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, 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",
43 ``-machine [type=]name[,prop=value[,...]]``
44 Select the emulated machine by name. Use ``-machine help`` to list
47 For architectures which aim to support live migration compatibility
48 across releases, each release will introduce a new versioned machine
49 type. For example, the 2.8.0 release introduced machine types
50 "pc-i440fx-2.8" and "pc-q35-2.8" for the x86\_64/i686 architectures.
52 To allow live migration of guests from QEMU version 2.8.0, to QEMU
53 version 2.9.0, the 2.9.0 version must support the "pc-i440fx-2.8"
54 and "pc-q35-2.8" machines too. To allow users live migrating VMs to
55 skip multiple intermediate releases when upgrading, new releases of
56 QEMU will support machine types from many previous versions.
58 Supported machine properties are:
60 ``accel=accels1[:accels2[:...]]``
61 This is used to enable an accelerator. Depending on the target
62 architecture, kvm, xen, hvf, nvmm, whpx or tcg can be available.
63 By default, tcg is used. If there is more than one accelerator
64 specified, the next one is used if the previous one fails to
67 ``vmport=on|off|auto``
68 Enables emulation of VMWare IO port, for vmmouse etc. auto says
69 to select the value based on accel. For accel=xen the default is
70 off otherwise the default is on.
72 ``dump-guest-core=on|off``
73 Include guest memory in a core dump. The default is on.
76 Enables or disables memory merge support. This feature, when
77 supported by the host, de-duplicates identical memory pages
78 among VMs instances (enabled by default).
80 ``aes-key-wrap=on|off``
81 Enables or disables AES key wrapping support on s390-ccw hosts.
82 This feature controls whether AES wrapping keys will be created
83 to allow execution of AES cryptographic functions. The default
86 ``dea-key-wrap=on|off``
87 Enables or disables DEA key wrapping support on s390-ccw hosts.
88 This feature controls whether DEA wrapping keys will be created
89 to allow execution of DEA cryptographic functions. The default
93 Enables or disables NVDIMM support. The default is off.
95 ``memory-encryption=``
96 Memory encryption object to use. The default is none.
99 Enables or disables ACPI Heterogeneous Memory Attribute Table
100 (HMAT) support. The default is off.
102 ``memory-backend='id'``
103 An alternative to legacy ``-mem-path`` and ``mem-prealloc`` options.
104 Allows to use a memory backend as main RAM.
109 -object memory-backend-file,id=pc.ram,size=512M,mem-path=/hugetlbfs,prealloc=on,share=on
110 -machine memory-backend=pc.ram
113 Migration compatibility note:
115 * as backend id one shall use value of 'default-ram-id', advertised by
116 machine type (available via ``query-machines`` QMP command), if migration
117 to/from old QEMU (<5.0) is expected.
118 * for machine types 4.0 and older, user shall
119 use ``x-use-canonical-path-for-ramblock-id=off`` backend option
120 if migration to/from old QEMU (<5.0) is expected.
125 -object memory-backend-ram,id=pc.ram,size=512M,x-use-canonical-path-for-ramblock-id=off
126 -machine memory-backend=pc.ram
129 ``cxl-fmw.0.targets.0=firsttarget,cxl-fmw.0.targets.1=secondtarget,cxl-fmw.0.size=size[,cxl-fmw.0.interleave-granularity=granularity]``
130 Define a CXL Fixed Memory Window (CFMW).
132 Described in the CXL 2.0 ECN: CEDT CFMWS & QTG _DSM.
134 They are regions of Host Physical Addresses (HPA) on a system which
135 may be interleaved across one or more CXL host bridges. The system
136 software will assign particular devices into these windows and
137 configure the downstream Host-managed Device Memory (HDM) decoders
138 in root ports, switch ports and devices appropriately to meet the
139 interleave requirements before enabling the memory devices.
141 ``targets.X=target`` provides the mapping to CXL host bridges
142 which may be identified by the id provided in the -device entry.
143 Multiple entries are needed to specify all the targets when
144 the fixed memory window represents interleaved memory. X is the
147 ``size=size`` sets the size of the CFMW. This must be a multiple of
148 256MiB. The region will be aligned to 256MiB but the location is
149 platform and configuration dependent.
151 ``interleave-granularity=granularity`` sets the granularity of
152 interleave. Default 256KiB. Only 256KiB, 512KiB, 1024KiB, 2048KiB
153 4096KiB, 8192KiB and 16384KiB granularities supported.
159 -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 DEF("M", HAS_ARG, QEMU_OPTION_M,
163 " sgx-epc.0.memdev=memid,sgx-epc.0.node=numaid\n",
167 ``sgx-epc.0.memdev=@var{memid},sgx-epc.0.node=@var{numaid}``
168 Define an SGX EPC section.
171 DEF("cpu", HAS_ARG, QEMU_OPTION_cpu,
172 "-cpu cpu select CPU ('-cpu help' for list)\n", QEMU_ARCH_ALL)
175 Select CPU model (``-cpu help`` for list and additional feature
179 DEF("accel", HAS_ARG, QEMU_OPTION_accel,
180 "-accel [accel=]accelerator[,prop[=value][,...]]\n"
181 " select accelerator (kvm, xen, hvf, nvmm, whpx or tcg; use 'help' for a list)\n"
182 " igd-passthru=on|off (enable Xen integrated Intel graphics passthrough, default=off)\n"
183 " kernel-irqchip=on|off|split controls accelerated irqchip support (default=on)\n"
184 " kvm-shadow-mem=size of KVM shadow MMU in bytes\n"
185 " one-insn-per-tb=on|off (one guest instruction per TCG translation block)\n"
186 " split-wx=on|off (enable TCG split w^x mapping)\n"
187 " tb-size=n (TCG translation block cache size)\n"
188 " dirty-ring-size=n (KVM dirty ring GFN count, default 0)\n"
189 " eager-split-size=n (KVM Eager Page Split chunk size, default 0, disabled. ARM only)\n"
190 " notify-vmexit=run|internal-error|disable,notify-window=n (enable notify VM exit and set notify window, x86 only)\n"
191 " thread=single|multi (enable multi-threaded TCG)\n", QEMU_ARCH_ALL)
193 ``-accel name[,prop=value[,...]]``
194 This is used to enable an accelerator. Depending on the target
195 architecture, kvm, xen, hvf, nvmm, whpx or tcg can be available. By
196 default, tcg is used. If there is more than one accelerator
197 specified, the next one is used if the previous one fails to
200 ``igd-passthru=on|off``
201 When Xen is in use, this option controls whether Intel
202 integrated graphics devices can be passed through to the guest
205 ``kernel-irqchip=on|off|split``
206 Controls KVM in-kernel irqchip support. The default is full
207 acceleration of the interrupt controllers. On x86, split irqchip
208 reduces the kernel attack surface, at a performance cost for
209 non-MSI interrupts. Disabling the in-kernel irqchip completely
210 is not recommended except for debugging purposes.
212 ``kvm-shadow-mem=size``
213 Defines the size of the KVM shadow MMU.
215 ``one-insn-per-tb=on|off``
216 Makes the TCG accelerator put only one guest instruction into
217 each translation block. This slows down emulation a lot, but
218 can be useful in some situations, such as when trying to analyse
219 the logs produced by the ``-d`` option.
222 Controls the use of split w^x mapping for the TCG code generation
223 buffer. Some operating systems require this to be enabled, and in
224 such a case this will default on. On other operating systems, this
225 will default off, but one may enable this for testing or debugging.
228 Controls the size (in MiB) of the TCG translation block cache.
230 ``thread=single|multi``
231 Controls number of TCG threads. When the TCG is multi-threaded
232 there will be one thread per vCPU therefore taking advantage of
233 additional host cores. The default is to enable multi-threading
234 where both the back-end and front-ends support it and no
235 incompatible TCG features have been enabled (e.g.
238 ``dirty-ring-size=n``
239 When the KVM accelerator is used, it controls the size of the per-vCPU
240 dirty page ring buffer (number of entries for each vCPU). It should
241 be a value that is power of two, and it should be 1024 or bigger (but
242 still less than the maximum value that the kernel supports). 4096
243 could be a good initial value if you have no idea which is the best.
244 Set this value to 0 to disable the feature. By default, this feature
245 is disabled (dirty-ring-size=0). When enabled, KVM will instead
246 record dirty pages in a bitmap.
248 ``eager-split-size=n``
249 KVM implements dirty page logging at the PAGE_SIZE granularity and
250 enabling dirty-logging on a huge-page requires breaking it into
251 PAGE_SIZE pages in the first place. KVM on ARM does this splitting
252 lazily by default. There are performance benefits in doing huge-page
253 split eagerly, especially in situations where TLBI costs associated
254 with break-before-make sequences are considerable and also if guest
255 workloads are read intensive. The size here specifies how many pages
256 to break at a time and needs to be a valid block size which is
257 1GB/2MB/4KB, 32MB/16KB and 512MB/64KB for 4KB/16KB/64KB PAGE_SIZE
258 respectively. Be wary of specifying a higher size as it will have an
259 impact on the memory. By default, this feature is disabled
260 (eager-split-size=0).
262 ``notify-vmexit=run|internal-error|disable,notify-window=n``
263 Enables or disables notify VM exit support on x86 host and specify
264 the corresponding notify window to trigger the VM exit if enabled.
265 ``run`` option enables the feature. It does nothing and continue
266 if the exit happens. ``internal-error`` option enables the feature.
267 It raises a internal error. ``disable`` option doesn't enable the feature.
268 This feature can mitigate the CPU stuck issue due to event windows don't
269 open up for a specified of time (i.e. notify-window).
270 Default: notify-vmexit=run,notify-window=0.
274 DEF("smp", HAS_ARG, QEMU_OPTION_smp,
275 "-smp [[cpus=]n][,maxcpus=maxcpus][,drawers=drawers][,books=books][,sockets=sockets]\n"
276 " [,dies=dies][,clusters=clusters][,cores=cores][,threads=threads]\n"
277 " set the number of initial CPUs to 'n' [default=1]\n"
278 " maxcpus= maximum number of total CPUs, including\n"
279 " offline CPUs for hotplug, etc\n"
280 " drawers= number of drawers on the machine board\n"
281 " books= number of books in one drawer\n"
282 " sockets= number of sockets in one book\n"
283 " dies= number of dies in one socket\n"
284 " clusters= number of clusters in one die\n"
285 " cores= number of cores in one cluster\n"
286 " threads= number of threads in one core\n"
287 "Note: Different machines may have different subsets of the CPU topology\n"
288 " parameters supported, so the actual meaning of the supported parameters\n"
289 " will vary accordingly. For example, for a machine type that supports a\n"
290 " three-level CPU hierarchy of sockets/cores/threads, the parameters will\n"
291 " sequentially mean as below:\n"
292 " sockets means the number of sockets on the machine board\n"
293 " cores means the number of cores in one socket\n"
294 " threads means the number of threads in one core\n"
295 " For a particular machine type board, an expected CPU topology hierarchy\n"
296 " can be defined through the supported sub-option. Unsupported parameters\n"
297 " can also be provided in addition to the sub-option, but their values\n"
298 " must be set as 1 in the purpose of correct parsing.\n",
301 ``-smp [[cpus=]n][,maxcpus=maxcpus][,sockets=sockets][,dies=dies][,clusters=clusters][,cores=cores][,threads=threads]``
302 Simulate a SMP system with '\ ``n``\ ' CPUs initially present on
303 the machine type board. On boards supporting CPU hotplug, the optional
304 '\ ``maxcpus``\ ' parameter can be set to enable further CPUs to be
305 added at runtime. When both parameters are omitted, the maximum number
306 of CPUs will be calculated from the provided topology members and the
307 initial CPU count will match the maximum number. When only one of them
308 is given then the omitted one will be set to its counterpart's value.
309 Both parameters may be specified, but the maximum number of CPUs must
310 be equal to or greater than the initial CPU count. Product of the
311 CPU topology hierarchy must be equal to the maximum number of CPUs.
312 Both parameters are subject to an upper limit that is determined by
313 the specific machine type chosen.
315 To control reporting of CPU topology information, values of the topology
316 parameters can be specified. Machines may only support a subset of the
317 parameters and different machines may have different subsets supported
318 which vary depending on capacity of the corresponding CPU targets. So
319 for a particular machine type board, an expected topology hierarchy can
320 be defined through the supported sub-option. Unsupported parameters can
321 also be provided in addition to the sub-option, but their values must be
322 set as 1 in the purpose of correct parsing.
324 Either the initial CPU count, or at least one of the topology parameters
325 must be specified. The specified parameters must be greater than zero,
326 explicit configuration like "cpus=0" is not allowed. Values for any
327 omitted parameters will be computed from those which are given.
329 For example, the following sub-option defines a CPU topology hierarchy
330 (2 sockets totally on the machine, 2 cores per socket, 2 threads per
331 core) for a machine that only supports sockets/cores/threads.
332 Some members of the option can be omitted but their values will be
333 automatically computed:
337 -smp 8,sockets=2,cores=2,threads=2,maxcpus=8
339 The following sub-option defines a CPU topology hierarchy (2 sockets
340 totally on the machine, 2 dies per socket, 2 cores per die, 2 threads
341 per core) for PC machines which support sockets/dies/cores/threads.
342 Some members of the option can be omitted but their values will be
343 automatically computed:
347 -smp 16,sockets=2,dies=2,cores=2,threads=2,maxcpus=16
349 The following sub-option defines a CPU topology hierarchy (2 sockets
350 totally on the machine, 2 clusters per socket, 2 cores per cluster,
351 2 threads per core) for ARM virt machines which support sockets/clusters
352 /cores/threads. Some members of the option can be omitted but their values
353 will be automatically computed:
357 -smp 16,sockets=2,clusters=2,cores=2,threads=2,maxcpus=16
359 Historically preference was given to the coarsest topology parameters
360 when computing missing values (ie sockets preferred over cores, which
361 were preferred over threads), however, this behaviour is considered
362 liable to change. Prior to 6.2 the preference was sockets over cores
363 over threads. Since 6.2 the preference is cores over sockets over threads.
365 For example, the following option defines a machine board with 2 sockets
366 of 1 core before 6.2 and 1 socket of 2 cores after 6.2:
372 Note: The cluster topology will only be generated in ACPI and exposed
373 to guest if it's explicitly specified in -smp.
376 DEF("numa", HAS_ARG, QEMU_OPTION_numa,
377 "-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
378 "-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n"
379 "-numa dist,src=source,dst=destination,val=distance\n"
380 "-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]\n"
381 "-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"
382 "-numa hmat-cache,node-id=node,size=size,level=level[,associativity=none|direct|complex][,policy=none|write-back|write-through][,line=size]\n",
385 ``-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
387 ``-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]``
389 ``-numa dist,src=source,dst=destination,val=distance``
391 ``-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]``
393 ``-numa hmat-lb,initiator=node,target=node,hierarchy=hierarchy,data-type=type[,latency=lat][,bandwidth=bw]``
395 ``-numa hmat-cache,node-id=node,size=size,level=level[,associativity=str][,policy=str][,line=size]``
396 Define a NUMA node and assign RAM and VCPUs to it. Set the NUMA
397 distance from a source node to a destination node. Set the ACPI
398 Heterogeneous Memory Attributes for the given nodes.
400 Legacy VCPU assignment uses '\ ``cpus``\ ' option where firstcpu and
401 lastcpu are CPU indexes. Each '\ ``cpus``\ ' option represent a
402 contiguous range of CPU indexes (or a single VCPU if lastcpu is
403 omitted). A non-contiguous set of VCPUs can be represented by
404 providing multiple '\ ``cpus``\ ' options. If '\ ``cpus``\ ' is
405 omitted on all nodes, VCPUs are automatically split between them.
407 For example, the following option assigns VCPUs 0, 1, 2 and 5 to a
412 -numa node,cpus=0-2,cpus=5
414 '\ ``cpu``\ ' option is a new alternative to '\ ``cpus``\ ' option
415 which uses '\ ``socket-id|core-id|thread-id``\ ' properties to
416 assign CPU objects to a node using topology layout properties of
417 CPU. The set of properties is machine specific, and depends on used
418 machine type/'\ ``smp``\ ' options. It could be queried with
419 '\ ``hotpluggable-cpus``\ ' monitor command. '\ ``node-id``\ '
420 property specifies node to which CPU object will be assigned, it's
421 required for node to be declared with '\ ``node``\ ' option before
422 it's used with '\ ``cpu``\ ' option.
429 -smp 1,sockets=2,maxcpus=2 \
430 -numa node,nodeid=0 -numa node,nodeid=1 \
431 -numa cpu,node-id=0,socket-id=0 -numa cpu,node-id=1,socket-id=1
433 '\ ``memdev``\ ' option assigns RAM from a given memory backend
434 device to a node. It is recommended to use '\ ``memdev``\ ' option
435 over legacy '\ ``mem``\ ' option. This is because '\ ``memdev``\ '
436 option provides better performance and more control over the
437 backend's RAM (e.g. '\ ``prealloc``\ ' parameter of
438 '\ ``-memory-backend-ram``\ ' allows memory preallocation).
440 For compatibility reasons, legacy '\ ``mem``\ ' option is
441 supported in 5.0 and older machine types. Note that '\ ``mem``\ '
442 and '\ ``memdev``\ ' are mutually exclusive. If one node uses
443 '\ ``memdev``\ ', the rest nodes have to use '\ ``memdev``\ '
444 option, and vice versa.
446 Users must specify memory for all NUMA nodes by '\ ``memdev``\ '
447 (or legacy '\ ``mem``\ ' if available). In QEMU 5.2, the support
448 for '\ ``-numa node``\ ' without memory specified was removed.
450 '\ ``initiator``\ ' is an additional option that points to an
451 initiator NUMA node that has best performance (the lowest latency or
452 largest bandwidth) to this NUMA node. Note that this option can be
453 set only when the machine property 'hmat' is set to 'on'.
455 Following example creates a machine with 2 NUMA nodes, node 0 has
456 CPU. node 1 has only memory, and its initiator is node 0. Note that
457 because node 0 has CPU, by default the initiator of node 0 is itself
463 -m 2G,slots=2,maxmem=4G \
464 -object memory-backend-ram,size=1G,id=m0 \
465 -object memory-backend-ram,size=1G,id=m1 \
466 -numa node,nodeid=0,memdev=m0 \
467 -numa node,nodeid=1,memdev=m1,initiator=0 \
468 -smp 2,sockets=2,maxcpus=2 \
469 -numa cpu,node-id=0,socket-id=0 \
470 -numa cpu,node-id=0,socket-id=1
472 source and destination are NUMA node IDs. distance is the NUMA
473 distance from source to destination. The distance from a node to
474 itself is always 10. If any pair of nodes is given a distance, then
475 all pairs must be given distances. Although, when distances are only
476 given in one direction for each pair of nodes, then the distances in
477 the opposite directions are assumed to be the same. If, however, an
478 asymmetrical pair of distances is given for even one node pair, then
479 all node pairs must be provided distance values for both directions,
480 even when they are symmetrical. When a node is unreachable from
481 another node, set the pair's distance to 255.
483 Note that the -``numa`` option doesn't allocate any of the specified
484 resources, it just assigns existing resources to NUMA nodes. This
485 means that one still has to use the ``-m``, ``-smp`` options to
486 allocate RAM and VCPUs respectively.
488 Use '\ ``hmat-lb``\ ' to set System Locality Latency and Bandwidth
489 Information between initiator and target NUMA nodes in ACPI
490 Heterogeneous Attribute Memory Table (HMAT). Initiator NUMA node can
491 create memory requests, usually it has one or more processors.
492 Target NUMA node contains addressable memory.
494 In '\ ``hmat-lb``\ ' option, node are NUMA node IDs. hierarchy is
495 the memory hierarchy of the target NUMA node: if hierarchy is
496 'memory', the structure represents the memory performance; if
497 hierarchy is 'first-level\|second-level\|third-level', this
498 structure represents aggregated performance of memory side caches
499 for each domain. type of 'data-type' is type of data represented by
500 this structure instance: if 'hierarchy' is 'memory', 'data-type' is
501 'access\|read\|write' latency or 'access\|read\|write' bandwidth of
502 the target memory; if 'hierarchy' is
503 'first-level\|second-level\|third-level', 'data-type' is
504 'access\|read\|write' hit latency or 'access\|read\|write' hit
505 bandwidth of the target memory side cache.
