be used to
4827 reference this memory region in other parameters, e.g. ``-numa``,
4828 ``-device nvdimm``, etc.
4830 The ``size`` option provides the size of the memory region, and
4831 accepts common suffixes, e.g. ``500M``.
4833 The ``mem-path`` provides the path to either a shared memory or
4834 huge page filesystem mount.
4836 The ``share`` boolean option determines whether the memory
4837 region is marked as private to QEMU, or shared. The latter
4838 allows a co-operating external process to access the QEMU memory
4841 The ``share`` is also required for pvrdma devices due to
4842 limitations in the RDMA API provided by Linux.
4844 Setting share=on might affect the ability to configure NUMA
4845 bindings for the memory backend under some circumstances, see
4846 Documentation/vm/numa\_memory\_policy.txt on the Linux kernel
4847 source tree for additional details.
4849 Setting the ``discard-data`` boolean option to on indicates that
4850 file contents can be destroyed when QEMU exits, to avoid
4851 unnecessarily flushing data to the backing file. Note that
4852 ``discard-data`` is only an optimization, and QEMU might not
4853 discard file contents if it aborts unexpectedly or is terminated
4856 The ``merge`` boolean option enables memory merge, also known as
4857 MADV\_MERGEABLE, so that Kernel Samepage Merging will consider
4858 the pages for memory deduplication.
4860 Setting the ``dump`` boolean option to off excludes the memory
4861 from core dumps. This feature is also known as MADV\_DONTDUMP.
4863 The ``prealloc`` boolean option enables memory preallocation.
4865 The ``host-nodes`` option binds the memory range to a list of
4868 The ``policy`` option sets the NUMA policy to one of the
4875 prefer the given host node list for allocation
4878 restrict memory allocation to the given host node list
4881 interleave memory allocations across the given host node
4884 The ``align`` option specifies the base address alignment when
4885 QEMU mmap(2) ``mem-path``, and accepts common suffixes, eg
4886 ``2M``. Some backend store specified by ``mem-path`` requires an
4887 alignment different than the default one used by QEMU, eg the
4888 device DAX /dev/dax0.0 requires 2M alignment rather than 4K. In
4889 such cases, users can specify the required alignment via this
4892 The ``pmem`` option specifies whether the backing file specified
4893 by ``mem-path`` is in host persistent memory that can be
4894 accessed using the SNIA NVM programming model (e.g. Intel
4895 NVDIMM). If ``pmem`` is set to 'on', QEMU will take necessary
4896 operations to guarantee the persistence of its own writes to
4897 ``mem-path`` (e.g. in vNVDIMM label emulation and live
4898 migration). Also, we will map the backend-file with MAP\_SYNC
4899 flag, which ensures the file metadata is in sync for
4900 ``mem-path`` in case of host crash or a power failure. MAP\_SYNC
4901 requires support from both the host kernel (since Linux kernel
4902 4.15) and the filesystem of ``mem-path`` mounted with DAX
4905 The ``readonly`` option specifies whether the backing file is opened
4906 read-only or read-write (default).
4908 ``-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``
4909 Creates a memory backend object, which can be used to back the
4910 guest RAM. Memory backend objects offer more control than the
4911 ``-m`` option that is traditionally used to define guest RAM.
4912 Please refer to ``memory-backend-file`` for a description of the
4915 ``-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``
4916 Creates an anonymous memory file backend object, which allows
4917 QEMU to share the memory with an external process (e.g. when
4918 using vhost-user). The memory is allocated with memfd and
4919 optional sealing. (Linux only)
4921 The ``seal`` option creates a sealed-file, that will block
4922 further resizing the memory ('on' by default).