507 lat is latency value in nanoseconds. bw is bandwidth value, the
508 possible value and units are NUM[M\|G\|T], mean that the bandwidth
509 value are NUM byte per second (or MB/s, GB/s or TB/s depending on
510 used suffix). Note that if latency or bandwidth value is 0, means
511 the corresponding latency or bandwidth information is not provided.
513 In '\ ``hmat-cache``\ ' option, node-id is the NUMA-id of the memory
514 belongs. size is the size of memory side cache in bytes. level is
515 the cache level described in this structure, note that the cache
516 level 0 should not be used with '\ ``hmat-cache``\ ' option.
517 associativity is the cache associativity, the possible value is
518 'none/direct(direct-mapped)/complex(complex cache indexing)'. policy
519 is the write policy. line is the cache Line size in bytes.
521 For example, the following options describe 2 NUMA nodes. Node 0 has
522 2 cpus and a ram, node 1 has only a ram. The processors in node 0
523 access memory in node 0 with access-latency 5 nanoseconds,
524 access-bandwidth is 200 MB/s; The processors in NUMA node 0 access
525 memory in NUMA node 1 with access-latency 10 nanoseconds,
526 access-bandwidth is 100 MB/s. And for memory side cache information,
527 NUMA node 0 and 1 both have 1 level memory cache, size is 10KB,
528 policy is write-back, the cache Line size is 8 bytes:
534 -object memory-backend-ram,size=1G,id=m0 \
535 -object memory-backend-ram,size=1G,id=m1 \
536 -smp 2,sockets=2,maxcpus=2 \
537 -numa node,nodeid=0,memdev=m0 \
538 -numa node,nodeid=1,memdev=m1,initiator=0 \
539 -numa cpu,node-id=0,socket-id=0 \
540 -numa cpu,node-id=0,socket-id=1 \
541 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-latency,latency=5 \
542 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-bandwidth,bandwidth=200M \
543 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-latency,latency=10 \
544 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-bandwidth,bandwidth=100M \
545 -numa hmat-cache,node-id=0,size=10K,level=1,associativity=direct,policy=write-back,line=8 \
546 -numa hmat-cache,node-id=1,size=10K,level=1,associativity=direct,policy=write-back,line=8
549 DEF("add-fd", HAS_ARG, QEMU_OPTION_add_fd,
550 "-add-fd fd=fd,set=set[,opaque=opaque]\n"
551 " Add 'fd' to fd 'set'\n", QEMU_ARCH_ALL)
553 ``-add-fd fd=fd,set=set[,opaque=opaque]``
554 Add a file descriptor to an fd set. Valid options are:
557 This option defines the file descriptor of which a duplicate is
558 added to fd set. The file descriptor cannot be stdin, stdout, or
562 This option defines the ID of the fd set to add the file
566 This option defines a free-form string that can be used to
569 You can open an image using pre-opened file descriptors from an fd
575 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
576 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
577 -drive file=/dev/fdset/2,index=0,media=disk
580 DEF("set", HAS_ARG, QEMU_OPTION_set,
581 "-set group.id.arg=value\n"
582 " set <arg> parameter for item <id> of type <group>\n"
583 " i.e. -set drive.$id.file=/path/to/image\n", QEMU_ARCH_ALL)
585 ``-set group.id.arg=value``
586 Set parameter arg for item id of type group
589 DEF("global", HAS_ARG, QEMU_OPTION_global,
590 "-global driver.property=value\n"
591 "-global driver=driver,property=property,value=value\n"
592 " set a global default for a driver property\n",
595 ``-global driver.prop=value``
597 ``-global driver=driver,property=property,value=value``
598 Set default value of driver's property prop to value, e.g.:
602 |qemu_system_x86| -global ide-hd.physical_block_size=4096 disk-image.img
604 In particular, you can use this to set driver properties for devices
605 which are created automatically by the machine model. To create a
606 device which is not created automatically and set properties on it,
609 -global driver.prop=value is shorthand for -global
610 driver=driver,property=prop,value=value. The longhand syntax works
611 even when driver contains a dot.
614 DEF("boot", HAS_ARG, QEMU_OPTION_boot,
615 "-boot [order=drives][,once=drives][,menu=on|off]\n"
616 " [,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_time][,strict=on|off]\n"
617 " 'drives': floppy (a), hard disk (c), CD-ROM (d), network (n)\n"
618 " 'sp_name': the file's name that would be passed to bios as logo picture, if menu=on\n"
619 " 'sp_time': the period that splash picture last if menu=on, unit is ms\n"
620 " 'rb_timeout': the timeout before guest reboot when boot failed, unit is ms\n",
623 ``-boot [order=drives][,once=drives][,menu=on|off][,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_timeout][,strict=on|off]``
624 Specify boot order drives as a string of drive letters. Valid drive
625 letters depend on the target architecture. The x86 PC uses: a, b
626 (floppy 1 and 2), c (first hard disk), d (first CD-ROM), n-p
627 (Etherboot from network adapter 1-4), hard disk boot is the default.
628 To apply a particular boot order only on the first startup, specify
629 it via ``once``. Note that the ``order`` or ``once`` parameter
630 should not be used together with the ``bootindex`` property of
631 devices, since the firmware implementations normally do not support
632 both at the same time.
634 Interactive boot menus/prompts can be enabled via ``menu=on`` as far
635 as firmware/BIOS supports them. The default is non-interactive boot.
637 A splash picture could be passed to bios, enabling user to show it
638 as logo, when option splash=sp\_name is given and menu=on, If
639 firmware/BIOS supports them. Currently Seabios for X86 system
640 support it. limitation: The splash file could be a jpeg file or a
641 BMP file in 24 BPP format(true color). The resolution should be
642 supported by the SVGA mode, so the recommended is 320x240, 640x480,
645 A timeout could be passed to bios, guest will pause for rb\_timeout
646 ms when boot failed, then reboot. If rb\_timeout is '-1', guest will
647 not reboot, qemu passes '-1' to bios by default. Currently Seabios
648 for X86 system support it.
650 Do strict boot via ``strict=on`` as far as firmware/BIOS supports
651 it. This only effects when boot priority is changed by bootindex
652 options. The default is non-strict boot.
656 # try to boot from network first, then from hard disk
657 |qemu_system_x86| -boot order=nc
658 # boot from CD-ROM first, switch back to default order after reboot
659 |qemu_system_x86| -boot once=d
660 # boot with a splash picture for 5 seconds.
661 |qemu_system_x86| -boot menu=on,splash=/root/boot.bmp,splash-time=5000
663 Note: The legacy format '-boot drives' is still supported but its
664 use is discouraged as it may be removed from future versions.
667 DEF("m", HAS_ARG, QEMU_OPTION_m,
668 "-m [size=]megs[,slots=n,maxmem=size]\n"
669 " configure guest RAM\n"
670 " size: initial amount of guest memory\n"
671 " slots: number of hotplug slots (default: none)\n"
672 " maxmem: maximum amount of guest memory (default: none)\n"
673 " Note: Some architectures might enforce a specific granularity\n",
676 ``-m [size=]megs[,slots=n,maxmem=size]``
677 Sets guest startup RAM size to megs megabytes. Default is 128 MiB.
678 Optionally, a suffix of "M" or "G" can be used to signify a value in
679 megabytes or gigabytes respectively. Optional pair slots, maxmem
680 could be used to set amount of hotpluggable memory slots and maximum
681 amount of memory. Note that maxmem must be aligned to the page size.
683 For example, the following command-line sets the guest startup RAM
684 size to 1GB, creates 3 slots to hotplug additional memory and sets
685 the maximum memory the guest can reach to 4GB:
689 |qemu_system| -m 1G,slots=3,maxmem=4G
691 If slots and maxmem are not specified, memory hotplug won't be
692 enabled and the guest startup RAM will never increase.
695 DEF("mem-path", HAS_ARG, QEMU_OPTION_mempath,
696 "-mem-path FILE provide backing storage for guest RAM\n", QEMU_ARCH_ALL)
699 Allocate guest RAM from a temporarily created file in path.
702 DEF("mem-prealloc", 0, QEMU_OPTION_mem_prealloc,
703 "-mem-prealloc preallocate guest memory (use with -mem-path)\n",
707 Preallocate memory when using -mem-path.
710 DEF("k", HAS_ARG, QEMU_OPTION_k,
711 "-k language use keyboard layout (for example 'fr' for French)\n",
715 Use keyboard layout language (for example ``fr`` for French). This
716 option is only needed where it is not easy to get raw PC keycodes
717 (e.g. on Macs, with some X11 servers or with a VNC or curses
718 display). You don't normally need to use it on PC/Linux or
721 The available layouts are:
725 ar de-ch es fo fr-ca hu ja mk no pt-br sv
726 da en-gb et fr fr-ch is lt nl pl ru th
727 de en-us fi fr-be hr it lv nl-be pt sl tr
729 The default is ``en-us``.
733 DEF("audio", HAS_ARG, QEMU_OPTION_audio,
734 "-audio [driver=]driver[,prop[=value][,...]]\n"
735 " specifies default audio backend when `audiodev` is not\n"
736 " used to create a machine or sound device;"
737 " options are the same as for -audiodev\n"
738 "-audio [driver=]driver,model=value[,prop[=value][,...]]\n"
739 " specifies the audio backend and device to use;\n"
740 " apart from 'model', options are the same as for -audiodev.\n"
741 " use '-audio model=help' to show possible devices.\n",
744 ``-audio [driver=]driver[,model=value][,prop[=value][,...]]``
745 If the ``model`` option is specified, ``-audio`` is a shortcut
746 for configuring both the guest audio hardware and the host audio
747 backend in one go. The guest hardware model can be set with
748 ``model=modelname``. Use ``model=help`` to list the available
751 The following two example do exactly the same, to show how ``-audio``
752 can be used to shorten the command line length:
756 |qemu_system| -audiodev pa,id=pa -device sb16,audiodev=pa
757 |qemu_system| -audio pa,model=sb16
759 If the ``model`` option is not specified, ``-audio`` is used to
760 configure a default audio backend that will be used whenever the
761 ``audiodev`` property is not set on a device or machine. In
762 particular, ``-audio none`` ensures that no audio is produced even
763 for machines that have embedded sound hardware.
765 In both cases, the driver option is the same as with the corresponding
766 ``-audiodev`` option below. Use ``driver=help`` to list the available
771 DEF("audiodev", HAS_ARG, QEMU_OPTION_audiodev,
772 "-audiodev [driver=]driver,id=id[,prop[=value][,...]]\n"
773 " specifies the audio backend to use\n"
774 " Use ``-audiodev help`` to list the available drivers\n"
775 " id= identifier of the backend\n"
776 " timer-period= timer period in microseconds\n"
777 " in|out.mixing-engine= use mixing engine to mix streams inside QEMU\n"
778 " in|out.fixed-settings= use fixed settings for host audio\n"
779 " in|out.frequency= frequency to use with fixed settings\n"
780 " in|out.channels= number of channels to use with fixed settings\n"
781 " in|out.format= sample format to use with fixed settings\n"
782 " valid values: s8, s16, s32, u8, u16, u32, f32\n"
783 " in|out.voices= number of voices to use\n"
784 " in|out.buffer-length= length of buffer in microseconds\n"
785 "-audiodev none,id=id,[,prop[=value][,...]]\n"
786 " dummy driver that discards all output\n"
787 #ifdef CONFIG_AUDIO_ALSA
788 "-audiodev alsa,id=id[,prop[=value][,...]]\n"
789 " in|out.dev= name of the audio device to use\n"
790 " in|out.period-length= length of period in microseconds\n"
791 " in|out.try-poll= attempt to use poll mode\n"
792 " threshold= threshold (in microseconds) when playback starts\n"
794 #ifdef CONFIG_AUDIO_COREAUDIO
795 "-audiodev coreaudio,id=id[,prop[=value][,...]]\n"
796 " in|out.buffer-count= number of buffers\n"
798 #ifdef CONFIG_AUDIO_DSOUND
799 "-audiodev dsound,id=id[,prop[=value][,...]]\n"
800 " latency= add extra latency to playback in microseconds\n"
802 #ifdef CONFIG_AUDIO_OSS
803 "-audiodev oss,id=id[,prop[=value][,...]]\n"
804 " in|out.dev= path of the audio device to use\n"
805 " in|out.buffer-count= number of buffers\n"
806 " in|out.try-poll= attempt to use poll mode\n"
807 " try-mmap= try using memory mapped access\n"
808 " exclusive= open device in exclusive mode\n"
809 " dsp-policy= set timing policy (0..10), -1 to use fragment mode\n"
811 #ifdef CONFIG_AUDIO_PA
812 "-audiodev pa,id=id[,prop[=value][,...]]\n"
813 " server= PulseAudio server address\n"
814 " in|out.name= source/sink device name\n"
815 " in|out.latency= desired latency in microseconds\n"
817 #ifdef CONFIG_AUDIO_PIPEWIRE
818 "-audiodev pipewire,id=id[,prop[=value][,...]]\n"
819 " in|out.name= source/sink device name\n"
820 " in|out.stream-name= name of pipewire stream\n"
821 " in|out.latency= desired latency in microseconds\n"
823 #ifdef CONFIG_AUDIO_SDL
824 "-audiodev sdl,id=id[,prop[=value][,...]]\n"
825 " in|out.buffer-count= number of buffers\n"
827 #ifdef CONFIG_AUDIO_SNDIO
828 "-audiodev sndio,id=id[,prop[=value][,...]]\n"
831 "-audiodev spice,id=id[,prop[=value][,...]]\n"
833 #ifdef CONFIG_DBUS_DISPLAY
834 "-audiodev dbus,id=id[,prop[=value][,...]]\n"
836 "-audiodev wav,id=id[,prop[=value][,...]]\n"
837 " path= path of wav file to record\n",
840 ``-audiodev [driver=]driver,id=id[,prop[=value][,...]]``
841 Adds a new audio backend driver identified by id. There are global
842 and driver specific properties. Some values can be set differently
843 for input and output, they're marked with ``in|out.``. You can set
844 the input's property with ``in.prop`` and the output's property with
845 ``out.prop``. For example:
849 -audiodev alsa,id=example,in.frequency=44110,out.frequency=8000
850 -audiodev alsa,id=example,out.channels=1 # leaves in.channels unspecified
852 NOTE: parameter validation is known to be incomplete, in many cases
853 specifying an invalid option causes QEMU to print an error message
854 and continue emulation without sound.
856 Valid global options are:
859 Identifies the audio backend.
861 ``timer-period=period``
862 Sets the timer period used by the audio subsystem in
863 microseconds. Default is 10000 (10 ms).
865 ``in|out.mixing-engine=on|off``
866 Use QEMU's mixing engine to mix all streams inside QEMU and
867 convert audio formats when not supported by the backend. When
868 off, fixed-settings must be off too. Note that disabling this
869 option means that the selected backend must support multiple
870 streams and the audio formats used by the virtual cards,
871 otherwise you'll get no sound. It's not recommended to disable
872 this option unless you want to use 5.1 or 7.1 audio, as mixing
873 engine only supports mono and stereo audio. Default is on.
875 ``in|out.fixed-settings=on|off``
876 Use fixed settings for host audio. When off, it will change
877 based on how the guest opens the sound card. In this case you
878 must not specify frequency, channels or format. Default is on.
880 ``in|out.frequency=frequency``
881 Specify the frequency to use when using fixed-settings. Default
884 ``in|out.channels=channels``
885 Specify the number of channels to use when using fixed-settings.
886 Default is 2 (stereo).
888 ``in|out.format=format``
889 Specify the sample format to use when using fixed-settings.
890 Valid values are: ``s8``, ``s16``, ``s32``, ``u8``, ``u16``,
891 ``u32``, ``f32``. Default is ``s16``.
893 ``in|out.voices=voices``
894 Specify the number of voices to use. Default is 1.
896 ``in|out.buffer-length=usecs``
897 Sets the size of the buffer in microseconds.
899 ``-audiodev none,id=id[,prop[=value][,...]]``
900 Creates a dummy backend that discards all outputs. This backend has
901 no backend specific properties.
903 ``-audiodev alsa,id=id[,prop[=value][,...]]``
904 Creates backend using the ALSA. This backend is only available on
907 ALSA specific options are:
909 ``in|out.dev=device``
910 Specify the ALSA device to use for input and/or output. Default
913 ``in|out.period-length=usecs``
914 Sets the period length in microseconds.
916 ``in|out.try-poll=on|off``
917 Attempt to use poll mode with the device. Default is on.
919 ``threshold=threshold``
920 Threshold (in microseconds) when playback starts. Default is 0.
922 ``-audiodev coreaudio,id=id[,prop[=value][,...]]``
923 Creates a backend using Apple's Core Audio. This backend is only
924 available on Mac OS and only supports playback.
926 Core Audio specific options are:
928 ``in|out.buffer-count=count``
929 Sets the count of the buffers.
931 ``-audiodev dsound,id=id[,prop[=value][,...]]``
932 Creates a backend using Microsoft's DirectSound. This backend is
933 only available on Windows and only supports playback.
935 DirectSound specific options are:
938 Add extra usecs microseconds latency to playback. Default is
941 ``-audiodev oss,id=id[,prop[=value][,...]]``
942 Creates a backend using OSS. This backend is available on most
945 OSS specific options are:
947 ``in|out.dev=device``
948 Specify the file name of the OSS device to use. Default is
951 ``in|out.buffer-count=count``
952 Sets the count of the buffers.
954 ``in|out.try-poll=on|of``
955 Attempt to use poll mode with the device. Default is on.
958 Try using memory mapped device access. Default is off.
961 Open the device in exclusive mode (vmix won't work in this
962 case). Default is off.
964 ``dsp-policy=policy``
965 Sets the timing policy (between 0 and 10, where smaller number
966 means smaller latency but higher CPU usage). Use -1 to use
967 buffer sizes specified by ``buffer`` and ``buffer-count``. This
968 option is ignored if you do not have OSS 4. Default is 5.
970 ``-audiodev pa,id=id[,prop[=value][,...]]``
971 Creates a backend using PulseAudio. This backend is available on
974 PulseAudio specific options are:
977 Sets the PulseAudio server to connect to.
980 Use the specified source/sink for recording/playback.
982 ``in|out.latency=usecs``
983 Desired latency in microseconds. The PulseAudio server will try
984 to honor this value but actual latencies may be lower or higher.
986 ``-audiodev pipewire,id=id[,prop[=value][,...]]``
987 Creates a backend using PipeWire. This backend is available on
990 PipeWire specific options are:
992 ``in|out.latency=usecs``
993 Desired latency in microseconds.
996 Use the specified source/sink for recording/playback.
998 ``in|out.stream-name``
999 Specify the name of pipewire stream.
1001 ``-audiodev sdl,id=id[,prop[=value][,...]]``
1002 Creates a backend using SDL. This backend is available on most
1003 systems, but you should use your platform's native backend if
1006 SDL specific options are:
1008 ``in|out.buffer-count=count``
1009 Sets the count of the buffers.
1011 ``-audiodev sndio,id=id[,prop[=value][,...]]``
1012 Creates a backend using SNDIO. This backend is available on
1013 OpenBSD and most other Unix-like systems.
1015 Sndio specific options are:
1017 ``in|out.dev=device``
1018 Specify the sndio device to use for input and/or output. Default
1021 ``in|out.latency=usecs``
1022 Sets the desired period length in microseconds.
1024 ``-audiodev spice,id=id[,prop[=value][,...]]``
1025 Creates a backend that sends audio through SPICE. This backend
1026 requires ``-spice`` and automatically selected in that case, so
1027 usually you can ignore this option. This backend has no backend
1028 specific properties.
1030 ``-audiodev wav,id=id[,prop[=value][,...]]``
1031 Creates a backend that writes audio to a WAV file.
1033 Backend specific options are:
1036 Write recorded audio into the specified file. Default is
1040 DEF("device", HAS_ARG, QEMU_OPTION_device,
1041 "-device driver[,prop[=value][,...]]\n"
1042 " add device (based on driver)\n"
1043 " prop=value,... sets driver properties\n"
1044 " use '-device help' to print all possible drivers\n"
1045 " use '-device driver,help' to print all possible properties\n",
1048 ``-device driver[,prop[=value][,...]]``
1049 Add device driver. prop=value sets driver properties. Valid
1050 properties depend on the driver. To get help on possible drivers and
1051 properties, use ``-device help`` and ``-device driver,help``.