4924 The ``hugetlb`` option specify the file to be created resides in
4925 the hugetlbfs filesystem (since Linux 4.14). Used in conjunction
4926 with the ``hugetlb`` option, the ``hugetlbsize`` option specify
4927 the hugetlb page size on systems that support multiple hugetlb
4928 page sizes (it must be a power of 2 value supported by the
4931 In some versions of Linux, the ``hugetlb`` option is
4932 incompatible with the ``seal`` option (requires at least Linux
4935 Please refer to ``memory-backend-file`` for a description of the
4938 The ``share`` boolean option is on by default with memfd.
4940 ``-object rng-builtin,id=id``
4941 Creates a random number generator backend which obtains entropy
4942 from QEMU builtin functions. The ``id`` parameter is a unique ID
4943 that will be used to reference this entropy backend from the
4944 ``virtio-rng`` device. By default, the ``virtio-rng`` device
4945 uses this RNG backend.
4947 ``-object rng-random,id=id,filename=/dev/random``
4948 Creates a random number generator backend which obtains entropy
4949 from a device on the host. The ``id`` parameter is a unique ID
4950 that will be used to reference this entropy backend from the
4951 ``virtio-rng`` device. The ``filename`` parameter specifies
4952 which file to obtain entropy from and if omitted defaults to
4955 ``-object rng-egd,id=id,chardev=chardevid``
4956 Creates a random number generator backend which obtains entropy
4957 from an external daemon running on the host. The ``id``
4958 parameter is a unique ID that will be used to reference this
4959 entropy backend from the ``virtio-rng`` device. The ``chardev``
4960 parameter is the unique ID of a character device backend that
4961 provides the connection to the RNG daemon.
4963 ``-object tls-creds-anon,id=id,endpoint=endpoint,dir=/path/to/cred/dir,verify-peer=on|off``
4964 Creates a TLS anonymous credentials object, which can be used to
4965 provide TLS support on network backends. The ``id`` parameter is
4966 a unique ID which network backends will use to access the
4967 credentials. The ``endpoint`` is either ``server`` or ``client``
4968 depending on whether the QEMU network backend that uses the
4969 credentials will be acting as a client or as a server. If
4970 ``verify-peer`` is enabled (the default) then once the handshake
4971 is completed, the peer credentials will be verified, though this
4972 is a no-op for anonymous credentials.
4974 The dir parameter tells QEMU where to find the credential files.
4975 For server endpoints, this directory may contain a file
4976 dh-params.pem providing diffie-hellman parameters to use for the
4977 TLS server. If the file is missing, QEMU will generate a set of
4978 DH parameters at startup. This is a computationally expensive
4979 operation that consumes random pool entropy, so it is
4980 recommended that a persistent set of parameters be generated
4983 ``-object tls-creds-psk,id=id,endpoint=endpoint,dir=/path/to/keys/dir[,username=username]``
4984 Creates a TLS Pre-Shared Keys (PSK) credentials object, which
4985 can be used to provide TLS support on network backends. The
4986 ``id`` parameter is a unique ID which network backends will use
4987 to access the credentials. The ``endpoint`` is either ``server``
4988 or ``client`` depending on whether the QEMU network backend that
4989 uses the credentials will be acting as a client or as a server.
4990 For clients only, ``username`` is the username which will be
4991 sent to the server. If omitted it defaults to "qemu".
4993 The dir parameter tells QEMU where to find the keys file. It is
4994 called "dir/keys.psk" and contains "username:key" pairs. This
4995 file can most easily be created using the GnuTLS ``psktool``
4998 For server endpoints, dir may also contain a file dh-params.pem
4999 providing diffie-hellman parameters to use for the TLS server.
5000 If the file is missing, QEMU will generate a set of DH
5001 parameters at startup. This is a computationally expensive
5002 operation that consumes random pool entropy, so it is
5003 recommended that a persistent set of parameters be generated up
5006 ``-object tls-creds-x509,id=id,endpoint=endpoint,dir=/path/to/cred/dir,priority=priority,verify-peer=on|off,passwordid=id``
5007 Creates a TLS anonymous credentials object, which can be used to
5008 provide TLS support on network backends. The ``id`` parameter is
5009 a unique ID which network backends will use to access the
5010 credentials. The ``endpoint`` is either ``server`` or ``client``
5011 depending on whether the QEMU network backend that uses the
5012 credentials will be acting as a client or as a server. If
5013 ``verify-peer`` is enabled (the default) then once the handshake
5014 is completed, the peer credentials will be verified. With x509
5015 certificates, this implies that the clients must be provided
5016 with valid client certificates too.
5018 The dir parameter tells QEMU where to find the credential files.
5019 For server endpoints, this directory may contain a file
5020 dh-params.pem providing diffie-hellman parameters to use for the
5021 TLS server. If the file is missing, QEMU will generate a set of
5022 DH parameters at startup. This is a computationally expensive
5023 operation that consumes random pool entropy, so it is
5024 recommended that a persistent set of parameters be generated
5027 For x509 certificate credentials the directory will contain
5028 further files providing the x509 certificates. The certificates
5029 must be stored in PEM format, in filenames ca-cert.pem,
5030 ca-crl.pem (optional), server-cert.pem (only servers),
5031 server-key.pem (only servers), client-cert.pem (only clients),
5032 and client-key.pem (only clients).
5034 For the server-key.pem and client-key.pem files which contain
5035 sensitive private keys, it is possible to use an encrypted
5036 version by providing the passwordid parameter. This provides the
5037 ID of a previously created ``secret`` object containing the
5038 password for decryption.
5040 The priority parameter allows to override the global default
5041 priority used by gnutls. This can be useful if the system
5042 administrator needs to use a weaker set of crypto priorities for
5043 QEMU without potentially forcing the weakness onto all
5044 applications. Or conversely if one wants wants a stronger
5045 default for QEMU than for all other applications, they can do
5046 this through this parameter. Its format is a gnutls priority
5047 string as described at
5048 https://gnutls.org/manual/html_node/Priority-Strings.html.
5050 ``-object tls-cipher-suites,id=id,priority=priority``
5051 Creates a TLS cipher suites object, which can be used to control
5052 the TLS cipher/protocol algorithms that applications are permitted
5055 The ``id`` parameter is a unique ID which frontends will use to
5056 access the ordered list of permitted TLS cipher suites from the
5059 The ``priority`` parameter allows to override the global default
5060 priority used by gnutls. This can be useful if the system
5061 administrator needs to use a weaker set of crypto priorities for
5062 QEMU without potentially forcing the weakness onto all
5063 applications. Or conversely if one wants wants a stronger
5064 default for QEMU than for all other applications, they can do
5065 this through this parameter. Its format is a gnutls priority
5066 string as described at
5067 https://gnutls.org/manual/html_node/Priority-Strings.html.
5069 An example of use of this object is to control UEFI HTTPS Boot.
5070 The tls-cipher-suites object exposes the ordered list of permitted
5071 TLS cipher suites from the host side to the guest firmware, via
5072 fw_cfg. The list is represented as an array of IANA_TLS_CIPHER
5073 objects. The firmware uses the IANA_TLS_CIPHER array for configuring
5076 In the following example, the priority at which the host-side policy
5077 is retrieved is given by the ``priority`` property.
5078 Given that QEMU uses GNUTLS, ``priority=@SYSTEM`` may be used to
5079 refer to /etc/crypto-policies/back-ends/gnutls.config.
5084 -object tls-cipher-suites,id=mysuite0,priority=@SYSTEM \\
5085 -fw_cfg name=etc/edk2/https/ciphers,gen_id=mysuite0
5087 ``-object filter-buffer,id=id,netdev=netdevid,interval=t[,queue=all|rx|tx][,status=on|off][,position=head|tail|id=<id>][,insert=behind|before]``
5088 Interval t can't be 0, this filter batches the packet delivery:
5089 all packets arriving in a given interval on netdev netdevid are
5090 delayed until the end of the interval. Interval is in
5091 microseconds. ``status`` is optional that indicate whether the
5092 netfilter is on (enabled) or off (disabled), the default status
5093 for netfilter will be 'on'.
5095 queue all\|rx\|tx is an option that can be applied to any
5098 ``all``: the filter is attached both to the receive and the
5099 transmit queue of the netdev (default).
5101 ``rx``: the filter is attached to the receive queue of the
5102 netdev, where it will receive packets sent to the netdev.
5104 ``tx``: the filter is attached to the transmit queue of the
5105 netdev, where it will receive packets sent by the netdev.