1055 ``-device ipmi-bmc-sim,id=id[,prop[=value][,...]]``
1056 Add an IPMI BMC. This is a simulation of a hardware management
1057 interface processor that normally sits on a system. It provides a
1058 watchdog and the ability to reset and power control the system. You
1059 need to connect this to an IPMI interface to make it useful
1061 The IPMI slave address to use for the BMC. The default is 0x20. This
1062 address is the BMC's address on the I2C network of management
1063 controllers. If you don't know what this means, it is safe to ignore
1067 The BMC id for interfaces to use this device.
1070 Define slave address to use for the BMC. The default is 0x20.
1073 file containing raw Sensor Data Records (SDR) data. The default
1077 size of a Field Replaceable Unit (FRU) area. The default is
1080 ``frudatafile=file``
1081 file containing raw Field Replaceable Unit (FRU) inventory data.
1082 The default is none.
1085 value for the GUID for the BMC, in standard UUID format. If this
1086 is set, get "Get GUID" command to the BMC will return it.
1087 Otherwise "Get GUID" will return an error.
1089 ``-device ipmi-bmc-extern,id=id,chardev=id[,slave_addr=val]``
1090 Add a connection to an external IPMI BMC simulator. Instead of
1091 locally emulating the BMC like the above item, instead connect to an
1092 external entity that provides the IPMI services.
1094 A connection is made to an external BMC simulator. If you do this,
1095 it is strongly recommended that you use the "reconnect=" chardev
1096 option to reconnect to the simulator if the connection is lost. Note
1097 that if this is not used carefully, it can be a security issue, as
1098 the interface has the ability to send resets, NMIs, and power off
1099 the VM. It's best if QEMU makes a connection to an external
1100 simulator running on a secure port on localhost, so neither the
1101 simulator nor QEMU is exposed to any outside network.
1103 See the "lanserv/README.vm" file in the OpenIPMI library for more
1104 details on the external interface.
1106 ``-device isa-ipmi-kcs,bmc=id[,ioport=val][,irq=val]``
1107 Add a KCS IPMI interface on the ISA bus. This also adds a
1108 corresponding ACPI and SMBIOS entries, if appropriate.
1111 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern
1115 Define the I/O address of the interface. The default is 0xca0
1119 Define the interrupt to use. The default is 5. To disable
1120 interrupts, set this to 0.
1122 ``-device isa-ipmi-bt,bmc=id[,ioport=val][,irq=val]``
1123 Like the KCS interface, but defines a BT interface. The default port
1124 is 0xe4 and the default interrupt is 5.
1126 ``-device pci-ipmi-kcs,bmc=id``
1127 Add a KCS IPMI interface on the PCI bus.
1130 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern above.
1132 ``-device pci-ipmi-bt,bmc=id``
1133 Like the KCS interface, but defines a BT interface on the PCI bus.
1135 ``-device intel-iommu[,option=...]``
1136 This is only supported by ``-machine q35``, which will enable Intel VT-d
1137 emulation within the guest. It supports below options:
1139 ``intremap=on|off`` (default: auto)
1140 This enables interrupt remapping feature. It's required to enable
1141 complete x2apic. Currently it only supports kvm kernel-irqchip modes
1142 ``off`` or ``split``, while full kernel-irqchip is not yet supported.
1143 The default value is "auto", which will be decided by the mode of
1146 ``caching-mode=on|off`` (default: off)
1147 This enables caching mode for the VT-d emulated device. When
1148 caching-mode is enabled, each guest DMA buffer mapping will generate an
1149 IOTLB invalidation from the guest IOMMU driver to the vIOMMU device in
1150 a synchronous way. It is required for ``-device vfio-pci`` to work
1151 with the VT-d device, because host assigned devices requires to setup
1152 the DMA mapping on the host before guest DMA starts.
1154 ``device-iotlb=on|off`` (default: off)
1155 This enables device-iotlb capability for the emulated VT-d device. So
1156 far virtio/vhost should be the only real user for this parameter,
1157 paired with ats=on configured for the device.
1159 ``aw-bits=39|48`` (default: 39)
1160 This decides the address width of IOVA address space. The address
1161 space has 39 bits width for 3-level IOMMU page tables, and 48 bits for
1162 4-level IOMMU page tables.
1164 Please also refer to the wiki page for general scenarios of VT-d
1165 emulation in QEMU: https://wiki.qemu.org/Features/VT-d.
1169 DEF("name", HAS_ARG, QEMU_OPTION_name,
1170 "-name string1[,process=string2][,debug-threads=on|off]\n"
1171 " set the name of the guest\n"
1172 " string1 sets the window title and string2 the process name\n"
1173 " When debug-threads is enabled, individual threads are given a separate name\n"
1174 " NOTE: The thread names are for debugging and not a stable API.\n",
1178 Sets the name of the guest. This name will be displayed in the SDL
1179 window caption. The name will also be used for the VNC server. Also
1180 optionally set the top visible process name in Linux. Naming of
1181 individual threads can also be enabled on Linux to aid debugging.
1184 DEF("uuid", HAS_ARG, QEMU_OPTION_uuid,
1185 "-uuid %08x-%04x-%04x-%04x-%012x\n"
1186 " specify machine UUID\n", QEMU_ARCH_ALL)
1194 DEFHEADING(Block device options:)
1197 The QEMU block device handling options have a long history and
1198 have gone through several iterations as the feature set and complexity
1199 of the block layer have grown. Many online guides to QEMU often
1200 reference older and deprecated options, which can lead to confusion.
1202 The most explicit way to describe disks is to use a combination of
1203 ``-device`` to specify the hardware device and ``-blockdev`` to
1204 describe the backend. The device defines what the guest sees and the
1205 backend describes how QEMU handles the data. It is the only guaranteed
1206 stable interface for describing block devices and as such is
1207 recommended for management tools and scripting.
1209 The ``-drive`` option combines the device and backend into a single
1210 command line option which is a more human friendly. There is however no
1211 interface stability guarantee although some older board models still
1212 need updating to work with the modern blockdev forms.
1214 Older options like ``-hda`` are essentially macros which expand into
1215 ``-drive`` options for various drive interfaces. The original forms
1216 bake in a lot of assumptions from the days when QEMU was emulating a
1217 legacy PC, they are not recommended for modern configurations.
1221 DEF("fda", HAS_ARG, QEMU_OPTION_fda,
1222 "-fda/-fdb file use 'file' as floppy disk 0/1 image\n", QEMU_ARCH_ALL)
1223 DEF("fdb", HAS_ARG, QEMU_OPTION_fdb, "", QEMU_ARCH_ALL)
1228 Use file as floppy disk 0/1 image (see the :ref:`disk images` chapter in
1229 the System Emulation Users Guide).
1232 DEF("hda", HAS_ARG, QEMU_OPTION_hda,
1233 "-hda/-hdb file use 'file' as hard disk 0/1 image\n", QEMU_ARCH_ALL)
1234 DEF("hdb", HAS_ARG, QEMU_OPTION_hdb, "", QEMU_ARCH_ALL)
1235 DEF("hdc", HAS_ARG, QEMU_OPTION_hdc,
1236 "-hdc/-hdd file use 'file' as hard disk 2/3 image\n", QEMU_ARCH_ALL)
1237 DEF("hdd", HAS_ARG, QEMU_OPTION_hdd, "", QEMU_ARCH_ALL)
1246 Use file as hard disk 0, 1, 2 or 3 image on the default bus of the
1247 emulated machine (this is for example the IDE bus on most x86 machines,
1248 but it can also be SCSI, virtio or something else on other target
1249 architectures). See also the :ref:`disk images` chapter in the System
1250 Emulation Users Guide.
1253 DEF("cdrom", HAS_ARG, QEMU_OPTION_cdrom,
1254 "-cdrom file use 'file' as CD-ROM image\n",
1258 Use file as CD-ROM image on the default bus of the emulated machine
1259 (which is IDE1 master on x86, so you cannot use ``-hdc`` and ``-cdrom``
1260 at the same time there). On systems that support it, you can use the
1261 host CD-ROM by using ``/dev/cdrom`` as filename.
1264 DEF("blockdev", HAS_ARG, QEMU_OPTION_blockdev,
1265 "-blockdev [driver=]driver[,node-name=N][,discard=ignore|unmap]\n"
1266 " [,cache.direct=on|off][,cache.no-flush=on|off]\n"
1267 " [,read-only=on|off][,auto-read-only=on|off]\n"
1268 " [,force-share=on|off][,detect-zeroes=on|off|unmap]\n"
1269 " [,driver specific parameters...]\n"
1270 " configure a block backend\n", QEMU_ARCH_ALL)
1272 ``-blockdev option[,option[,option[,...]]]``
1273 Define a new block driver node. Some of the options apply to all
1274 block drivers, other options are only accepted for a specific block
1275 driver. See below for a list of generic options and options for the
1276 most common block drivers.
1278 Options that expect a reference to another node (e.g. ``file``) can
1279 be given in two ways. Either you specify the node name of an already
1280 existing node (file=node-name), or you define a new node inline,
1281 adding options for the referenced node after a dot
1282 (file.filename=path,file.aio=native).
1284 A block driver node created with ``-blockdev`` can be used for a
1285 guest device by specifying its node name for the ``drive`` property
1286 in a ``-device`` argument that defines a block device.
1288 ``Valid options for any block driver node:``
1290 Specifies the block driver to use for the given node.
1293 This defines the name of the block driver node by which it
1294 will be referenced later. The name must be unique, i.e. it
1295 must not match the name of a different block driver node, or
1296 (if you use ``-drive`` as well) the ID of a drive.
1298 If no node name is specified, it is automatically generated.
1299 The generated node name is not intended to be predictable
1300 and changes between QEMU invocations. For the top level, an
1301 explicit node name must be specified.
1304 Open the node read-only. Guest write attempts will fail.
1306 Note that some block drivers support only read-only access,
1307 either generally or in certain configurations. In this case,
1308 the default value ``read-only=off`` does not work and the
1309 option must be specified explicitly.
1312 If ``auto-read-only=on`` is set, QEMU may fall back to
1313 read-only usage even when ``read-only=off`` is requested, or
1314 even switch between modes as needed, e.g. depending on
1315 whether the image file is writable or whether a writing user
1316 is attached to the node.
1319 Override the image locking system of QEMU by forcing the
1320 node to utilize weaker shared access for permissions where
1321 it would normally request exclusive access. When there is
1322 the potential for multiple instances to have the same file
1323 open (whether this invocation of QEMU is the first or the
1324 second instance), both instances must permit shared access
1325 for the second instance to succeed at opening the file.
1327 Enabling ``force-share=on`` requires ``read-only=on``.
1330 The host page cache can be avoided with ``cache.direct=on``.
1331 This will attempt to do disk IO directly to the guest's
1332 memory. QEMU may still perform an internal copy of the data.
1335 In case you don't care about data integrity over host
1336 failures, you can use ``cache.no-flush=on``. This option
1337 tells QEMU that it never needs to write any data to the disk
1338 but can instead keep things in cache. If anything goes
1339 wrong, like your host losing power, the disk storage getting
1340 disconnected accidentally, etc. your image will most
1341 probably be rendered unusable.
1344 discard is one of "ignore" (or "off") or "unmap" (or "on")
1345 and controls whether ``discard`` (also known as ``trim`` or
1346 ``unmap``) requests are ignored or passed to the filesystem.
1347 Some machine types may not support discard requests.
1349 ``detect-zeroes=detect-zeroes``
1350 detect-zeroes is "off", "on" or "unmap" and enables the
1351 automatic conversion of plain zero writes by the OS to
1352 driver specific optimized zero write commands. You may even
1353 choose "unmap" if discard is set to "unmap" to allow a zero
1354 write to be converted to an ``unmap`` operation.
1356 ``Driver-specific options for file``
1357 This is the protocol-level block driver for accessing regular
1361 The path to the image file in the local filesystem
1364 Specifies the AIO backend (threads/native/io_uring,
1368 Specifies whether the image file is protected with Linux OFD
1369 / POSIX locks. The default is to use the Linux Open File
1370 Descriptor API if available, otherwise no lock is applied.
1371 (auto/on/off, default: auto)
1377 -blockdev driver=file,node-name=disk,filename=disk.img
1379 ``Driver-specific options for raw``
1380 This is the image format block driver for raw images. It is
1381 usually stacked on top of a protocol level block driver such as
1385 Reference to or definition of the data source block driver
1386 node (e.g. a ``file`` driver node)
1392 -blockdev driver=file,node-name=disk_file,filename=disk.img
1393 -blockdev driver=raw,node-name=disk,file=disk_file
1399 -blockdev driver=raw,node-name=disk,file.driver=file,file.filename=disk.img
1401 ``Driver-specific options for qcow2``
1402 This is the image format block driver for qcow2 images. It is
1403 usually stacked on top of a protocol level block driver such as
1407 Reference to or definition of the data source block driver
1408 node (e.g. a ``file`` driver node)
1411 Reference to or definition of the backing file block device
1412 (default is taken from the image file). It is allowed to
1413 pass ``null`` here in order to disable the default backing
1417 Whether to enable the lazy refcounts feature (on/off;
1418 default is taken from the image file)
1421 The maximum total size of the L2 table and refcount block
1422 caches in bytes (default: the sum of l2-cache-size and
1423 refcount-cache-size)
1426 The maximum size of the L2 table cache in bytes (default: if
1427 cache-size is not specified - 32M on Linux platforms, and 8M
1428 on non-Linux platforms; otherwise, as large as possible
1429 within the cache-size, while permitting the requested or the
1430 minimal refcount cache size)
1432 ``refcount-cache-size``
1433 The maximum size of the refcount block cache in bytes
1434 (default: 4 times the cluster size; or if cache-size is
1435 specified, the part of it which is not used for the L2
1438 ``cache-clean-interval``
1439 Clean unused entries in the L2 and refcount caches. The
1440 interval is in seconds. The default value is 600 on
1441 supporting platforms, and 0 on other platforms. Setting it
1442 to 0 disables this feature.
1444 ``pass-discard-request``
1445 Whether discard requests to the qcow2 device should be
1446 forwarded to the data source (on/off; default: on if
1447 discard=unmap is specified, off otherwise)
1449 ``pass-discard-snapshot``
1450 Whether discard requests for the data source should be
1451 issued when a snapshot operation (e.g. deleting a snapshot)
1452 frees clusters in the qcow2 file (on/off; default: on)
1454 ``pass-discard-other``
1455 Whether discard requests for the data source should be
1456 issued on other occasions where a cluster gets freed
1457 (on/off; default: off)
1459 ``discard-no-unref``
1460 When enabled, data clusters will remain preallocated when they are
1461 no longer used, e.g. because they are discarded or converted to
1462 zero clusters. As usual, whether the old data is discarded or kept
1463 on the protocol level (i.e. in the image file) depends on the
1464 setting of the pass-discard-request option. Keeping the clusters
1465 preallocated prevents qcow2 fragmentation that would otherwise be
1466 caused by freeing and re-allocating them later. Besides potential
1467 performance degradation, such fragmentation can lead to increased
1468 allocation of clusters past the end of the image file,
1469 resulting in image files whose file length can grow much larger
1470 than their guest disk size would suggest.
1471 If image file length is of concern (e.g. when storing qcow2
1472 images directly on block devices), you should consider enabling
1476 Which overlap checks to perform for writes to the image
1477 (none/constant/cached/all; default: cached). For details or
1478 finer granularity control refer to the QAPI documentation of
1485 -blockdev driver=file,node-name=my_file,filename=/tmp/disk.qcow2
1486 -blockdev driver=qcow2,node-name=hda,file=my_file,overlap-check=none,cache-size=16777216
1492 -blockdev driver=qcow2,node-name=disk,file.driver=http,file.filename=http://example.com/image.qcow2
1494 ``Driver-specific options for other drivers``
1495 Please refer to the QAPI documentation of the ``blockdev-add``
1499 DEF("drive", HAS_ARG, QEMU_OPTION_drive,
1500 "-drive [file=file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n"
1501 " [,cache=writethrough|writeback|none|directsync|unsafe][,format=f]\n"
1502 " [,snapshot=on|off][,rerror=ignore|stop|report]\n"
1503 " [,werror=ignore|stop|report|enospc][,id=name]\n"
1504 " [,aio=threads|native|io_uring]\n"
1505 " [,readonly=on|off][,copy-on-read=on|off]\n"
1506 " [,discard=ignore|unmap][,detect-zeroes=on|off|unmap]\n"
1507 " [[,bps=b]|[[,bps_rd=r][,bps_wr=w]]]\n"
1508 " [[,iops=i]|[[,iops_rd=r][,iops_wr=w]]]\n"
1509 " [[,bps_max=bm]|[[,bps_rd_max=rm][,bps_wr_max=wm]]]\n"
1510 " [[,iops_max=im]|[[,iops_rd_max=irm][,iops_wr_max=iwm]]]\n"
1511 " [[,iops_size=is]]\n"
1513 " use 'file' as a drive image\n", QEMU_ARCH_ALL)
1515 ``-drive option[,option[,option[,...]]]``
1516 Define a new drive. This includes creating a block driver node (the
1517 backend) as well as a guest device, and is mostly a shortcut for
1518 defining the corresponding ``-blockdev`` and ``-device`` options.
1520 ``-drive`` accepts all options that are accepted by ``-blockdev``.
1521 In addition, it knows the following options:
1524 This option defines which disk image (see the :ref:`disk images`
1525 chapter in the System Emulation Users Guide) to use with this drive.
1526 If the filename contains comma, you must double it (for instance,
1527 "file=my,,file" to use file "my,file").
1529 Special files such as iSCSI devices can be specified using
1530 protocol specific URLs. See the section for "Device URL Syntax"
1531 for more information.
1534 This option defines on which type on interface the drive is
1535 connected. Available types are: ide, scsi, sd, mtd, floppy,
1536 pflash, virtio, none.
1538 ``bus=bus,unit=unit``
1539 These options define where is connected the drive by defining
1540 the bus number and the unit id.
1543 This option defines where the drive is connected by using an
1544 index in the list of available connectors of a given interface
1548 This option defines the type of the media: disk or cdrom.
1550 ``snapshot=snapshot``
1551 snapshot is "on" or "off" and controls snapshot mode for the
1552 given drive (see ``-snapshot``).
1555 cache is "none", "writeback", "unsafe", "directsync" or
1556 "writethrough" and controls how the host cache is used to access
1557 block data. This is a shortcut that sets the ``cache.direct``
1558 and ``cache.no-flush`` options (as in ``-blockdev``), and
1559 additionally ``cache.writeback``, which provides a default for
1560 the ``write-cache`` option of block guest devices (as in
1561 ``-device``). The modes correspond to the following settings:
1563 ============= =============== ============ ==============
1564 \ cache.writeback cache.direct cache.no-flush
1565 ============= =============== ============ ==============
1566 writeback on off off
1568 writethrough off off off
1569 directsync off on off
1571 ============= =============== ============ ==============
1573 The default mode is ``cache=writeback``.
1576 aio is "threads", "native", or "io_uring" and selects between pthread
1577 based disk I/O, native Linux AIO, or Linux io_uring API.
1580 Specify which disk format will be used rather than detecting the
1581 format. Can be used to specify format=raw to avoid interpreting
1582 an untrusted format header.
1584 ``werror=action,rerror=action``
1585 Specify which action to take on write and read errors. Valid
1586 actions are: "ignore" (ignore the error and try to continue),
1587 "stop" (pause QEMU), "report" (report the error to the guest),
1588 "enospc" (pause QEMU only if the host disk is full; report the
1589 error to the guest otherwise). The default setting is
1590 ``werror=enospc`` and ``rerror=report``.
1592 ``copy-on-read=copy-on-read``
1593 copy-on-read is "on" or "off" and enables whether to copy read
1594 backing file sectors into the image file.
1596 ``bps=b,bps_rd=r,bps_wr=w``
1597 Specify bandwidth throttling limits in bytes per second, either
1598 for all request types or for reads or writes only. Small values
1599 can lead to timeouts or hangs inside the guest. A safe minimum
1600 for disks is 2 MB/s.
1602 ``bps_max=bm,bps_rd_max=rm,bps_wr_max=wm``
1603 Specify bursts in bytes per second, either for all request types
1604 or for reads or writes only. Bursts allow the guest I/O to spike
1605 above the limit temporarily.