5107 position head\|tail\|id=<id> is an option to specify where the
5108 filter should be inserted in the filter list. It can be applied
5111 ``head``: the filter is inserted at the head of the filter list,
5112 before any existing filters.
5114 ``tail``: the filter is inserted at the tail of the filter list,
5115 behind any existing filters (default).
5117 ``id=<id>``: the filter is inserted before or behind the filter
5118 specified by <id>, see the insert option below.
5120 insert behind\|before is an option to specify where to insert
5121 the new filter relative to the one specified with
5122 position=id=<id>. It can be applied to any netfilter.
5124 ``before``: insert before the specified filter.
5126 ``behind``: insert behind the specified filter (default).
5128 ``-object filter-mirror,id=id,netdev=netdevid,outdev=chardevid,queue=all|rx|tx[,vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5129 filter-mirror on netdev netdevid,mirror net packet to
5130 chardevchardevid, if it has the vnet\_hdr\_support flag,
5131 filter-mirror will mirror packet with vnet\_hdr\_len.
5133 ``-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]``
5134 filter-redirector on netdev netdevid,redirect filter's net
5135 packet to chardev chardevid,and redirect indev's packet to
5136 filter.if it has the vnet\_hdr\_support flag, filter-redirector
5137 will redirect packet with vnet\_hdr\_len. Create a
5138 filter-redirector we need to differ outdev id from indev id, id
5139 can not be the same. we can just use indev or outdev, but at
5140 least one of indev or outdev need to be specified.
5142 ``-object filter-rewriter,id=id,netdev=netdevid,queue=all|rx|tx,[vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5143 Filter-rewriter is a part of COLO project.It will rewrite tcp
5144 packet to secondary from primary to keep secondary tcp
5145 connection,and rewrite tcp packet to primary from secondary make
5146 tcp packet can be handled by client.if it has the
5147 vnet\_hdr\_support flag, we can parse packet with vnet header.
5149 usage: colo secondary: -object
5150 filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 -object
5151 filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 -object
5152 filter-rewriter,id=rew0,netdev=hn0,queue=all
5154 ``-object filter-dump,id=id,netdev=dev[,file=filename][,maxlen=len][,position=head|tail|id=<id>][,insert=behind|before]``
5155 Dump the network traffic on netdev dev to the file specified by
5156 filename. At most len bytes (64k by default) per packet are
5157 stored. The file format is libpcap, so it can be analyzed with
5158 tools such as tcpdump or Wireshark.
5160 ``-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}]``
5161 Colo-compare gets packet from primary\_in chardevid and
5162 secondary\_in, then compare whether the payload of primary packet
5163 and secondary packet are the same. If same, it will output
5164 primary packet to out\_dev, else it will notify COLO-framework to do
5165 checkpoint and send primary packet to out\_dev. In order to
5166 improve efficiency, we need to put the task of comparison in
5167 another iothread. If it has the vnet\_hdr\_support flag,
5168 colo compare will send/recv packet with vnet\_hdr\_len.
5169 The compare\_timeout=@var{ms} determines the maximum time of the
5170 colo-compare hold the packet. The expired\_scan\_cycle=@var{ms}
5171 is to set the period of scanning expired primary node network packets.
5172 The max\_queue\_size=@var{size} is to set the max compare queue
5173 size depend on user environment.
5174 If user want to use Xen COLO, need to add the notify\_dev to
5175 notify Xen colo-frame to do checkpoint.
5177 COLO-compare must be used with the help of filter-mirror,
5178 filter-redirector and filter-rewriter.
5185 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5186 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5187 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5188 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5189 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5190 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5191 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5192 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5193 -object iothread,id=iothread1
5194 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5195 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5196 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5197 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,iothread=iothread1
5200 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5201 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5202 -chardev socket,id=red0,host=3.3.3.3,port=9003
5203 -chardev socket,id=red1,host=3.3.3.3,port=9004
5204 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5205 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5211 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5212 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5213 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5214 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5215 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5216 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5217 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5218 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5219 -chardev socket,id=notify_way,host=3.3.3.3,port=9009,server=on,wait=off
5220 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5221 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5222 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5223 -object iothread,id=iothread1
5224 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,notify_dev=nofity_way,iothread=iothread1
5227 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5228 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5229 -chardev socket,id=red0,host=3.3.3.3,port=9003
5230 -chardev socket,id=red1,host=3.3.3.3,port=9004
5231 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5232 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5234 If you want to know the detail of above command line, you can
5235 read the colo-compare git log.
5237 ``-object cryptodev-backend-builtin,id=id[,queues=queues]``
5238 Creates a cryptodev backend which executes crypto opreation from
5239 the QEMU cipher APIS. The id parameter is a unique ID that will
5240 be used to reference this cryptodev backend from the
5241 ``virtio-crypto`` device. The queues parameter is optional,
5242 which specify the queue number of cryptodev backend, the default
5249 -object cryptodev-backend-builtin,id=cryptodev0 \\
5250 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5253 ``-object cryptodev-vhost-user,id=id,chardev=chardevid[,queues=queues]``
5254 Creates a vhost-user cryptodev backend, backed by a chardev
5255 chardevid. The id parameter is a unique ID that will be used to
5256 reference this cryptodev backend from the ``virtio-crypto``
5257 device. The chardev should be a unix domain socket backed one.
5258 The vhost-user uses a specifically defined protocol to pass
5259 vhost ioctl replacement messages to an application on the other
5260 end of the socket. The queues parameter is optional, which
5261 specify the queue number of cryptodev backend for multiqueue
5262 vhost-user, the default of queues is 1.
5268 -chardev socket,id=chardev0,path=/path/to/socket \\
5269 -object cryptodev-vhost-user,id=cryptodev0,chardev=chardev0 \\
5270 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5273 ``-object secret,id=id,data=string,format=raw|base64[,keyid=secretid,iv=string]``
5275 ``-object secret,id=id,file=filename,format=raw|base64[,keyid=secretid,iv=string]``
5276 Defines a secret to store a password, encryption key, or some
5277 other sensitive data. The sensitive data can either be passed
5278 directly via the data parameter, or indirectly via the file
5279 parameter. Using the data parameter is insecure unless the
5280 sensitive data is encrypted.
5282 The sensitive data can be provided in raw format (the default),
5283 or base64. When encoded as JSON, the raw format only supports
5284 valid UTF-8 characters, so base64 is recommended for sending
5285 binary data. QEMU will convert from which ever format is
5286 provided to the format it needs internally. eg, an RBD password
5287 can be provided in raw format, even though it will be base64
5288 encoded when passed onto the RBD sever.
5290 For added protection, it is possible to encrypt the data
5291 associated with a secret using the AES-256-CBC cipher. Use of
5292 encryption is indicated by providing the keyid and iv
5293 parameters. The keyid parameter provides the ID of a previously
5294 defined secret that contains the AES-256 decryption key. This
5295 key should be 32-bytes long and be base64 encoded. The iv
5296 parameter provides the random initialization vector used for
5297 encryption of this particular secret and should be a base64
5298 encrypted string of the 16-byte IV.
5300 The simplest (insecure) usage is to provide the secret inline
5304 # |qemu_system| -object secret,id=sec0,data=letmein,format=raw
5306 The simplest secure usage is to provide the secret via a file
5308 # printf "letmein" > mypasswd.txt # QEMU\_SYSTEM\_MACRO -object
5309 secret,id=sec0,file=mypasswd.txt,format=raw
5311 For greater security, AES-256-CBC should be used. To illustrate
5312 usage, consider the openssl command line tool which can encrypt
5313 the data. Note that when encrypting, the plaintext must be
5314 padded to the cipher block size (32 bytes) using the standard
5315 PKCS#5/6 compatible padding algorithm.
5317 First a master key needs to be created in base64 encoding:
5321 # openssl rand -base64 32 > key.b64
5322 # KEY=$(base64 -d key.b64 | hexdump -v -e '/1 "%02X"')
5324 Each secret to be encrypted needs to have a random
5325 initialization vector generated. These do not need to be kept
5330 # openssl rand -base64 16 > iv.b64
5331 # IV=$(base64 -d iv.b64 | hexdump -v -e '/1 "%02X"')
5333 The secret to be defined can now be encrypted, in this case
5334 we're telling openssl to base64 encode the result, but it could
5335 be left as raw bytes if desired.