1607 ``iops=i,iops_rd=r,iops_wr=w``
1608 Specify request rate limits in requests per second, either for
1609 all request types or for reads or writes only.
1611 ``iops_max=bm,iops_rd_max=rm,iops_wr_max=wm``
1612 Specify bursts in requests per second, either for all request
1613 types or for reads or writes only. Bursts allow the guest I/O to
1614 spike above the limit temporarily.
1617 Let every is bytes of a request count as a new request for iops
1618 throttling purposes. Use this option to prevent guests from
1619 circumventing iops limits by sending fewer but larger requests.
1622 Join a throttling quota group with given name g. All drives that
1623 are members of the same group are accounted for together. Use
1624 this option to prevent guests from circumventing throttling
1625 limits by using many small disks instead of a single larger
1628 By default, the ``cache.writeback=on`` mode is used. It will report
1629 data writes as completed as soon as the data is present in the host
1630 page cache. This is safe as long as your guest OS makes sure to
1631 correctly flush disk caches where needed. If your guest OS does not
1632 handle volatile disk write caches correctly and your host crashes or
1633 loses power, then the guest may experience data corruption.
1635 For such guests, you should consider using ``cache.writeback=off``.
1636 This means that the host page cache will be used to read and write
1637 data, but write notification will be sent to the guest only after
1638 QEMU has made sure to flush each write to the disk. Be aware that
1639 this has a major impact on performance.
1641 When using the ``-snapshot`` option, unsafe caching is always used.
1643 Copy-on-read avoids accessing the same backing file sectors
1644 repeatedly and is useful when the backing file is over a slow
1645 network. By default copy-on-read is off.
1647 Instead of ``-cdrom`` you can use:
1651 |qemu_system| -drive file=file,index=2,media=cdrom
1653 Instead of ``-hda``, ``-hdb``, ``-hdc``, ``-hdd``, you can use:
1657 |qemu_system| -drive file=file,index=0,media=disk
1658 |qemu_system| -drive file=file,index=1,media=disk
1659 |qemu_system| -drive file=file,index=2,media=disk
1660 |qemu_system| -drive file=file,index=3,media=disk
1662 You can open an image using pre-opened file descriptors from an fd
1668 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\
1669 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\
1670 -drive file=/dev/fdset/2,index=0,media=disk
1672 You can connect a CDROM to the slave of ide0:
1676 |qemu_system_x86| -drive file=file,if=ide,index=1,media=cdrom
1678 If you don't specify the "file=" argument, you define an empty
1683 |qemu_system_x86| -drive if=ide,index=1,media=cdrom
1685 Instead of ``-fda``, ``-fdb``, you can use:
1689 |qemu_system_x86| -drive file=file,index=0,if=floppy
1690 |qemu_system_x86| -drive file=file,index=1,if=floppy
1692 By default, interface is "ide" and index is automatically
1697 |qemu_system_x86| -drive file=a -drive file=b
1699 is interpreted like:
1703 |qemu_system_x86| -hda a -hdb b
1706 DEF("mtdblock", HAS_ARG, QEMU_OPTION_mtdblock,
1707 "-mtdblock file use 'file' as on-board Flash memory image\n",
1711 Use file as on-board Flash memory image.
1714 DEF("sd", HAS_ARG, QEMU_OPTION_sd,
1715 "-sd file use 'file' as SecureDigital card image\n", QEMU_ARCH_ALL)
1718 Use file as SecureDigital card image.
1721 DEF("snapshot", 0, QEMU_OPTION_snapshot,
1722 "-snapshot write to temporary files instead of disk image files\n",
1726 Write to temporary files instead of disk image files. In this case,
1727 the raw disk image you use is not written back. You can however
1728 force the write back by pressing C-a s (see the :ref:`disk images`
1729 chapter in the System Emulation Users Guide).
1732 snapshot is incompatible with ``-blockdev`` (instead use qemu-img
1733 to manually create snapshot images to attach to your blockdev).
1734 If you have mixed ``-blockdev`` and ``-drive`` declarations you
1735 can use the 'snapshot' property on your drive declarations
1736 instead of this global option.
1740 DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev,
1741 "-fsdev local,id=id,path=path,security_model=mapped-xattr|mapped-file|passthrough|none\n"
1742 " [,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode]\n"
1743 " [[,throttling.bps-total=b]|[[,throttling.bps-read=r][,throttling.bps-write=w]]]\n"
1744 " [[,throttling.iops-total=i]|[[,throttling.iops-read=r][,throttling.iops-write=w]]]\n"
1745 " [[,throttling.bps-total-max=bm]|[[,throttling.bps-read-max=rm][,throttling.bps-write-max=wm]]]\n"
1746 " [[,throttling.iops-total-max=im]|[[,throttling.iops-read-max=irm][,throttling.iops-write-max=iwm]]]\n"
1747 " [[,throttling.iops-size=is]]\n"
1748 "-fsdev proxy,id=id,socket=socket[,writeout=immediate][,readonly=on]\n"
1749 "-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=immediate][,readonly=on]\n"
1750 "-fsdev synth,id=id\n",
1754 ``-fsdev local,id=id,path=path,security_model=security_model [,writeout=writeout][,readonly=on][,fmode=fmode][,dmode=dmode] [,throttling.option=value[,throttling.option=value[,...]]]``
1756 ``-fsdev proxy,id=id,socket=socket[,writeout=writeout][,readonly=on]``
1758 ``-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=writeout][,readonly=on]``
1760 ``-fsdev synth,id=id[,readonly=on]``
1761 Define a new file system device. Valid options are:
1764 Accesses to the filesystem are done by QEMU.
1767 Accesses to the filesystem are done by virtfs-proxy-helper(1). This
1768 option is deprecated (since QEMU 8.1) and will be removed in a future
1769 version of QEMU. Use ``local`` instead.
1772 Synthetic filesystem, only used by QTests.
1775 Specifies identifier for this device.
1778 Specifies the export path for the file system device. Files
1779 under this path will be available to the 9p client on the guest.
1781 ``security_model=security_model``
1782 Specifies the security model to be used for this export path.
1783 Supported security models are "passthrough", "mapped-xattr",
1784 "mapped-file" and "none". In "passthrough" security model, files
1785 are stored using the same credentials as they are created on the
1786 guest. This requires QEMU to run as root. In "mapped-xattr"
1787 security model, some of the file attributes like uid, gid, mode
1788 bits and link target are stored as file attributes. For
1789 "mapped-file" these attributes are stored in the hidden
1790 .virtfs\_metadata directory. Directories exported by this
1791 security model cannot interact with other unix tools. "none"
1792 security model is same as passthrough except the sever won't
1793 report failures if it fails to set file attributes like
1794 ownership. Security model is mandatory only for local fsdriver.
1795 Other fsdrivers (like proxy) don't take security model as a
1798 ``writeout=writeout``
1799 This is an optional argument. The only supported value is
1800 "immediate". This means that host page cache will be used to
1801 read and write data but write notification will be sent to the
1802 guest only when the data has been reported as written by the
1806 Enables exporting 9p share as a readonly mount for guests. By
1807 default read-write access is given.
1810 Enables proxy filesystem driver to use passed socket file for
1811 communicating with virtfs-proxy-helper(1).
1814 Enables proxy filesystem driver to use passed socket descriptor
1815 for communicating with virtfs-proxy-helper(1). Usually a helper
1816 like libvirt will create socketpair and pass one of the fds as
1820 Specifies the default mode for newly created files on the host.
1821 Works only with security models "mapped-xattr" and
1825 Specifies the default mode for newly created directories on the
1826 host. Works only with security models "mapped-xattr" and
1829 ``throttling.bps-total=b,throttling.bps-read=r,throttling.bps-write=w``
1830 Specify bandwidth throttling limits in bytes per second, either
1831 for all request types or for reads or writes only.
1833 ``throttling.bps-total-max=bm,bps-read-max=rm,bps-write-max=wm``
1834 Specify bursts in bytes per second, either for all request types
1835 or for reads or writes only. Bursts allow the guest I/O to spike
1836 above the limit temporarily.
1838 ``throttling.iops-total=i,throttling.iops-read=r, throttling.iops-write=w``
1839 Specify request rate limits in requests per second, either for
1840 all request types or for reads or writes only.
1842 ``throttling.iops-total-max=im,throttling.iops-read-max=irm, throttling.iops-write-max=iwm``
1843 Specify bursts in requests per second, either for all request
1844 types or for reads or writes only. Bursts allow the guest I/O to
1845 spike above the limit temporarily.
1847 ``throttling.iops-size=is``
1848 Let every is bytes of a request count as a new request for iops
1849 throttling purposes.
1851 -fsdev option is used along with -device driver "virtio-9p-...".
1853 ``-device virtio-9p-type,fsdev=id,mount_tag=mount_tag``
1854 Options for virtio-9p-... driver are:
1857 Specifies the variant to be used. Supported values are "pci",
1858 "ccw" or "device", depending on the machine type.
1861 Specifies the id value specified along with -fsdev option.
1863 ``mount_tag=mount_tag``
1864 Specifies the tag name to be used by the guest to mount this
1868 DEF("virtfs", HAS_ARG, QEMU_OPTION_virtfs,
1869 "-virtfs local,path=path,mount_tag=tag,security_model=mapped-xattr|mapped-file|passthrough|none\n"
1870 " [,id=id][,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode][,multidevs=remap|forbid|warn]\n"
1871 "-virtfs proxy,mount_tag=tag,socket=socket[,id=id][,writeout=immediate][,readonly=on]\n"
1872 "-virtfs proxy,mount_tag=tag,sock_fd=sock_fd[,id=id][,writeout=immediate][,readonly=on]\n"
1873 "-virtfs synth,mount_tag=tag[,id=id][,readonly=on]\n",
1877 ``-virtfs local,path=path,mount_tag=mount_tag ,security_model=security_model[,writeout=writeout][,readonly=on] [,fmode=fmode][,dmode=dmode][,multidevs=multidevs]``
1879 ``-virtfs proxy,socket=socket,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1881 ``-virtfs proxy,sock_fd=sock_fd,mount_tag=mount_tag [,writeout=writeout][,readonly=on]``
1883 ``-virtfs synth,mount_tag=mount_tag``
1884 Define a new virtual filesystem device and expose it to the guest using
1885 a virtio-9p-device (a.k.a. 9pfs), which essentially means that a certain
1886 directory on host is made directly accessible by guest as a pass-through
1887 file system by using the 9P network protocol for communication between
1888 host and guests, if desired even accessible, shared by several guests
1891 Note that ``-virtfs`` is actually just a convenience shortcut for its
1892 generalized form ``-fsdev -device virtio-9p-pci``.
1894 The general form of pass-through file system options are:
1897 Accesses to the filesystem are done by QEMU.
1900 Accesses to the filesystem are done by virtfs-proxy-helper(1).
1901 This option is deprecated (since QEMU 8.1) and will be removed in a
1902 future version of QEMU. Use ``local`` instead.
1905 Synthetic filesystem, only used by QTests.
1908 Specifies identifier for the filesystem device
1911 Specifies the export path for the file system device. Files
1912 under this path will be available to the 9p client on the guest.
1914 ``security_model=security_model``
1915 Specifies the security model to be used for this export path.
1916 Supported security models are "passthrough", "mapped-xattr",
1917 "mapped-file" and "none". In "passthrough" security model, files
1918 are stored using the same credentials as they are created on the
1919 guest. This requires QEMU to run as root. In "mapped-xattr"
1920 security model, some of the file attributes like uid, gid, mode
1921 bits and link target are stored as file attributes. For
1922 "mapped-file" these attributes are stored in the hidden
1923 .virtfs\_metadata directory. Directories exported by this
1924 security model cannot interact with other unix tools. "none"
1925 security model is same as passthrough except the sever won't
1926 report failures if it fails to set file attributes like
1927 ownership. Security model is mandatory only for local fsdriver.
1928 Other fsdrivers (like proxy) don't take security model as a
1931 ``writeout=writeout``
1932 This is an optional argument. The only supported value is
1933 "immediate". This means that host page cache will be used to
1934 read and write data but write notification will be sent to the
1935 guest only when the data has been reported as written by the
1939 Enables exporting 9p share as a readonly mount for guests. By
1940 default read-write access is given.
1943 Enables proxy filesystem driver to use passed socket file for
1944 communicating with virtfs-proxy-helper(1). Usually a helper like
1945 libvirt will create socketpair and pass one of the fds as
1949 Enables proxy filesystem driver to use passed 'sock\_fd' as the
1950 socket descriptor for interfacing with virtfs-proxy-helper(1).
1953 Specifies the default mode for newly created files on the host.
1954 Works only with security models "mapped-xattr" and
1958 Specifies the default mode for newly created directories on the
1959 host. Works only with security models "mapped-xattr" and
1962 ``mount_tag=mount_tag``
1963 Specifies the tag name to be used by the guest to mount this
1966 ``multidevs=multidevs``
1967 Specifies how to deal with multiple devices being shared with a
1968 9p export. Supported behaviours are either "remap", "forbid" or
1969 "warn". The latter is the default behaviour on which virtfs 9p
1970 expects only one device to be shared with the same export, and
1971 if more than one device is shared and accessed via the same 9p
1972 export then only a warning message is logged (once) by qemu on
1973 host side. In order to avoid file ID collisions on guest you
1974 should either create a separate virtfs export for each device to
1975 be shared with guests (recommended way) or you might use "remap"
1976 instead which allows you to share multiple devices with only one
1977 export instead, which is achieved by remapping the original
1978 inode numbers from host to guest in a way that would prevent
1979 such collisions. Remapping inodes in such use cases is required
1980 because the original device IDs from host are never passed and
1981 exposed on guest. Instead all files of an export shared with
1982 virtfs always share the same device id on guest. So two files
1983 with identical inode numbers but from actually different devices
1984 on host would otherwise cause a file ID collision and hence
1985 potential misbehaviours on guest. "forbid" on the other hand
1986 assumes like "warn" that only one device is shared by the same
1987 export, however it will not only log a warning message but also
1988 deny access to additional devices on guest. Note though that
1989 "forbid" does currently not block all possible file access
1990 operations (e.g. readdir() would still return entries from other
1994 DEF("iscsi", HAS_ARG, QEMU_OPTION_iscsi,
1995 "-iscsi [user=user][,password=password][,password-secret=secret-id]\n"
1996 " [,header-digest=CRC32C|CR32C-NONE|NONE-CRC32C|NONE]\n"
1997 " [,initiator-name=initiator-iqn][,id=target-iqn]\n"
1998 " [,timeout=timeout]\n"
1999 " iSCSI session parameters\n", QEMU_ARCH_ALL)
2003 Configure iSCSI session parameters.
2008 DEFHEADING(USB convenience options:)
2010 DEF("usb", 0, QEMU_OPTION_usb,
2011 "-usb enable on-board USB host controller (if not enabled by default)\n",
2015 Enable USB emulation on machine types with an on-board USB host
2016 controller (if not enabled by default). Note that on-board USB host
2017 controllers may not support USB 3.0. In this case
2018 ``-device qemu-xhci`` can be used instead on machines with PCI.
2021 DEF("usbdevice", HAS_ARG, QEMU_OPTION_usbdevice,
2022 "-usbdevice name add the host or guest USB device 'name'\n",
2025 ``-usbdevice devname``
2026 Add the USB device devname, and enable an on-board USB controller
2027 if possible and necessary (just like it can be done via
2028 ``-machine usb=on``). Note that this option is mainly intended for
2029 the user's convenience only. More fine-grained control can be
2030 achieved by selecting a USB host controller (if necessary) and the
2031 desired USB device via the ``-device`` option instead. For example,
2032 instead of using ``-usbdevice mouse`` it is possible to use
2033 ``-device qemu-xhci -device usb-mouse`` to connect the USB mouse
2034 to a USB 3.0 controller instead (at least on machines that support
2035 PCI and do not have an USB controller enabled by default yet).
2036 For more details, see the chapter about
2037 :ref:`Connecting USB devices` in the System Emulation Users Guide.
2038 Possible devices for devname are:
2041 Braille device. This will use BrlAPI to display the braille
2042 output on a real or fake device (i.e. it also creates a
2043 corresponding ``braille`` chardev automatically beside the
2044 ``usb-braille`` USB device).
2047 Standard USB keyboard. Will override the PS/2 keyboard (if present).
2050 Virtual Mouse. This will override the PS/2 mouse emulation when
2054 Pointer device that uses absolute coordinates (like a
2055 touchscreen). This means QEMU is able to report the mouse
2056 position without having to grab the mouse. Also overrides the
2057 PS/2 mouse emulation when activated.
2060 Wacom PenPartner USB tablet.
2067 DEFHEADING(Display options:)
2069 DEF("display", HAS_ARG, QEMU_OPTION_display,
2070 #if defined(CONFIG_SPICE)
2071 "-display spice-app[,gl=on|off]\n"
2073 #if defined(CONFIG_SDL)
2074 "-display sdl[,gl=on|core|es|off][,grab-mod=<mod>][,show-cursor=on|off]\n"
2075 " [,window-close=on|off]\n"
2077 #if defined(CONFIG_GTK)
2078 "-display gtk[,full-screen=on|off][,gl=on|off][,grab-on-hover=on|off]\n"
2079 " [,show-tabs=on|off][,show-cursor=on|off][,window-close=on|off]\n"
2080 " [,show-menubar=on|off]\n"
2082 #if defined(CONFIG_VNC)
2083 "-display vnc=<display>[,<optargs>]\n"
2085 #if defined(CONFIG_CURSES)
2086 "-display curses[,charset=<encoding>]\n"
2088 #if defined(CONFIG_COCOA)
2089 "-display cocoa[,full-grab=on|off][,swap-opt-cmd=on|off]\n"
2091 #if defined(CONFIG_OPENGL)
2092 "-display egl-headless[,rendernode=<file>]\n"
2094 #if defined(CONFIG_DBUS_DISPLAY)
2095 "-display dbus[,addr=<dbusaddr>]\n"
2096 " [,gl=on|core|es|off][,rendernode=<file>]\n"
2098 #if defined(CONFIG_COCOA)
2099 "-display cocoa[,show-cursor=on|off][,left-command-key=on|off]\n"
2102 " select display backend type\n"
2103 " The default display is equivalent to\n "
2104 #if defined(CONFIG_GTK)
2105 "\"-display gtk\"\n"
2106 #elif defined(CONFIG_SDL)
2107 "\"-display sdl\"\n"
2108 #elif defined(CONFIG_COCOA)
2109 "\"-display cocoa\"\n"
2110 #elif defined(CONFIG_VNC)
2111 "\"-vnc localhost:0,to=99,id=default\"\n"
2113 "\"-display none\"\n"
2118 Select type of display to use. Use ``-display help`` to list the available
2119 display types. Valid values for type are
2121 ``spice-app[,gl=on|off]``
2122 Start QEMU as a Spice server and launch the default Spice client
2123 application. The Spice server will redirect the serial consoles
2124 and QEMU monitors. (Since 4.0)
2127 Export the display over D-Bus interfaces. (Since 7.0)
2129 The connection is registered with the "org.qemu" name (and queued when
2132 ``addr=<dbusaddr>`` : D-Bus bus address to connect to.
2134 ``p2p=yes|no`` : Use peer-to-peer connection, accepted via QMP ``add_client``.
2136 ``gl=on|off|core|es`` : Use OpenGL for rendering (the D-Bus interface
2137 will share framebuffers with DMABUF file descriptors).
2140 Display video output via SDL (usually in a separate graphics
2141 window; see the SDL documentation for other possibilities).
2142 Valid parameters are:
2144 ``grab-mod=<mods>`` : Used to select the modifier keys for toggling
2145 the mouse grabbing in conjunction with the "g" key. ``<mods>`` can be
2146 either ``lshift-lctrl-lalt`` or ``rctrl``.
2148 ``gl=on|off|core|es`` : Use OpenGL for displaying
2150 ``show-cursor=on|off`` : Force showing the mouse cursor
2152 ``window-close=on|off`` : Allow to quit qemu with window close button
2155 Display video output in a GTK window. This interface provides
2156 drop-down menus and other UI elements to configure and control
2157 the VM during runtime. Valid parameters are:
2159 ``full-screen=on|off`` : Start in fullscreen mode
2161 ``gl=on|off`` : Use OpenGL for displaying
2163 ``grab-on-hover=on|off`` : Grab keyboard input on mouse hover
2165 ``show-tabs=on|off`` : Display the tab bar for switching between the
2166 various graphical interfaces (e.g. VGA and
2167 virtual console character devices) by default.