5339 # SECRET=$(printf "letmein" |
5340 openssl enc -aes-256-cbc -a -K $KEY -iv $IV)
5342 When launching QEMU, create a master secret pointing to
5343 ``key.b64`` and specify that to be used to decrypt the user
5344 password. Pass the contents of ``iv.b64`` to the second secret
5349 -object secret,id=secmaster0,format=base64,file=key.b64 \\
5350 -object secret,id=sec0,keyid=secmaster0,format=base64,\\
5351 data=$SECRET,iv=$(<iv.b64)
5353 ``-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]``
5354 Create a Secure Encrypted Virtualization (SEV) guest object,
5355 which can be used to provide the guest memory encryption support
5358 When memory encryption is enabled, one of the physical address
5359 bit (aka the C-bit) is utilized to mark if a memory page is
5360 protected. The ``cbitpos`` is used to provide the C-bit
5361 position. The C-bit position is Host family dependent hence user
5362 must provide this value. On EPYC, the value should be 47.
5364 When memory encryption is enabled, we loose certain bits in
5365 physical address space. The ``reduced-phys-bits`` is used to
5366 provide the number of bits we loose in physical address space.
5367 Similar to C-bit, the value is Host family dependent. On EPYC,
5368 the value should be 5.
5370 The ``sev-device`` provides the device file to use for
5371 communicating with the SEV firmware running inside AMD Secure
5372 Processor. The default device is '/dev/sev'. If hardware
5373 supports memory encryption then /dev/sev devices are created by
5376 The ``policy`` provides the guest policy to be enforced by the
5377 SEV firmware and restrict what configuration and operational
5378 commands can be performed on this guest by the hypervisor. The
5379 policy should be provided by the guest owner and is bound to the
5380 guest and cannot be changed throughout the lifetime of the
5381 guest. The default is 0.
5383 If guest ``policy`` allows sharing the key with another SEV
5384 guest then ``handle`` can be use to provide handle of the guest
5385 from which to share the key.
5387 The ``dh-cert-file`` and ``session-file`` provides the guest
5388 owner's Public Diffie-Hillman key defined in SEV spec. The PDH
5389 and session parameters are used for establishing a cryptographic
5390 session with the guest owner to negotiate keys used for
5391 attestation. The file must be encoded in base64.
5393 The ``kernel-hashes`` adds the hashes of given kernel/initrd/
5394 cmdline to a designated guest firmware page for measured Linux
5395 boot with -kernel. The default is off. (Since 6.2)
5397 e.g to launch a SEV guest
5401 # |qemu_system_x86| \\
5403 -object sev-guest,id=sev0,cbitpos=47,reduced-phys-bits=5 \\
5404 -machine ...,memory-encryption=sev0 \\
5407 ``-object authz-simple,id=id,identity=string``
5408 Create an authorization object that will control access to
5411 The ``identity`` parameter is identifies the user and its format
5412 depends on the network service that authorization object is
5413 associated with. For authorizing based on TLS x509 certificates,
5414 the identity must be the x509 distinguished name. Note that care
5415 must be taken to escape any commas in the distinguished name.
5417 An example authorization object to validate a x509 distinguished
5418 name would look like:
5424 -object 'authz-simple,id=auth0,identity=CN=laptop.example.com,,O=Example Org,,L=London,,ST=London,,C=GB' \\
5427 Note the use of quotes due to the x509 distinguished name
5428 containing whitespace, and escaping of ','.
5430 ``-object authz-listfile,id=id,filename=path,refresh=on|off``
5431 Create an authorization object that will control access to
5434 The ``filename`` parameter is the fully qualified path to a file
5435 containing the access control list rules in JSON format.