2169 ``show-cursor=on|off`` : Force showing the mouse cursor
2171 ``window-close=on|off`` : Allow to quit qemu with window close button
2173 ``show-menubar=on|off`` : Display the main window menubar, defaults to "on"
2175 ``zoom-to-fit=on|off`` : Expand video output to the window size,
2178 ``curses[,charset=<encoding>]``
2179 Display video output via curses. For graphics device models
2180 which support a text mode, QEMU can display this output using a
2181 curses/ncurses interface. Nothing is displayed when the graphics
2182 device is in graphical mode or if the graphics device does not
2183 support a text mode. Generally only the VGA device models
2184 support text mode. The font charset used by the guest can be
2185 specified with the ``charset`` option, for example
2186 ``charset=CP850`` for IBM CP850 encoding. The default is
2190 Display video output in a Cocoa window. Mac only. This interface
2191 provides drop-down menus and other UI elements to configure and
2192 control the VM during runtime. Valid parameters are:
2194 ``show-cursor=on|off`` : Force showing the mouse cursor
2196 ``left-command-key=on|off`` : Disable forwarding left command key to host
2198 ``egl-headless[,rendernode=<file>]``
2199 Offload all OpenGL operations to a local DRI device. For any
2200 graphical display, this display needs to be paired with either
2201 VNC or SPICE displays.
2204 Start a VNC server on display <display>
2207 Do not display video output. The guest will still see an
2208 emulated graphics card, but its output will not be displayed to
2209 the QEMU user. This option differs from the -nographic option in
2210 that it only affects what is done with video output; -nographic
2211 also changes the destination of the serial and parallel port
2215 DEF("nographic", 0, QEMU_OPTION_nographic,
2216 "-nographic disable graphical output and redirect serial I/Os to console\n",
2220 Normally, if QEMU is compiled with graphical window support, it
2221 displays output such as guest graphics, guest console, and the QEMU
2222 monitor in a window. With this option, you can totally disable
2223 graphical output so that QEMU is a simple command line application.
2224 The emulated serial port is redirected on the console and muxed with
2225 the monitor (unless redirected elsewhere explicitly). Therefore, you
2226 can still use QEMU to debug a Linux kernel with a serial console.
2227 Use C-a h for help on switching between the console and monitor.
2231 DEF("spice", HAS_ARG, QEMU_OPTION_spice,
2232 "-spice [port=port][,tls-port=secured-port][,x509-dir=<dir>]\n"
2233 " [,x509-key-file=<file>][,x509-key-password=<file>]\n"
2234 " [,x509-cert-file=<file>][,x509-cacert-file=<file>]\n"
2235 " [,x509-dh-key-file=<file>][,addr=addr]\n"
2236 " [,ipv4=on|off][,ipv6=on|off][,unix=on|off]\n"
2237 " [,tls-ciphers=<list>]\n"
2238 " [,tls-channel=[main|display|cursor|inputs|record|playback]]\n"
2239 " [,plaintext-channel=[main|display|cursor|inputs|record|playback]]\n"
2240 " [,sasl=on|off][,disable-ticketing=on|off]\n"
2241 " [,password-secret=<secret-id>]\n"
2242 " [,image-compression=[auto_glz|auto_lz|quic|glz|lz|off]]\n"
2243 " [,jpeg-wan-compression=[auto|never|always]]\n"
2244 " [,zlib-glz-wan-compression=[auto|never|always]]\n"
2245 " [,streaming-video=[off|all|filter]][,disable-copy-paste=on|off]\n"
2246 " [,disable-agent-file-xfer=on|off][,agent-mouse=[on|off]]\n"
2247 " [,playback-compression=[on|off]][,seamless-migration=[on|off]]\n"
2248 " [,gl=[on|off]][,rendernode=<file>]\n"
2250 " at least one of {port, tls-port} is mandatory\n",
2254 ``-spice option[,option[,...]]``
2255 Enable the spice remote desktop protocol. Valid options are
2258 Set the TCP port spice is listening on for plaintext channels.
2261 Set the IP address spice is listening on. Default is any
2264 ``ipv4=on|off``; \ ``ipv6=on|off``; \ ``unix=on|off``
2265 Force using the specified IP version.
2267 ``password-secret=<secret-id>``
2268 Set the ID of the ``secret`` object containing the password
2269 you need to authenticate.
2272 Require that the client use SASL to authenticate with the spice.
2273 The exact choice of authentication method used is controlled
2274 from the system / user's SASL configuration file for the 'qemu'
2275 service. This is typically found in /etc/sasl2/qemu.conf. If
2276 running QEMU as an unprivileged user, an environment variable
2277 SASL\_CONF\_PATH can be used to make it search alternate
2278 locations for the service config. While some SASL auth methods
2279 can also provide data encryption (eg GSSAPI), it is recommended
2280 that SASL always be combined with the 'tls' and 'x509' settings
2281 to enable use of SSL and server certificates. This ensures a
2282 data encryption preventing compromise of authentication
2285 ``disable-ticketing=on|off``
2286 Allow client connects without authentication.
2288 ``disable-copy-paste=on|off``
2289 Disable copy paste between the client and the guest.
2291 ``disable-agent-file-xfer=on|off``
2292 Disable spice-vdagent based file-xfer between the client and the
2296 Set the TCP port spice is listening on for encrypted channels.
2299 Set the x509 file directory. Expects same filenames as -vnc
2302 ``x509-key-file=<file>``; \ ``x509-key-password=<file>``; \ ``x509-cert-file=<file>``; \ ``x509-cacert-file=<file>``; \ ``x509-dh-key-file=<file>``
2303 The x509 file names can also be configured individually.
2305 ``tls-ciphers=<list>``
2306 Specify which ciphers to use.
2308 ``tls-channel=[main|display|cursor|inputs|record|playback]``; \ ``plaintext-channel=[main|display|cursor|inputs|record|playback]``
2309 Force specific channel to be used with or without TLS
2310 encryption. The options can be specified multiple times to
2311 configure multiple channels. The special name "default" can be
2312 used to set the default mode. For channels which are not
2313 explicitly forced into one mode the spice client is allowed to
2314 pick tls/plaintext as he pleases.
2316 ``image-compression=[auto_glz|auto_lz|quic|glz|lz|off]``
2317 Configure image compression (lossless). Default is auto\_glz.
2319 ``jpeg-wan-compression=[auto|never|always]``; \ ``zlib-glz-wan-compression=[auto|never|always]``
2320 Configure wan image compression (lossy for slow links). Default
2323 ``streaming-video=[off|all|filter]``
2324 Configure video stream detection. Default is off.
2326 ``agent-mouse=[on|off]``
2327 Enable/disable passing mouse events via vdagent. Default is on.
2329 ``playback-compression=[on|off]``
2330 Enable/disable audio stream compression (using celt 0.5.1).
2333 ``seamless-migration=[on|off]``
2334 Enable/disable spice seamless migration. Default is off.
2337 Enable/disable OpenGL context. Default is off.
2339 ``rendernode=<file>``
2340 DRM render node for OpenGL rendering. If not specified, it will
2341 pick the first available. (Since 2.9)
2344 DEF("portrait", 0, QEMU_OPTION_portrait,
2345 "-portrait rotate graphical output 90 deg left (only PXA LCD)\n",
2349 Rotate graphical output 90 deg left (only PXA LCD).
2352 DEF("rotate", HAS_ARG, QEMU_OPTION_rotate,
2353 "-rotate <deg> rotate graphical output some deg left (only PXA LCD)\n",
2357 Rotate graphical output some deg left (only PXA LCD).
2360 DEF("vga", HAS_ARG, QEMU_OPTION_vga,
2361 "-vga [std|cirrus|vmware|qxl|xenfb|tcx|cg3|virtio|none]\n"
2362 " select video card type\n", QEMU_ARCH_ALL)
2365 Select type of VGA card to emulate. Valid values for type are
2368 Cirrus Logic GD5446 Video card. All Windows versions starting
2369 from Windows 95 should recognize and use this graphic card. For
2370 optimal performances, use 16 bit color depth in the guest and
2371 the host OS. (This card was the default before QEMU 2.2)
2374 Standard VGA card with Bochs VBE extensions. If your guest OS
2375 supports the VESA 2.0 VBE extensions (e.g. Windows XP) and if
2376 you want to use high resolution modes (>= 1280x1024x16) then you
2377 should use this option. (This card is the default since QEMU
2381 VMWare SVGA-II compatible adapter. Use it if you have
2382 sufficiently recent XFree86/XOrg server or Windows guest with a
2383 driver for this card.
2386 QXL paravirtual graphic card. It is VGA compatible (including
2387 VESA 2.0 VBE support). Works best with qxl guest drivers
2388 installed though. Recommended choice when using the spice
2392 (sun4m only) Sun TCX framebuffer. This is the default
2393 framebuffer for sun4m machines and offers both 8-bit and 24-bit
2394 colour depths at a fixed resolution of 1024x768.
2397 (sun4m only) Sun cgthree framebuffer. This is a simple 8-bit
2398 framebuffer for sun4m machines available in both 1024x768
2399 (OpenBIOS) and 1152x900 (OBP) resolutions aimed at people
2400 wishing to run older Solaris versions.
2409 DEF("full-screen", 0, QEMU_OPTION_full_screen,
2410 "-full-screen start in full screen\n", QEMU_ARCH_ALL)
2413 Start in full screen.
2416 DEF("g", HAS_ARG, QEMU_OPTION_g ,
2417 "-g WxH[xDEPTH] Set the initial graphical resolution and depth\n",
2418 QEMU_ARCH_PPC | QEMU_ARCH_SPARC | QEMU_ARCH_M68K)
2420 ``-g`` *width*\ ``x``\ *height*\ ``[x``\ *depth*\ ``]``
2421 Set the initial graphical resolution and depth (PPC, SPARC only).
2423 For PPC the default is 800x600x32.
2425 For SPARC with the TCX graphics device, the default is 1024x768x8
2426 with the option of 1024x768x24. For cgthree, the default is
2427 1024x768x8 with the option of 1152x900x8 for people who wish to use
2432 DEF("vnc", HAS_ARG, QEMU_OPTION_vnc ,
2433 "-vnc <display> shorthand for -display vnc=<display>\n", QEMU_ARCH_ALL)
2436 ``-vnc display[,option[,option[,...]]]``
2437 Normally, if QEMU is compiled with graphical window support, it
2438 displays output such as guest graphics, guest console, and the QEMU
2439 monitor in a window. With this option, you can have QEMU listen on
2440 VNC display display and redirect the VGA display over the VNC
2441 session. It is very useful to enable the usb tablet device when
2442 using this option (option ``-device usb-tablet``). When using the
2443 VNC display, you must use the ``-k`` parameter to set the keyboard
2444 layout if you are not using en-us. Valid syntax for the display is
2447 With this option, QEMU will try next available VNC displays,
2448 until the number L, if the origianlly defined "-vnc display" is
2449 not available, e.g. port 5900+display is already used by another
2450 application. By default, to=0.
2453 TCP connections will only be allowed from host on display d. By
2454 convention the TCP port is 5900+d. Optionally, host can be
2455 omitted in which case the server will accept connections from
2459 Connections will be allowed over UNIX domain sockets where path
2460 is the location of a unix socket to listen for connections on.
2463 VNC is initialized but not started. The monitor ``change``
2464 command can be used to later start the VNC server.
2466 Following the display value there may be one or more option flags
2467 separated by commas. Valid options are
2470 Connect to a listening VNC client via a "reverse" connection.
2471 The client is specified by the display. For reverse network
2472 connections (host:d,``reverse``), the d argument is a TCP port
2473 number, not a display number.
2475 ``websocket=on|off``
2476 Opens an additional TCP listening port dedicated to VNC
2477 Websocket connections. If a bare websocket option is given, the
2478 Websocket port is 5700+display. An alternative port can be
2479 specified with the syntax ``websocket``\ =port.
2481 If host is specified connections will only be allowed from this
2482 host. It is possible to control the websocket listen address
2483 independently, using the syntax ``websocket``\ =host:port.
2485 If no TLS credentials are provided, the websocket connection
2486 runs in unencrypted mode. If TLS credentials are provided, the
2487 websocket connection requires encrypted client connections.
2490 Require that password based authentication is used for client
2493 The password must be set separately using the ``set_password``
2494 command in the :ref:`QEMU monitor`. The
2495 syntax to change your password is:
2496 ``set_password <protocol> <password>`` where <protocol> could be
2497 either "vnc" or "spice".
2499 If you would like to change <protocol> password expiration, you
2500 should use ``expire_password <protocol> <expiration-time>``
2501 where expiration time could be one of the following options:
2502 now, never, +seconds or UNIX time of expiration, e.g. +60 to
2503 make password expire in 60 seconds, or 1335196800 to make
2504 password expire on "Mon Apr 23 12:00:00 EDT 2012" (UNIX time for
2505 this date and time).
2507 You can also use keywords "now" or "never" for the expiration
2508 time to allow <protocol> password to expire immediately or never
2511 ``password-secret=<secret-id>``
2512 Require that password based authentication is used for client
2513 connections, using the password provided by the ``secret``
2514 object identified by ``secret-id``.
2517 Provides the ID of a set of TLS credentials to use to secure the
2518 VNC server. They will apply to both the normal VNC server socket
2519 and the websocket socket (if enabled). Setting TLS credentials
2520 will cause the VNC server socket to enable the VeNCrypt auth
2521 mechanism. The credentials should have been previously created
2522 using the ``-object tls-creds`` argument.
2525 Provides the ID of the QAuthZ authorization object against which
2526 the client's x509 distinguished name will validated. This object
2527 is only resolved at time of use, so can be deleted and recreated
2528 on the fly while the VNC server is active. If missing, it will
2529 default to denying access.
2532 Require that the client use SASL to authenticate with the VNC
2533 server. The exact choice of authentication method used is
2534 controlled from the system / user's SASL configuration file for
2535 the 'qemu' service. This is typically found in
2536 /etc/sasl2/qemu.conf. If running QEMU as an unprivileged user,
2537 an environment variable SASL\_CONF\_PATH can be used to make it
2538 search alternate locations for the service config. While some
2539 SASL auth methods can also provide data encryption (eg GSSAPI),
2540 it is recommended that SASL always be combined with the 'tls'
2541 and 'x509' settings to enable use of SSL and server
2542 certificates. This ensures a data encryption preventing
2543 compromise of authentication credentials. See the
2544 :ref:`VNC security` section in the System Emulation Users Guide
2545 for details on using SASL authentication.
2548 Provides the ID of the QAuthZ authorization object against which
2549 the client's SASL username will validated. This object is only
2550 resolved at time of use, so can be deleted and recreated on the
2551 fly while the VNC server is active. If missing, it will default
2555 Legacy method for enabling authorization of clients against the
2556 x509 distinguished name and SASL username. It results in the
2557 creation of two ``authz-list`` objects with IDs of
2558 ``vnc.username`` and ``vnc.x509dname``. The rules for these
2559 objects must be configured with the HMP ACL commands.
2561 This option is deprecated and should no longer be used. The new
2562 ``sasl-authz`` and ``tls-authz`` options are a replacement.
2565 Enable lossy compression methods (gradient, JPEG, ...). If this
2566 option is set, VNC client may receive lossy framebuffer updates
2567 depending on its encoding settings. Enabling this option can
2568 save a lot of bandwidth at the expense of quality.
2570 ``non-adaptive=on|off``
2571 Disable adaptive encodings. Adaptive encodings are enabled by
2572 default. An adaptive encoding will try to detect frequently
2573 updated screen regions, and send updates in these regions using
2574 a lossy encoding (like JPEG). This can be really helpful to save
2575 bandwidth when playing videos. Disabling adaptive encodings
2576 restores the original static behavior of encodings like Tight.
2578 ``share=[allow-exclusive|force-shared|ignore]``
2579 Set display sharing policy. 'allow-exclusive' allows clients to
2580 ask for exclusive access. As suggested by the rfb spec this is
2581 implemented by dropping other connections. Connecting multiple
2582 clients in parallel requires all clients asking for a shared
2583 session (vncviewer: -shared switch). This is the default.
2584 'force-shared' disables exclusive client access. Useful for
2585 shared desktop sessions, where you don't want someone forgetting
2586 specify -shared disconnect everybody else. 'ignore' completely
2587 ignores the shared flag and allows everybody connect
2588 unconditionally. Doesn't conform to the rfb spec but is
2589 traditional QEMU behavior.
2592 Set keyboard delay, for key down and key up events, in
2593 milliseconds. Default is 10. Keyboards are low-bandwidth
2594 devices, so this slowdown can help the device and guest to keep
2595 up and not lose events in case events are arriving in bulk.
2596 Possible causes for the latter are flaky network connections, or
2597 scripts for automated testing.
2599 ``audiodev=audiodev``
2600 Use the specified audiodev when the VNC client requests audio
2601 transmission. When not using an -audiodev argument, this option
2602 must be omitted, otherwise is must be present and specify a
2605 ``power-control=on|off``
2606 Permit the remote client to issue shutdown, reboot or reset power
2610 ARCHHEADING(, QEMU_ARCH_I386)
2612 ARCHHEADING(i386 target only:, QEMU_ARCH_I386)
2614 DEF("win2k-hack", 0, QEMU_OPTION_win2k_hack,
2615 "-win2k-hack use it when installing Windows 2000 to avoid a disk full bug\n",
2619 Use it when installing Windows 2000 to avoid a disk full bug. After
2620 Windows 2000 is installed, you no longer need this option (this
2621 option slows down the IDE transfers).
2624 DEF("no-fd-bootchk", 0, QEMU_OPTION_no_fd_bootchk,
2625 "-no-fd-bootchk disable boot signature checking for floppy disks\n",
2629 Disable boot signature checking for floppy disks in BIOS. May be
2630 needed to boot from old floppy disks.
2633 DEF("no-acpi", 0, QEMU_OPTION_no_acpi,
2634 "-no-acpi disable ACPI\n", QEMU_ARCH_I386 | QEMU_ARCH_ARM)
2637 Disable ACPI (Advanced Configuration and Power Interface) support.
2638 Use it if your guest OS complains about ACPI problems (PC target
2642 DEF("no-hpet", 0, QEMU_OPTION_no_hpet,
2643 "-no-hpet disable HPET\n", QEMU_ARCH_I386)
2646 Disable HPET support. Deprecated, use '-machine hpet=off' instead.
2649 DEF("acpitable", HAS_ARG, QEMU_OPTION_acpitable,
2650 "-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"
2651 " ACPI table description\n", QEMU_ARCH_I386)
2653 ``-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]...]``
2654 Add ACPI table with specified header fields and context from
2655 specified files. For file=, take whole ACPI table from the specified
2656 files, including all ACPI headers (possible overridden by other
2657 options). For data=, only data portion of the table is used, all
2658 header information is specified in the command line. If a SLIC table
2659 is supplied to QEMU, then the SLIC's oem\_id and oem\_table\_id
2660 fields will override the same in the RSDT and the FADT (a.k.a.
2661 FACP), in order to ensure the field matches required by the
2662 Microsoft SLIC spec and the ACPI spec.
2665 DEF("smbios", HAS_ARG, QEMU_OPTION_smbios,
2666 "-smbios file=binary\n"
2667 " load SMBIOS entry from binary file\n"
2668 "-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d]\n"
2670 " specify SMBIOS type 0 fields\n"
2671 "-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str]\n"
2672 " [,uuid=uuid][,sku=str][,family=str]\n"
2673 " specify SMBIOS type 1 fields\n"
2674 "-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str]\n"
2675 " [,asset=str][,location=str]\n"
2676 " specify SMBIOS type 2 fields\n"
2677 "-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str]\n"
2679 " specify SMBIOS type 3 fields\n"
2680 "-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str]\n"
2681 " [,asset=str][,part=str][,max-speed=%d][,current-speed=%d]\n"
2682 " [,processor-id=%d]\n"
2683 " specify SMBIOS type 4 fields\n"
2684 "-smbios type=8[,external_reference=str][,internal_reference=str][,connector_type=%d][,port_type=%d]\n"
2685 " specify SMBIOS type 8 fields\n"
2686 "-smbios type=11[,value=str][,path=filename]\n"
2687 " specify SMBIOS type 11 fields\n"
2688 "-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str]\n"
2689 " [,asset=str][,part=str][,speed=%d]\n"
2690 " specify SMBIOS type 17 fields\n"
2691 "-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]\n"
2692 " specify SMBIOS type 41 fields\n",
2693 QEMU_ARCH_I386 | QEMU_ARCH_ARM | QEMU_ARCH_LOONGARCH)
2695 ``-smbios file=binary``
2696 Load SMBIOS entry from binary file.
2698 ``-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d][,uefi=on|off]``
2699 Specify SMBIOS type 0 fields
2701 ``-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str][,uuid=uuid][,sku=str][,family=str]``
2702 Specify SMBIOS type 1 fields
2704 ``-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str][,asset=str][,location=str]``
2705 Specify SMBIOS type 2 fields
2707 ``-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str][,sku=str]``
2708 Specify SMBIOS type 3 fields
2710 ``-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str][,asset=str][,part=str][,processor-id=%d]``
2711 Specify SMBIOS type 4 fields
2713 ``-smbios type=11[,value=str][,path=filename]``
2714 Specify SMBIOS type 11 fields
2716 This argument can be repeated multiple times, and values are added in the order they are parsed.
2717 Applications intending to use OEM strings data are encouraged to use their application name as
2718 a prefix for the value string. This facilitates passing information for multiple applications
2721 The ``value=str`` syntax provides the string data inline, while the ``path=filename`` syntax
2722 loads data from a file on disk. Note that the file is not permitted to contain any NUL bytes.
2724 Both the ``value`` and ``path`` options can be repeated multiple times and will be added to
2725 the SMBIOS table in the order in which they appear.
2727 Note that on the x86 architecture, the total size of all SMBIOS tables is limited to 65535
2728 bytes. Thus the OEM strings data is not suitable for passing large amounts of data into the
2729 guest. Instead it should be used as a indicator to inform the guest where to locate the real
2730 data set, for example, by specifying the serial ID of a block device.
2732 An example passing three strings is
2736 -smbios type=11,value=cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/,\\
2737 value=anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os,\\
2738 path=/some/file/with/oemstringsdata.txt
2740 In the guest OS this is visible with the ``dmidecode`` command
2745 Handle 0x0E00, DMI type 11, 5 bytes
2747 String 1: cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/
2748 String 2: anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os
2749 String 3: myapp:some extra data
2752 ``-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str][,asset=str][,part=str][,speed=%d]``
2753 Specify SMBIOS type 17 fields
2755 ``-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]``
2756 Specify SMBIOS type 41 fields
2758 This argument can be repeated multiple times. Its main use is to allow network interfaces be created
2759 as ``enoX`` on Linux, with X being the instance number, instead of the name depending on the interface
2760 position on the PCI bus.
2762 Here is an example of use:
2766 -netdev user,id=internet \\
2767 -device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \\
2768 -smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev
2770 In the guest OS, the device should then appear as ``eno1``:
2775 lo UNKNOWN 00:00:00:00:00:00 <LOOPBACK,UP,LOWER_UP>
2776 eno1 UP 50:54:00:00:00:42 <BROADCAST,MULTICAST,UP,LOWER_UP>
2778 Currently, the PCI device has to be attached to the root bus.
2784 DEFHEADING(Network options:)
2786 DEF("netdev", HAS_ARG, QEMU_OPTION_netdev,
2788 "-netdev user,id=str[,ipv4=on|off][,net=addr[/mask]][,host=addr]\n"
2789 " [,ipv6=on|off][,ipv6-net=addr[/int]][,ipv6-host=addr]\n"
2790 " [,restrict=on|off][,hostname=host][,dhcpstart=addr]\n"
2791 " [,dns=addr][,ipv6-dns=addr][,dnssearch=domain][,domainname=domain]\n"
2792 " [,tftp=dir][,tftp-server-name=name][,bootfile=f][,hostfwd=rule][,guestfwd=rule]"
2794 "[,smb=dir[,smbserver=addr]]\n"
2796 " configure a user mode network backend with ID 'str',\n"
2797 " its DHCP server and optional services\n"
2800 "-netdev tap,id=str,ifname=name\n"
2801 " configure a host TAP network backend with ID 'str'\n"
2803 "-netdev tap,id=str[,fd=h][,fds=x:y:...:z][,ifname=name][,script=file][,downscript=dfile]\n"
2804 " [,br=bridge][,helper=helper][,sndbuf=nbytes][,vnet_hdr=on|off][,vhost=on|off]\n"
2805 " [,vhostfd=h][,vhostfds=x:y:...:z][,vhostforce=on|off][,queues=n]\n"
2807 " configure a host TAP network backend with ID 'str'\n"
2808 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2809 " use network scripts 'file' (default=" DEFAULT_NETWORK_SCRIPT ")\n"
2810 " to configure it and 'dfile' (default=" DEFAULT_NETWORK_DOWN_SCRIPT ")\n"
2811 " to deconfigure it\n"
2812 " use '[down]script=no' to disable script execution\n"
2813 " use network helper 'helper' (default=" DEFAULT_BRIDGE_HELPER ") to\n"
2815 " use 'fd=h' to connect to an already opened TAP interface\n"
2816 " use 'fds=x:y:...:z' to connect to already opened multiqueue capable TAP interfaces\n"
2817 " use 'sndbuf=nbytes' to limit the size of the send buffer (the\n"
2818 " default is disabled 'sndbuf=0' to enable flow control set 'sndbuf=1048576')\n"
2819 " use vnet_hdr=off to avoid enabling the IFF_VNET_HDR tap flag\n"
2820 " use vnet_hdr=on to make the lack of IFF_VNET_HDR support an error condition\n"
2821 " use vhost=on to enable experimental in kernel accelerator\n"
2822 " (only has effect for virtio guests which use MSIX)\n"
2823 " use vhostforce=on to force vhost on for non-MSIX virtio guests\n"
2824 " use 'vhostfd=h' to connect to an already opened vhost net device\n"
2825 " use 'vhostfds=x:y:...:z to connect to multiple already opened vhost net devices\n"
2826 " use 'queues=n' to specify the number of queues to be created for multiqueue TAP\n"
2827 " use 'poll-us=n' to specify the maximum number of microseconds that could be\n"
2828 " spent on busy polling for vhost net\n"
2829 "-netdev bridge,id=str[,br=bridge][,helper=helper]\n"
2830 " configure a host TAP network backend with ID 'str' that is\n"
2831 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n"
2832 " using the program 'helper (default=" DEFAULT_BRIDGE_HELPER ")\n"
2835 "-netdev l2tpv3,id=str,src=srcaddr,dst=dstaddr[,srcport=srcport][,dstport=dstport]\n"
2836 " [,rxsession=rxsession],txsession=txsession[,ipv6=on|off][,udp=on|off]\n"
2837 " [,cookie64=on|off][,counter][,pincounter][,txcookie=txcookie]\n"
2838 " [,rxcookie=rxcookie][,offset=offset]\n"
2839 " configure a network backend with ID 'str' connected to\n"
2840 " an Ethernet over L2TPv3 pseudowire.\n"
2841 " Linux kernel 3.3+ as well as most routers can talk\n"
2842 " L2TPv3. This transport allows connecting a VM to a VM,\n"
2843 " VM to a router and even VM to Host. It is a nearly-universal\n"
2844 " standard (RFC3931). Note - this implementation uses static\n"
2845 " pre-configured tunnels (same as the Linux kernel).\n"
2846 " use 'src=' to specify source address\n"
2847 " use 'dst=' to specify destination address\n"
2848 " use 'udp=on' to specify udp encapsulation\n"
2849 " use 'srcport=' to specify source udp port\n"
2850 " use 'dstport=' to specify destination udp port\n"
2851 " use 'ipv6=on' to force v6\n"
2852 " L2TPv3 uses cookies to prevent misconfiguration as\n"
2853 " well as a weak security measure\n"
2854 " use 'rxcookie=0x012345678' to specify a rxcookie\n"
2855 " use 'txcookie=0x012345678' to specify a txcookie\n"
2856 " use 'cookie64=on' to set cookie size to 64 bit, otherwise 32\n"
2857 " use 'counter=off' to force a 'cut-down' L2TPv3 with no counter\n"
2858 " use 'pincounter=on' to work around broken counter handling in peer\n"
2859 " use 'offset=X' to add an extra offset between header and data\n"
2861 "-netdev socket,id=str[,fd=h][,listen=[host]:port][,connect=host:port]\n"
2862 " configure a network backend to connect to another network\n"
2863 " using a socket connection\n"
2864 "-netdev socket,id=str[,fd=h][,mcast=maddr:port[,localaddr=addr]]\n"
2865 " configure a network backend to connect to a multicast maddr and port\n"
2866 " use 'localaddr=addr' to specify the host address to send packets from\n"
2867 "-netdev socket,id=str[,fd=h][,udp=host:port][,localaddr=host:port]\n"
2868 " configure a network backend to connect to another network\n"
2869 " using an UDP tunnel\n"
2870 "-netdev stream,id=str[,server=on|off],addr.type=inet,addr.host=host,addr.port=port[,to=maxport][,numeric=on|off][,keep-alive=on|off][,mptcp=on|off][,addr.ipv4=on|off][,addr.ipv6=on|off][,reconnect=seconds]\n"
2871 "-netdev stream,id=str[,server=on|off],addr.type=unix,addr.path=path[,abstract=on|off][,tight=on|off][,reconnect=seconds]\n"
2872 "-netdev stream,id=str[,server=on|off],addr.type=fd,addr.str=file-descriptor[,reconnect=seconds]\n"
2873 " configure a network backend to connect to another network\n"
2874 " using a socket connection in stream mode.\n"
2875 "-netdev dgram,id=str,remote.type=inet,remote.host=maddr,remote.port=port[,local.type=inet,local.host=addr]\n"
2876 "-netdev dgram,id=str,remote.type=inet,remote.host=maddr,remote.port=port[,local.type=fd,local.str=file-descriptor]\n"
2877 " configure a network backend to connect to a multicast maddr and port\n"
2878 " use ``local.host=addr`` to specify the host address to send packets from\n"
2879 "-netdev dgram,id=str,local.type=inet,local.host=addr,local.port=port[,remote.type=inet,remote.host=addr,remote.port=port]\n"
2880 "-netdev dgram,id=str,local.type=unix,local.path=path[,remote.type=unix,remote.path=path]\n"
2881 "-netdev dgram,id=str,local.type=fd,local.str=file-descriptor\n"
2882 " configure a network backend to connect to another network\n"
2883 " using an UDP tunnel\n"
2885 "-netdev vde,id=str[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]\n"
2886 " configure a network backend to connect to port 'n' of a vde switch\n"
2887 " running on host and listening for incoming connections on 'socketpath'.\n"
2888 " Use group 'groupname' and mode 'octalmode' to change default\n"
2889 " ownership and permissions for communication port.\n"
2891 #ifdef CONFIG_NETMAP
2892 "-netdev netmap,id=str,ifname=name[,devname=nmname]\n"
2893 " attach to the existing netmap-enabled network interface 'name', or to a\n"
2894 " VALE port (created on the fly) called 'name' ('nmname' is name of the \n"
2895 " netmap device, defaults to '/dev/netmap')\n"
2897 #ifdef CONFIG_AF_XDP
2898 "-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off]\n"
2899 " [,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]\n"
2900 " attach to the existing network interface 'name' with AF_XDP socket\n"
2901 " use 'mode=MODE' to specify an XDP program attach mode\n"
2902 " use 'force-copy=on|off' to force XDP copy mode even if device supports zero-copy (default: off)\n"
2903 " use 'inhibit=on|off' to inhibit loading of a default XDP program (default: off)\n"
2904 " with inhibit=on,\n"
2905 " use 'sock-fds' to provide file descriptors for already open AF_XDP sockets\n"
2906 " added to a socket map in XDP program. One socket per queue.\n"
2907 " use 'queues=n' to specify how many queues of a multiqueue interface should be used\n"
2908 " use 'start-queue=m' to specify the first queue that should be used\n"
2911 "-netdev vhost-user,id=str,chardev=dev[,vhostforce=on|off]\n"
2912 " configure a vhost-user network, backed by a chardev 'dev'\n"
2915 "-netdev vhost-vdpa,id=str[,vhostdev=/path/to/dev][,vhostfd=h]\n"
2916 " configure a vhost-vdpa network,Establish a vhost-vdpa netdev\n"
2917 " use 'vhostdev=/path/to/dev' to open a vhost vdpa device\n"
2918 " use 'vhostfd=h' to connect to an already opened vhost vdpa device\n"
2921 "-netdev vmnet-host,id=str[,isolated=on|off][,net-uuid=uuid]\n"
2922 " [,start-address=addr,end-address=addr,subnet-mask=mask]\n"
2923 " configure a vmnet network backend in host mode with ID 'str',\n"
2924 " isolate this interface from others with 'isolated',\n"
2925 " configure the address range and choose a subnet mask,\n"
2926 " specify network UUID 'uuid' to disable DHCP and interact with\n"
2927 " vmnet-host interfaces within this isolated network\n"
2928 "-netdev vmnet-shared,id=str[,isolated=on|off][,nat66-prefix=addr]\n"
2929 " [,start-address=addr,end-address=addr,subnet-mask=mask]\n"
2930 " configure a vmnet network backend in shared mode with ID 'str',\n"
2931 " configure the address range and choose a subnet mask,\n"
2932 " set IPv6 ULA prefix (of length 64) to use for internal network,\n"
2933 " isolate this interface from others with 'isolated'\n"
2934 "-netdev vmnet-bridged,id=str,ifname=name[,isolated=on|off]\n"
2935 " configure a vmnet network backend in bridged mode with ID 'str',\n"
2936 " use 'ifname=name' to select a physical network interface to be bridged,\n"
2937 " isolate this interface from others with 'isolated'\n"
2939 "-netdev hubport,id=str,hubid=n[,netdev=nd]\n"
2940 " configure a hub port on the hub with ID 'n'\n", QEMU_ARCH_ALL)
2941 DEF("nic", HAS_ARG, QEMU_OPTION_nic,
2952 #ifdef CONFIG_NETMAP
2955 #ifdef CONFIG_AF_XDP
2962 "vmnet-host|vmnet-shared|vmnet-bridged|"
2964 "socket][,option][,...][mac=macaddr]\n"
2965 " initialize an on-board / default host NIC (using MAC address\n"
2966 " macaddr) and connect it to the given host network backend\n"
2967 "-nic none use it alone to have zero network devices (the default is to\n"
2968 " provided a 'user' network connection)\n",
2970 DEF("net", HAS_ARG, QEMU_OPTION_net,
2971 "-net nic[,macaddr=mac][,model=type][,name=str][,addr=str][,vectors=v]\n"
2972 " configure or create an on-board (or machine default) NIC and\n"
2973 " connect it to hub 0 (please use -nic unless you need a hub)\n"
2983 #ifdef CONFIG_NETMAP
2986 #ifdef CONFIG_AF_XDP
2990 "vmnet-host|vmnet-shared|vmnet-bridged|"
2992 "socket][,option][,option][,...]\n"
2993 " old way to initialize a host network interface\n"
2994 " (use the -netdev option if possible instead)\n", QEMU_ARCH_ALL)
2996 ``-nic [tap|bridge|user|l2tpv3|vde|netmap|af-xdp|vhost-user|socket][,...][,mac=macaddr][,model=mn]``
2997 This option is a shortcut for configuring both the on-board
2998 (default) guest NIC hardware and the host network backend in one go.
2999 The host backend options are the same as with the corresponding
3000 ``-netdev`` options below. The guest NIC model can be set with
3001 ``model=modelname``. Use ``model=help`` to list the available device
3002 types. The hardware MAC address can be set with ``mac=macaddr``.
3004 The following two example do exactly the same, to show how ``-nic``
3005 can be used to shorten the command line length:
3009 |qemu_system| -netdev user,id=n1,ipv6=off -device e1000,netdev=n1,mac=52:54:98:76:54:32
3010 |qemu_system| -nic user,ipv6=off,model=e1000,mac=52:54:98:76:54:32
3013 Indicate that no network devices should be configured. It is used to
3014 override the default configuration (default NIC with "user" host
3015 network backend) which is activated if no other networking options
3018 ``-netdev user,id=id[,option][,option][,...]``
3019 Configure user mode host network backend which requires no
3020 administrator privilege to run. Valid options are:
3023 Assign symbolic name for use in monitor commands.
3025 ``ipv4=on|off and ipv6=on|off``
3026 Specify that either IPv4 or IPv6 must be enabled. If neither is
3027 specified both protocols are enabled.
3030 Set IP network address the guest will see. Optionally specify
3031 the netmask, either in the form a.b.c.d or as number of valid
3032 top-most bits. Default is 10.0.2.0/24.
3035 Specify the guest-visible address of the host. Default is the
3036 2nd IP in the guest network, i.e. x.x.x.2.
3038 ``ipv6-net=addr[/int]``
3039 Set IPv6 network address the guest will see (default is
3040 fec0::/64). The network prefix is given in the usual hexadecimal
3041 IPv6 address notation. The prefix size is optional, and is given
3042 as the number of valid top-most bits (default is 64).
3045 Specify the guest-visible IPv6 address of the host. Default is
3046 the 2nd IPv6 in the guest network, i.e. xxxx::2.
3049 If this option is enabled, the guest will be isolated, i.e. it
3050 will not be able to contact the host and no guest IP packets
3051 will be routed over the host to the outside. This option does
3052 not affect any explicitly set forwarding rules.
3055 Specifies the client hostname reported by the built-in DHCP
3059 Specify the first of the 16 IPs the built-in DHCP server can
3060 assign. Default is the 15th to 31st IP in the guest network,
3061 i.e. x.x.x.15 to x.x.x.31.
3064 Specify the guest-visible address of the virtual nameserver. The
3065 address must be different from the host address. Default is the
3066 3rd IP in the guest network, i.e. x.x.x.3.
3069 Specify the guest-visible address of the IPv6 virtual
3070 nameserver. The address must be different from the host address.
3071 Default is the 3rd IP in the guest network, i.e. xxxx::3.
3073 ``dnssearch=domain``
3074 Provides an entry for the domain-search list sent by the
3075 built-in DHCP server. More than one domain suffix can be
3076 transmitted by specifying this option multiple times. If
3077 supported, this will cause the guest to automatically try to
3078 append the given domain suffix(es) in case a domain name can not
3085 |qemu_system| -nic user,dnssearch=mgmt.example.org,dnssearch=example.org
3087 ``domainname=domain``
3088 Specifies the client domain name reported by the built-in DHCP
3092 When using the user mode network stack, activate a built-in TFTP
3093 server. The files in dir will be exposed as the root of a TFTP
3094 server. The TFTP client on the guest must be configured in
3095 binary mode (use the command ``bin`` of the Unix TFTP client).
3097 ``tftp-server-name=name``
3098 In BOOTP reply, broadcast name as the "TFTP server name"
3099 (RFC2132 option 66). This can be used to advise the guest to
3100 load boot files or configurations from a different server than
3104 When using the user mode network stack, broadcast file as the
3105 BOOTP filename. In conjunction with ``tftp``, this can be used
3106 to network boot a guest from a local directory.
3108 Example (using pxelinux):
3112 |qemu_system| -hda linux.img -boot n -device e1000,netdev=n1 \\
3113 -netdev user,id=n1,tftp=/path/to/tftp/files,bootfile=/pxelinux.0
3115 ``smb=dir[,smbserver=addr]``
3116 When using the user mode network stack, activate a built-in SMB
3117 server so that Windows OSes can access to the host files in
3118 ``dir`` transparently. The IP address of the SMB server can be
3119 set to addr. By default the 4th IP in the guest network is used,
3122 In the guest Windows OS, the line:
3128 must be added in the file ``C:\WINDOWS\LMHOSTS`` (for windows
3129 9x/Me) or ``C:\WINNT\SYSTEM32\DRIVERS\ETC\LMHOSTS`` (Windows
3132 Then ``dir`` can be accessed in ``\\smbserver\qemu``.
3134 Note that a SAMBA server must be installed on the host OS.
3136 ``hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport``
3137 Redirect incoming TCP or UDP connections to the host port
3138 hostport to the guest IP address guestaddr on guest port
3139 guestport. If guestaddr is not specified, its value is x.x.x.15
3140 (default first address given by the built-in DHCP server). By
3141 specifying hostaddr, the rule can be bound to a specific host
3142 interface. If no connection type is set, TCP is used. This
3143 option can be given multiple times.