5437 An example set of rules that match against SASL usernames might
5444 { "match": "fred", "policy": "allow", "format": "exact" },
5445 { "match": "bob", "policy": "allow", "format": "exact" },
5446 { "match": "danb", "policy": "deny", "format": "glob" },
5447 { "match": "dan*", "policy": "allow", "format": "exact" },
5452 When checking access the object will iterate over all the rules
5453 and the first rule to match will have its ``policy`` value
5454 returned as the result. If no rules match, then the default
5455 ``policy`` value is returned.
5457 The rules can either be an exact string match, or they can use
5458 the simple UNIX glob pattern matching to allow wildcards to be
5461 If ``refresh`` is set to true the file will be monitored and
5462 automatically reloaded whenever its content changes.
5464 As with the ``authz-simple`` object, the format of the identity
5465 strings being matched depends on the network service, but is
5466 usually a TLS x509 distinguished name, or a SASL username.
5468 An example authorization object to validate a SASL username
5475 -object authz-simple,id=auth0,filename=/etc/qemu/vnc-sasl.acl,refresh=on \\
5478 ``-object authz-pam,id=id,service=string``
5479 Create an authorization object that will control access to
5482 The ``service`` parameter provides the name of a PAM service to
5483 use for authorization. It requires that a file
5484 ``/etc/pam.d/service`` exist to provide the configuration for
5485 the ``account`` subsystem.
5487 An example authorization object to validate a TLS x509
5488 distinguished name would look like:
5494 -object authz-pam,id=auth0,service=qemu-vnc \\
5497 There would then be a corresponding config file for PAM at
5498 ``/etc/pam.d/qemu-vnc`` that contains:
5502 account requisite pam_listfile.so item=user sense=allow \
5503 file=/etc/qemu/vnc.allow
5505 Finally the ``/etc/qemu/vnc.allow`` file would contain the list
5506 of x509 distingished names that are permitted access
5510 CN=laptop.example.com,O=Example Home,L=London,ST=London,C=GB
5512 ``-object iothread,id=id,poll-max-ns=poll-max-ns,poll-grow=poll-grow,poll-shrink=poll-shrink,aio-max-batch=aio-max-batch``
5513 Creates a dedicated event loop thread that devices can be
5514 assigned to. This is known as an IOThread. By default device
5515 emulation happens in vCPU threads or the main event loop thread.
5516 This can become a scalability bottleneck. IOThreads allow device
5517 emulation and I/O to run on other host CPUs.
5519 The ``id`` parameter is a unique ID that will be used to
5520 reference this IOThread from ``-device ...,iothread=id``.
5521 Multiple devices can be assigned to an IOThread. Note that not
5522 all devices support an ``iothread`` parameter.
5524 The ``query-iothreads`` QMP command lists IOThreads and reports
5525 their thread IDs so that the user can configure host CPU
5528 IOThreads use an adaptive polling algorithm to reduce event loop
5529 latency. Instead of entering a blocking system call to monitor
5530 file descriptors and then pay the cost of being woken up when an
5531 event occurs, the polling algorithm spins waiting for events for
5532 a short time. The algorithm's default parameters are suitable
5533 for many cases but can be adjusted based on knowledge of the
5534 workload and/or host device latency.
5536 The ``poll-max-ns`` parameter is the maximum number of
5537 nanoseconds to busy wait for events. Polling can be disabled by
5538 setting this value to 0.
5540 The ``poll-grow`` parameter is the multiplier used to increase
5541 the polling time when the algorithm detects it is missing events
5542 due to not polling long enough.
5544 The ``poll-shrink`` parameter is the divisor used to decrease
5545 the polling time when the algorithm detects it is spending too
5546 long polling without encountering events.
5548 The ``aio-max-batch`` parameter is the maximum number of requests
5549 in a batch for the AIO engine, 0 means that the engine will use
5552 The IOThread parameters can be modified at run-time using the
5553 ``qom-set`` command (where ``iothread1`` is the IOThread's
5558 (qemu) qom-set /objects/iothread1 poll-max-ns 100000
5562 HXCOMM This is the last statement. Insert new options before this line!