3145 For example, to redirect host X11 connection from screen 1 to
3146 guest screen 0, use the following:
3151 |qemu_system| -nic user,hostfwd=tcp:127.0.0.1:6001-:6000
3152 # this host xterm should open in the guest X11 server
3155 To redirect telnet connections from host port 5555 to telnet
3156 port on the guest, use the following:
3161 |qemu_system| -nic user,hostfwd=tcp::5555-:23
3162 telnet localhost 5555
3164 Then when you use on the host ``telnet localhost 5555``, you
3165 connect to the guest telnet server.
3167 ``guestfwd=[tcp]:server:port-dev``; \ ``guestfwd=[tcp]:server:port-cmd:command``
3168 Forward guest TCP connections to the IP address server on port
3169 port to the character device dev or to a program executed by
3170 cmd:command which gets spawned for each connection. This option
3171 can be given multiple times.
3173 You can either use a chardev directly and have that one used
3174 throughout QEMU's lifetime, like in the following example:
3178 # open 10.10.1.1:4321 on bootup, connect 10.0.2.100:1234 to it whenever
3179 # the guest accesses it
3180 |qemu_system| -nic user,guestfwd=tcp:10.0.2.100:1234-tcp:10.10.1.1:4321
3182 Or you can execute a command on every TCP connection established
3183 by the guest, so that QEMU behaves similar to an inetd process
3184 for that virtual server:
3188 # call "netcat 10.10.1.1 4321" on every TCP connection to 10.0.2.100:1234
3189 # and connect the TCP stream to its stdin/stdout
3190 |qemu_system| -nic 'user,id=n1,guestfwd=tcp:10.0.2.100:1234-cmd:netcat 10.10.1.1 4321'
3192 ``-netdev tap,id=id[,fd=h][,ifname=name][,script=file][,downscript=dfile][,br=bridge][,helper=helper]``
3193 Configure a host TAP network backend with ID id.
3195 Use the network script file to configure it and the network script
3196 dfile to deconfigure it. If name is not provided, the OS
3197 automatically provides one. The default network configure script is
3198 ``/etc/qemu-ifup`` and the default network deconfigure script is
3199 ``/etc/qemu-ifdown``. Use ``script=no`` or ``downscript=no`` to
3200 disable script execution.
3202 If running QEMU as an unprivileged user, use the network helper
3203 to configure the TAP interface and attach it to the bridge.
3204 The default network helper executable is
3205 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3208 ``fd``\ =h can be used to specify the handle of an already opened
3215 #launch a QEMU instance with the default network script
3216 |qemu_system| linux.img -nic tap
3220 #launch a QEMU instance with two NICs, each one connected
3222 |qemu_system| linux.img \\
3223 -netdev tap,id=nd0,ifname=tap0 -device e1000,netdev=nd0 \\
3224 -netdev tap,id=nd1,ifname=tap1 -device rtl8139,netdev=nd1
3228 #launch a QEMU instance with the default network helper to
3229 #connect a TAP device to bridge br0
3230 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3231 -netdev tap,id=n1,"helper=/path/to/qemu-bridge-helper"
3233 ``-netdev bridge,id=id[,br=bridge][,helper=helper]``
3234 Connect a host TAP network interface to a host bridge device.
3236 Use the network helper helper to configure the TAP interface and
3237 attach it to the bridge. The default network helper executable is
3238 ``/path/to/qemu-bridge-helper`` and the default bridge device is
3245 #launch a QEMU instance with the default network helper to
3246 #connect a TAP device to bridge br0
3247 |qemu_system| linux.img -netdev bridge,id=n1 -device virtio-net,netdev=n1
3251 #launch a QEMU instance with the default network helper to
3252 #connect a TAP device to bridge qemubr0
3253 |qemu_system| linux.img -netdev bridge,br=qemubr0,id=n1 -device virtio-net,netdev=n1
3255 ``-netdev socket,id=id[,fd=h][,listen=[host]:port][,connect=host:port]``
3256 This host network backend can be used to connect the guest's network
3257 to another QEMU virtual machine using a TCP socket connection. If
3258 ``listen`` is specified, QEMU waits for incoming connections on port
3259 (host is optional). ``connect`` is used to connect to another QEMU
3260 instance using the ``listen`` option. ``fd``\ =h specifies an
3261 already opened TCP socket.
3267 # launch a first QEMU instance
3268 |qemu_system| linux.img \\
3269 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3270 -netdev socket,id=n1,listen=:1234
3271 # connect the network of this instance to the network of the first instance
3272 |qemu_system| linux.img \\
3273 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3274 -netdev socket,id=n2,connect=127.0.0.1:1234
3276 ``-netdev socket,id=id[,fd=h][,mcast=maddr:port[,localaddr=addr]]``
3277 Configure a socket host network backend to share the guest's network
3278 traffic with another QEMU virtual machines using a UDP multicast
3279 socket, effectively making a bus for every QEMU with same multicast
3280 address maddr and port. NOTES:
3282 1. Several QEMU can be running on different hosts and share same bus
3283 (assuming correct multicast setup for these hosts).
3285 2. mcast support is compatible with User Mode Linux (argument
3286 ``ethN=mcast``), see http://user-mode-linux.sf.net.
3288 3. Use ``fd=h`` to specify an already opened UDP multicast socket.
3294 # launch one QEMU instance
3295 |qemu_system| linux.img \\
3296 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3297 -netdev socket,id=n1,mcast=230.0.0.1:1234
3298 # launch another QEMU instance on same "bus"
3299 |qemu_system| linux.img \\
3300 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\
3301 -netdev socket,id=n2,mcast=230.0.0.1:1234
3302 # launch yet another QEMU instance on same "bus"
3303 |qemu_system| linux.img \\
3304 -device e1000,netdev=n3,mac=52:54:00:12:34:58 \\
3305 -netdev socket,id=n3,mcast=230.0.0.1:1234
3307 Example (User Mode Linux compat.):
3311 # launch QEMU instance (note mcast address selected is UML's default)
3312 |qemu_system| linux.img \\
3313 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3314 -netdev socket,id=n1,mcast=239.192.168.1:1102
3316 /path/to/linux ubd0=/path/to/root_fs eth0=mcast
3318 Example (send packets from host's 1.2.3.4):
3322 |qemu_system| linux.img \\
3323 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\
3324 -netdev socket,id=n1,mcast=239.192.168.1:1102,localaddr=1.2.3.4
3326 ``-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]``
3327 Configure a L2TPv3 pseudowire host network backend. L2TPv3 (RFC3931)
3328 is a popular protocol to transport Ethernet (and other Layer 2) data
3329 frames between two systems. It is present in routers, firewalls and
3330 the Linux kernel (from version 3.3 onwards).
3332 This transport allows a VM to communicate to another VM, router or
3336 source address (mandatory)
3339 destination address (mandatory)
3342 select udp encapsulation (default is ip).
3348 destination udp port.
3351 force v6, otherwise defaults to v4.
3353 ``rxcookie=rxcookie``; \ ``txcookie=txcookie``
3354 Cookies are a weak form of security in the l2tpv3 specification.
3355 Their function is mostly to prevent misconfiguration. By default
3359 Set cookie size to 64 bit instead of the default 32
3362 Force a 'cut-down' L2TPv3 with no counter as in
3363 draft-mkonstan-l2tpext-keyed-ipv6-tunnel-00
3366 Work around broken counter handling in peer. This may also help
3367 on networks which have packet reorder.
3370 Add an extra offset between header and data
3372 For example, to attach a VM running on host 4.3.2.1 via L2TPv3 to
3373 the bridge br-lan on the remote Linux host 1.2.3.4:
3377 # Setup tunnel on linux host using raw ip as encapsulation
3379 ip l2tp add tunnel remote 4.3.2.1 local 1.2.3.4 tunnel_id 1 peer_tunnel_id 1 \\
3380 encap udp udp_sport 16384 udp_dport 16384
3381 ip l2tp add session tunnel_id 1 name vmtunnel0 session_id \\
3382 0xFFFFFFFF peer_session_id 0xFFFFFFFF
3383 ifconfig vmtunnel0 mtu 1500
3384 ifconfig vmtunnel0 up
3385 brctl addif br-lan vmtunnel0
3389 # launch QEMU instance - if your network has reorder or is very lossy add ,pincounter
3391 |qemu_system| linux.img -device e1000,netdev=n1 \\
3392 -netdev l2tpv3,id=n1,src=4.2.3.1,dst=1.2.3.4,udp,srcport=16384,dstport=16384,rxsession=0xffffffff,txsession=0xffffffff,counter
3394 ``-netdev vde,id=id[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]``
3395 Configure VDE backend to connect to PORT n of a vde switch running
3396 on host and listening for incoming connections on socketpath. Use
3397 GROUP groupname and MODE octalmode to change default ownership and
3398 permissions for communication port. This option is only available if
3399 QEMU has been compiled with vde support enabled.
3406 vde_switch -F -sock /tmp/myswitch
3407 # launch QEMU instance
3408 |qemu_system| linux.img -nic vde,sock=/tmp/myswitch
3410 ``-netdev af-xdp,id=str,ifname=name[,mode=native|skb][,force-copy=on|off][,queues=n][,start-queue=m][,inhibit=on|off][,sock-fds=x:y:...:z]``
3411 Configure AF_XDP backend to connect to a network interface 'name'
3412 using AF_XDP socket. A specific program attach mode for a default
3413 XDP program can be forced with 'mode', defaults to best-effort,
3414 where the likely most performant mode will be in use. Number of queues
3415 'n' should generally match the number or queues in the interface,
3416 defaults to 1. Traffic arriving on non-configured device queues will
3417 not be delivered to the network backend.
3421 # set number of queues to 4
3422 ethtool -L eth0 combined 4
3423 # launch QEMU instance
3424 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3425 -netdev af-xdp,id=n1,ifname=eth0,queues=4
3427 'start-queue' option can be specified if a particular range of queues
3428 [m, m + n] should be in use. For example, this is may be necessary in
3429 order to use certain NICs in native mode. Kernel allows the driver to
3430 create a separate set of XDP queues on top of regular ones, and only
3431 these queues can be used for AF_XDP sockets. NICs that work this way
3432 may also require an additional traffic redirection with ethtool to these
3437 # set number of queues to 1
3438 ethtool -L eth0 combined 1
3439 # redirect all the traffic to the second queue (id: 1)
3440 # note: drivers may require non-empty key/mask pair.
3441 ethtool -N eth0 flow-type ether \\
3442 dst 00:00:00:00:00:00 m FF:FF:FF:FF:FF:FE action 1
3443 ethtool -N eth0 flow-type ether \\
3444 dst 00:00:00:00:00:01 m FF:FF:FF:FF:FF:FE action 1
3445 # launch QEMU instance
3446 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3447 -netdev af-xdp,id=n1,ifname=eth0,queues=1,start-queue=1
3449 XDP program can also be loaded externally. In this case 'inhibit' option
3450 should be set to 'on' and 'sock-fds' provided with file descriptors for
3451 already open but not bound XDP sockets already added to a socket map for
3452 corresponding queues. One socket per queue.
3456 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\
3457 -netdev af-xdp,id=n1,ifname=eth0,queues=3,inhibit=on,sock-fds=15:16:17
3459 ``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]``
3460 Establish a vhost-user netdev, backed by a chardev id. The chardev
3461 should be a unix domain socket backed one. The vhost-user uses a
3462 specifically defined protocol to pass vhost ioctl replacement
3463 messages to an application on the other end of the socket. On
3464 non-MSIX guests, the feature can be forced with vhostforce. Use
3465 'queues=n' to specify the number of queues to be created for
3466 multiqueue vhost-user.
3472 qemu -m 512 -object memory-backend-file,id=mem,size=512M,mem-path=/hugetlbfs,share=on \
3473 -numa node,memdev=mem \
3474 -chardev socket,id=chr0,path=/path/to/socket \
3475 -netdev type=vhost-user,id=net0,chardev=chr0 \
3476 -device virtio-net-pci,netdev=net0
3478 ``-netdev vhost-vdpa[,vhostdev=/path/to/dev][,vhostfd=h]``
3479 Establish a vhost-vdpa netdev.
3481 vDPA device is a device that uses a datapath which complies with
3482 the virtio specifications with a vendor specific control path.
3483 vDPA devices can be both physically located on the hardware or
3484 emulated by software.
3486 ``-netdev hubport,id=id,hubid=hubid[,netdev=nd]``
3487 Create a hub port on the emulated hub with ID hubid.
3489 The hubport netdev lets you connect a NIC to a QEMU emulated hub
3490 instead of a single netdev. Alternatively, you can also connect the
3491 hubport to another netdev with ID nd by using the ``netdev=nd``
3494 ``-net nic[,netdev=nd][,macaddr=mac][,model=type] [,name=name][,addr=addr][,vectors=v]``
3495 Legacy option to configure or create an on-board (or machine
3496 default) Network Interface Card(NIC) and connect it either to the
3497 emulated hub with ID 0 (i.e. the default hub), or to the netdev nd.
3498 If model is omitted, then the default NIC model associated with the
3499 machine type is used. Note that the default NIC model may change in
3500 future QEMU releases, so it is highly recommended to always specify
3501 a model. Optionally, the MAC address can be changed to mac, the
3502 device address set to addr (PCI cards only), and a name can be
3503 assigned for use in monitor commands. Optionally, for PCI cards, you
3504 can specify the number v of MSI-X vectors that the card should have;
3505 this option currently only affects virtio cards; set v = 0 to
3506 disable MSI-X. If no ``-net`` option is specified, a single NIC is
3507 created. QEMU can emulate several different models of network card.
3508 Use ``-net nic,model=help`` for a list of available devices for your
3511 ``-net user|tap|bridge|socket|l2tpv3|vde[,...][,name=name]``
3512 Configure a host network backend (with the options corresponding to
3513 the same ``-netdev`` option) and connect it to the emulated hub 0
3514 (the default hub). Use name to specify the name of the hub port.
3519 DEFHEADING(Character device options:)
3521 DEF("chardev", HAS_ARG, QEMU_OPTION_chardev,
3523 "-chardev null,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3524 "-chardev socket,id=id[,host=host],port=port[,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]\n"
3525 " [,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds][,mux=on|off]\n"
3526 " [,logfile=PATH][,logappend=on|off][,tls-creds=ID][,tls-authz=ID] (tcp)\n"
3527 "-chardev socket,id=id,path=path[,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds]\n"
3528 " [,mux=on|off][,logfile=PATH][,logappend=on|off][,abstract=on|off][,tight=on|off] (unix)\n"
3529 "-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr]\n"
3530 " [,localport=localport][,ipv4=on|off][,ipv6=on|off][,mux=on|off]\n"
3531 " [,logfile=PATH][,logappend=on|off]\n"
3532 "-chardev msmouse,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3533 "-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]\n"
3534 " [,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3535 "-chardev ringbuf,id=id[,size=size][,logfile=PATH][,logappend=on|off]\n"
3536 "-chardev file,id=id,path=path[,input-path=input-file][,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3537 "-chardev pipe,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3539 "-chardev console,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3540 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3542 "-chardev pty,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3543 "-chardev stdio,id=id[,mux=on|off][,signal=on|off][,logfile=PATH][,logappend=on|off]\n"
3545 #ifdef CONFIG_BRLAPI
3546 "-chardev braille,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3548 #if defined(__linux__) || defined(__sun__) || defined(__FreeBSD__) \
3549 || defined(__NetBSD__) || defined(__OpenBSD__) || defined(__DragonFly__)
3550 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3552 #if defined(__linux__) || defined(__FreeBSD__) || defined(__DragonFly__)
3553 "-chardev parallel,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n"
3555 #if defined(CONFIG_SPICE)
3556 "-chardev spicevmc,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n"
3557 "-chardev spiceport,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n"
3563 The general form of a character device option is:
3565 ``-chardev backend,id=id[,mux=on|off][,options]``
3566 Backend is one of: ``null``, ``socket``, ``udp``, ``msmouse``,
3567 ``vc``, ``ringbuf``, ``file``, ``pipe``, ``console``, ``serial``,
3568 ``pty``, ``stdio``, ``braille``, ``parallel``,
3569 ``spicevmc``, ``spiceport``. The specific backend will determine the
3572 Use ``-chardev help`` to print all available chardev backend types.
3574 All devices must have an id, which can be any string up to 127
3575 characters long. It is used to uniquely identify this device in
3576 other command line directives.
3578 A character device may be used in multiplexing mode by multiple
3579 front-ends. Specify ``mux=on`` to enable this mode. A multiplexer is
3580 a "1:N" device, and here the "1" end is your specified chardev
3581 backend, and the "N" end is the various parts of QEMU that can talk
3582 to a chardev. If you create a chardev with ``id=myid`` and
3583 ``mux=on``, QEMU will create a multiplexer with your specified ID,
3584 and you can then configure multiple front ends to use that chardev
3585 ID for their input/output. Up to four different front ends can be
3586 connected to a single multiplexed chardev. (Without multiplexing
3587 enabled, a chardev can only be used by a single front end.) For
3588 instance you could use this to allow a single stdio chardev to be
3589 used by two serial ports and the QEMU monitor:
3593 -chardev stdio,mux=on,id=char0 \
3594 -mon chardev=char0,mode=readline \
3595 -serial chardev:char0 \
3596 -serial chardev:char0
3598 You can have more than one multiplexer in a system configuration;
3599 for instance you could have a TCP port multiplexed between UART 0
3600 and UART 1, and stdio multiplexed between the QEMU monitor and a
3605 -chardev stdio,mux=on,id=char0 \
3606 -mon chardev=char0,mode=readline \
3607 -parallel chardev:char0 \
3608 -chardev tcp,...,mux=on,id=char1 \
3609 -serial chardev:char1 \
3610 -serial chardev:char1
3612 When you're using a multiplexed character device, some escape
3613 sequences are interpreted in the input. See the chapter about
3614 :ref:`keys in the character backend multiplexer` in the
3615 System Emulation Users Guide for more details.
3617 Note that some other command line options may implicitly create
3618 multiplexed character backends; for instance ``-serial mon:stdio``
3619 creates a multiplexed stdio backend connected to the serial port and
3620 the QEMU monitor, and ``-nographic`` also multiplexes the console
3621 and the monitor to stdio.
3623 There is currently no support for multiplexing in the other
3624 direction (where a single QEMU front end takes input and output from
3627 Every backend supports the ``logfile`` option, which supplies the
3628 path to a file to record all data transmitted via the backend. The
3629 ``logappend`` option controls whether the log file will be truncated
3630 or appended to when opened.
3632 The available backends are:
3634 ``-chardev null,id=id``
3635 A void device. This device will not emit any data, and will drop any
3636 data it receives. The null backend does not take any options.
3638 ``-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]``
3639 Create a two-way stream socket, which can be either a TCP or a unix
3640 socket. A unix socket will be created if ``path`` is specified.
3641 Behaviour is undefined if TCP options are specified for a unix
3644 ``server=on|off`` specifies that the socket shall be a listening socket.
3646 ``wait=on|off`` specifies that QEMU should not block waiting for a client
3647 to connect to a listening socket.
3649 ``telnet=on|off`` specifies that traffic on the socket should interpret
3650 telnet escape sequences.
3652 ``websocket=on|off`` specifies that the socket uses WebSocket protocol for
3655 ``reconnect`` sets the timeout for reconnecting on non-server
3656 sockets when the remote end goes away. qemu will delay this many
3657 seconds and then attempt to reconnect. Zero disables reconnecting,
3660 ``tls-creds`` requests enablement of the TLS protocol for
3661 encryption, and specifies the id of the TLS credentials to use for
3662 the handshake. The credentials must be previously created with the
3663 ``-object tls-creds`` argument.
3665 ``tls-auth`` provides the ID of the QAuthZ authorization object
3666 against which the client's x509 distinguished name will be
3667 validated. This object is only resolved at time of use, so can be
3668 deleted and recreated on the fly while the chardev server is active.
3669 If missing, it will default to denying access.
3671 TCP and unix socket options are given below:
3673 ``TCP options: port=port[,host=host][,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]``
3674 ``host`` for a listening socket specifies the local address to
3675 be bound. For a connecting socket species the remote host to
3676 connect to. ``host`` is optional for listening sockets. If not
3677 specified it defaults to ``0.0.0.0``.
3679 ``port`` for a listening socket specifies the local port to be
3680 bound. For a connecting socket specifies the port on the remote
3681 host to connect to. ``port`` can be given as either a port
3682 number or a service name. ``port`` is required.
3684 ``to`` is only relevant to listening sockets. If it is
3685 specified, and ``port`` cannot be bound, QEMU will attempt to
3686 bind to subsequent ports up to and including ``to`` until it
3687 succeeds. ``to`` must be specified as a port number.
3689 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4
3690 or IPv6 must be used. If neither is specified the socket may
3691 use either protocol.
3693 ``nodelay=on|off`` disables the Nagle algorithm.
3695 ``unix options: path=path[,abstract=on|off][,tight=on|off]``
3696 ``path`` specifies the local path of the unix socket. ``path``
3698 ``abstract=on|off`` specifies the use of the abstract socket namespace,
3699 rather than the filesystem. Optional, defaults to false.
3700 ``tight=on|off`` sets the socket length of abstract sockets to their minimum,
3701 rather than the full sun_path length. Optional, defaults to true.
3703 ``-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr][,localport=localport][,ipv4=on|off][,ipv6=on|off]``
3704 Sends all traffic from the guest to a remote host over UDP.
3706 ``host`` specifies the remote host to connect to. If not specified
3707 it defaults to ``localhost``.
3709 ``port`` specifies the port on the remote host to connect to.
3710 ``port`` is required.
3712 ``localaddr`` specifies the local address to bind to. If not
3713 specified it defaults to ``0.0.0.0``.
3715 ``localport`` specifies the local port to bind to. If not specified
3716 any available local port will be used.
3718 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4 or IPv6 must be used.
3719 If neither is specified the device may use either protocol.
3721 ``-chardev msmouse,id=id``
3722 Forward QEMU's emulated msmouse events to the guest. ``msmouse``
3723 does not take any options.
3725 ``-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]``
3726 Connect to a QEMU text console. ``vc`` may optionally be given a
3729 ``width`` and ``height`` specify the width and height respectively
3730 of the console, in pixels.
3732 ``cols`` and ``rows`` specify that the console be sized to fit a
3733 text console with the given dimensions.
3735 ``-chardev ringbuf,id=id[,size=size]``
3736 Create a ring buffer with fixed size ``size``. size must be a power
3737 of two and defaults to ``64K``.
3739 ``-chardev file,id=id,path=path[,input-path=input-path]``
3740 Log all traffic received from the guest to a file.
3742 ``path`` specifies the path of the file to be opened. This file will
3743 be created if it does not already exist, and overwritten if it does.
3744 ``path`` is required.
3746 If ``input-path`` is specified, this is the path of a second file
3747 which will be used for input. If ``input-path`` is not specified,
3748 no input will be available from the chardev.
3750 Note that ``input-path`` is not supported on Windows hosts.
3752 ``-chardev pipe,id=id,path=path``
3753 Create a two-way connection to the guest. The behaviour differs
3754 slightly between Windows hosts and other hosts:
3756 On Windows, a single duplex pipe will be created at
3759 On other hosts, 2 pipes will be created called ``path.in`` and
3760 ``path.out``. Data written to ``path.in`` will be received by the
3761 guest. Data written by the guest can be read from ``path.out``. QEMU
3762 will not create these fifos, and requires them to be present.
3764 ``path`` forms part of the pipe path as described above. ``path`` is
3767 ``-chardev console,id=id``
3768 Send traffic from the guest to QEMU's standard output. ``console``
3769 does not take any options.
3771 ``console`` is only available on Windows hosts.
3773 ``-chardev serial,id=id,path=path``
3774 Send traffic from the guest to a serial device on the host.
3776 On Unix hosts serial will actually accept any tty device, not only
3779 ``path`` specifies the name of the serial device to open.
3781 ``-chardev pty,id=id``
3782 Create a new pseudo-terminal on the host and connect to it. ``pty``
3783 does not take any options.
3785 ``pty`` is not available on Windows hosts.
3787 ``-chardev stdio,id=id[,signal=on|off]``
3788 Connect to standard input and standard output of the QEMU process.
3790 ``signal`` controls if signals are enabled on the terminal, that
3791 includes exiting QEMU with the key sequence Control-c. This option
3792 is enabled by default, use ``signal=off`` to disable it.
3794 ``-chardev braille,id=id``
3795 Connect to a local BrlAPI server. ``braille`` does not take any
3798 ``-chardev parallel,id=id,path=path``
3800 ``parallel`` is only available on Linux, FreeBSD and DragonFlyBSD
3803 Connect to a local parallel port.
3805 ``path`` specifies the path to the parallel port device. ``path`` is
3808 ``-chardev spicevmc,id=id,debug=debug,name=name``
3809 ``spicevmc`` is only available when spice support is built in.
3811 ``debug`` debug level for spicevmc
3813 ``name`` name of spice channel to connect to
3815 Connect to a spice virtual machine channel, such as vdiport.
3817 ``-chardev spiceport,id=id,debug=debug,name=name``
3818 ``spiceport`` is only available when spice support is built in.
3820 ``debug`` debug level for spicevmc
3822 ``name`` name of spice port to connect to
3824 Connect to a spice port, allowing a Spice client to handle the
3825 traffic identified by a name (preferably a fqdn).
3831 DEFHEADING(TPM device options:)
3833 DEF("tpmdev", HAS_ARG, QEMU_OPTION_tpmdev, \
3834 "-tpmdev passthrough,id=id[,path=path][,cancel-path=path]\n"
3835 " use path to provide path to a character device; default is /dev/tpm0\n"
3836 " use cancel-path to provide path to TPM's cancel sysfs entry; if\n"
3837 " not provided it will be searched for in /sys/class/misc/tpm?/device\n"
3838 "-tpmdev emulator,id=id,chardev=dev\n"
3839 " configure the TPM device using chardev backend\n",
3842 The general form of a TPM device option is:
3844 ``-tpmdev backend,id=id[,options]``
3845 The specific backend type will determine the applicable options. The
3846 ``-tpmdev`` option creates the TPM backend and requires a
3847 ``-device`` option that specifies the TPM frontend interface model.
3849 Use ``-tpmdev help`` to print all available TPM backend types.
3851 The available backends are:
3853 ``-tpmdev passthrough,id=id,path=path,cancel-path=cancel-path``
3854 (Linux-host only) Enable access to the host's TPM using the
3857 ``path`` specifies the path to the host's TPM device, i.e., on a
3858 Linux host this would be ``/dev/tpm0``. ``path`` is optional and by
3859 default ``/dev/tpm0`` is used.
3861 ``cancel-path`` specifies the path to the host TPM device's sysfs
3862 entry allowing for cancellation of an ongoing TPM command.
3863 ``cancel-path`` is optional and by default QEMU will search for the
3866 Some notes about using the host's TPM with the passthrough driver:
3868 The TPM device accessed by the passthrough driver must not be used
3869 by any other application on the host.
3871 Since the host's firmware (BIOS/UEFI) has already initialized the
3872 TPM, the VM's firmware (BIOS/UEFI) will not be able to initialize
3873 the TPM again and may therefore not show a TPM-specific menu that
3874 would otherwise allow the user to configure the TPM, e.g., allow the
3875 user to enable/disable or activate/deactivate the TPM. Further, if
3876 TPM ownership is released from within a VM then the host's TPM will
3877 get disabled and deactivated. To enable and activate the TPM again
3878 afterwards, the host has to be rebooted and the user is required to
3879 enter the firmware's menu to enable and activate the TPM. If the TPM
3880 is left disabled and/or deactivated most TPM commands will fail.
3882 To create a passthrough TPM use the following two options:
3886 -tpmdev passthrough,id=tpm0 -device tpm-tis,tpmdev=tpm0
3888 Note that the ``-tpmdev`` id is ``tpm0`` and is referenced by
3889 ``tpmdev=tpm0`` in the device option.
3891 ``-tpmdev emulator,id=id,chardev=dev``
3892 (Linux-host only) Enable access to a TPM emulator using Unix domain
3893 socket based chardev backend.
3895 ``chardev`` specifies the unique ID of a character device backend
3896 that provides connection to the software TPM server.
3898 To create a TPM emulator backend device with chardev socket backend:
3902 -chardev socket,id=chrtpm,path=/tmp/swtpm-sock -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0
3909 DEFHEADING(Boot Image or Kernel specific:)
3911 There are broadly 4 ways you can boot a system with QEMU.
3913 - specify a firmware and let it control finding a kernel
3914 - specify a firmware and pass a hint to the kernel to boot
3915 - direct kernel image boot
3916 - manually load files into the guest's address space
3918 The third method is useful for quickly testing kernels but as there is
3919 no firmware to pass configuration information to the kernel the
3920 hardware must either be probeable, the kernel built for the exact
3921 configuration or passed some configuration data (e.g. a DTB blob)
3922 which tells the kernel what drivers it needs. This exact details are
3923 often hardware specific.
3925 The final method is the most generic way of loading images into the
3926 guest address space and used mostly for ``bare metal`` type
3927 development where the reset vectors of the processor are taken into
3934 For x86 machines and some other architectures ``-bios`` will generally
3935 do the right thing with whatever it is given. For other machines the
3936 more strict ``-pflash`` option needs an image that is sized for the
3937 flash device for the given machine type.
3939 Please see the :ref:`system-targets-ref` section of the manual for
3940 more detailed documentation.
3944 DEF("bios", HAS_ARG, QEMU_OPTION_bios, \
3945 "-bios file set the filename for the BIOS\n", QEMU_ARCH_ALL)
3948 Set the filename for the BIOS.
3951 DEF("pflash", HAS_ARG, QEMU_OPTION_pflash,
3952 "-pflash file use 'file' as a parallel flash image\n", QEMU_ARCH_ALL)
3955 Use file as a parallel flash image.
3960 The kernel options were designed to work with Linux kernels although
3961 other things (like hypervisors) can be packaged up as a kernel
3962 executable image. The exact format of a executable image is usually
3963 architecture specific.
3965 The way in which the kernel is started (what address it is loaded at,
3966 what if any information is passed to it via CPU registers, the state
3967 of the hardware when it is started, and so on) is also architecture
3968 specific. Typically it follows the specification laid down by the
3969 Linux kernel for how kernels for that architecture must be started.
3973 DEF("kernel", HAS_ARG, QEMU_OPTION_kernel, \
3974 "-kernel bzImage use 'bzImage' as kernel image\n", QEMU_ARCH_ALL)
3977 Use bzImage as kernel image. The kernel can be either a Linux kernel
3978 or in multiboot format.
3981 DEF("append", HAS_ARG, QEMU_OPTION_append, \
3982 "-append cmdline use 'cmdline' as kernel command line\n", QEMU_ARCH_ALL)
3985 Use cmdline as kernel command line
3988 DEF("initrd", HAS_ARG, QEMU_OPTION_initrd, \
3989 "-initrd file use 'file' as initial ram disk\n", QEMU_ARCH_ALL)
3993 Use file as initial ram disk.
3995 ``-initrd "file1 arg=foo,file2"``
3996 This syntax is only available with multiboot.
3998 Use file1 and file2 as modules and pass ``arg=foo`` as parameter to the
3999 first module. Commas can be provided in module parameters by doubling
4000 them on the command line to escape them:
4002 ``-initrd "bzImage earlyprintk=xen,,keep root=/dev/xvda1,initrd.img"``
4003 Multiboot only. Use bzImage as the first module with
4004 "``earlyprintk=xen,keep root=/dev/xvda1``" as its command line,
4005 and initrd.img as the second module.
4009 DEF("dtb", HAS_ARG, QEMU_OPTION_dtb, \
4010 "-dtb file use 'file' as device tree image\n", QEMU_ARCH_ALL)
4013 Use file as a device tree binary (dtb) image and pass it to the
4019 Finally you can also manually load images directly into the address
4020 space of the guest. This is most useful for developers who already
4021 know the layout of their guest and take care to ensure something sane
4022 will happen when the reset vector executes.
4024 The generic loader can be invoked by using the loader device:
4026 ``-device loader,addr=<addr>,data=<data>,data-len=<data-len>[,data-be=<data-be>][,cpu-num=<cpu-num>]``
4028 there is also the guest loader which operates in a similar way but
4029 tweaks the DTB so a hypervisor loaded via ``-kernel`` can find where
4032 ``-device guest-loader,addr=<addr>[,kernel=<path>,[bootargs=<arguments>]][,initrd=<path>]``
4038 DEFHEADING(Debug/Expert options:)
4040 DEF("compat", HAS_ARG, QEMU_OPTION_compat,
4041 "-compat [deprecated-input=accept|reject|crash][,deprecated-output=accept|hide]\n"
4042 " Policy for handling deprecated management interfaces\n"
4043 "-compat [unstable-input=accept|reject|crash][,unstable-output=accept|hide]\n"
4044 " Policy for handling unstable management interfaces\n",
4047 ``-compat [deprecated-input=@var{input-policy}][,deprecated-output=@var{output-policy}]``
4048 Set policy for handling deprecated management interfaces (experimental):
4050 ``deprecated-input=accept`` (default)
4051 Accept deprecated commands and arguments
4052 ``deprecated-input=reject``
4053 Reject deprecated commands and arguments
4054 ``deprecated-input=crash``
4055 Crash on deprecated commands and arguments
4056 ``deprecated-output=accept`` (default)
4057 Emit deprecated command results and events
4058 ``deprecated-output=hide``
4059 Suppress deprecated command results and events
4061 Limitation: covers only syntactic aspects of QMP.
4063 ``-compat [unstable-input=@var{input-policy}][,unstable-output=@var{output-policy}]``
4064 Set policy for handling unstable management interfaces (experimental):
4066 ``unstable-input=accept`` (default)
4067 Accept unstable commands and arguments
4068 ``unstable-input=reject``
4069 Reject unstable commands and arguments
4070 ``unstable-input=crash``
4071 Crash on unstable commands and arguments
4072 ``unstable-output=accept`` (default)
4073 Emit unstable command results and events
4074 ``unstable-output=hide``
4075 Suppress unstable command results and events
4077 Limitation: covers only syntactic aspects of QMP.
4080 DEF("fw_cfg", HAS_ARG, QEMU_OPTION_fwcfg,
4081 "-fw_cfg [name=]<name>,file=<file>\n"
4082 " add named fw_cfg entry with contents from file\n"
4083 "-fw_cfg [name=]<name>,string=<str>\n"
4084 " add named fw_cfg entry with contents from string\n",
4087 ``-fw_cfg [name=]name,file=file``
4088 Add named fw\_cfg entry with contents from file file.
4090 ``-fw_cfg [name=]name,string=str``
4091 Add named fw\_cfg entry with contents from string str.
4093 The terminating NUL character of the contents of str will not be
4094 included as part of the fw\_cfg item data. To insert contents with
4095 embedded NUL characters, you have to use the file parameter.
4097 The fw\_cfg entries are passed by QEMU through to the guest.
4103 -fw_cfg name=opt/com.mycompany/blob,file=./my_blob.bin
4105 creates an fw\_cfg entry named opt/com.mycompany/blob with contents
4106 from ./my\_blob.bin.
4109 DEF("serial", HAS_ARG, QEMU_OPTION_serial, \
4110 "-serial dev redirect the serial port to char device 'dev'\n",
4114 Redirect the virtual serial port to host character device dev. The
4115 default device is ``vc`` in graphical mode and ``stdio`` in non
4118 This option can be used several times to simulate up to 4 serial
4121 Use ``-serial none`` to disable all serial ports.
4123 Available character devices are:
4126 Virtual console. Optionally, a width and height can be given in
4133 It is also possible to specify width or height in characters:
4140 [Linux only] Pseudo TTY (a new PTY is automatically allocated)
4143 No device is allocated.
4149 Use a named character device defined with the ``-chardev``
4153 [Linux only] Use host tty, e.g. ``/dev/ttyS0``. The host serial
4154 port parameters are set according to the emulated ones.
4157 [Linux only, parallel port only] Use host parallel port N.
4158 Currently SPP and EPP parallel port features can be used.
4161 Write output to filename. No character can be read.
4164 [Unix only] standard input/output
4170 [Windows only] Use host serial port n
4172 ``udp:[remote_host]:remote_port[@[src_ip]:src_port]``
4173 This implements UDP Net Console. When remote\_host or src\_ip
4174 are not specified they default to ``0.0.0.0``. When not using a
4175 specified src\_port a random port is automatically chosen.
4177 If you just want a simple readonly console you can use
4178 ``netcat`` or ``nc``, by starting QEMU with:
4179 ``-serial udp::4555`` and nc as: ``nc -u -l -p 4555``. Any time
4180 QEMU writes something to that port it will appear in the
4183 If you plan to send characters back via netconsole or you want
4184 to stop and start QEMU a lot of times, you should have QEMU use
4185 the same source port each time by using something like ``-serial
4186 udp::4555@:4556`` to QEMU. Another approach is to use a patched
4187 version of netcat which can listen to a TCP port and send and
4188 receive characters via udp. If you have a patched version of
4189 netcat which activates telnet remote echo and single char
4190 transfer, then you can use the following options to set up a
4191 netcat redirector to allow telnet on port 5555 to access the
4195 -serial udp::4555@:4556
4198 -u -P 4555 -L 0.0.0.0:4556 -t -p 5555 -I -T
4203 ``tcp:[host]:port[,server=on|off][,wait=on|off][,nodelay=on|off][,reconnect=seconds]``
4204 The TCP Net Console has two modes of operation. It can send the
4205 serial I/O to a location or wait for a connection from a
4206 location. By default the TCP Net Console is sent to host at the
4207 port. If you use the ``server=on`` option QEMU will wait for a client
4208 socket application to connect to the port before continuing,
4209 unless the ``wait=on|off`` option was specified. The ``nodelay=on|off``
4210 option disables the Nagle buffering algorithm. The ``reconnect=on``
4211 option only applies if ``server=no`` is set, if the connection goes
4212 down it will attempt to reconnect at the given interval. If host
4213 is omitted, 0.0.0.0 is assumed. Only one TCP connection at a
4214 time is accepted. You can use ``telnet=on`` to connect to the
4215 corresponding character device.
4217 ``Example to send tcp console to 192.168.0.2 port 4444``
4218 -serial tcp:192.168.0.2:4444
4220 ``Example to listen and wait on port 4444 for connection``
4221 -serial tcp::4444,server=on
4223 ``Example to not wait and listen on ip 192.168.0.100 port 4444``
4224 -serial tcp:192.168.0.100:4444,server=on,wait=off
4226 ``telnet:host:port[,server=on|off][,wait=on|off][,nodelay=on|off]``
4227 The telnet protocol is used instead of raw tcp sockets. The
4228 options work the same as if you had specified ``-serial tcp``.
4229 The difference is that the port acts like a telnet server or
4230 client using telnet option negotiation. This will also allow you
4231 to send the MAGIC\_SYSRQ sequence if you use a telnet that
4232 supports sending the break sequence. Typically in unix telnet
4233 you do it with Control-] and then type "send break" followed by
4234 pressing the enter key.
4236 ``websocket:host:port,server=on[,wait=on|off][,nodelay=on|off]``
4237 The WebSocket protocol is used instead of raw tcp socket. The
4238 port acts as a WebSocket server. Client mode is not supported.
4240 ``unix:path[,server=on|off][,wait=on|off][,reconnect=seconds]``
4241 A unix domain socket is used instead of a tcp socket. The option
4242 works the same as if you had specified ``-serial tcp`` except
4243 the unix domain socket path is used for connections.
4246 This is a special option to allow the monitor to be multiplexed
4247 onto another serial port. The monitor is accessed with key
4248 sequence of Control-a and then pressing c. dev\_string should be
4249 any one of the serial devices specified above. An example to
4250 multiplex the monitor onto a telnet server listening on port
4253 ``-serial mon:telnet::4444,server=on,wait=off``
4255 When the monitor is multiplexed to stdio in this way, Ctrl+C
4256 will not terminate QEMU any more but will be passed to the guest
4260 Braille device. This will use BrlAPI to display the braille
4261 output on a real or fake device.
4264 Three button serial mouse. Configure the guest to use Microsoft
4268 DEF("parallel", HAS_ARG, QEMU_OPTION_parallel, \
4269 "-parallel dev redirect the parallel port to char device 'dev'\n",
4273 Redirect the virtual parallel port to host device dev (same devices
4274 as the serial port). On Linux hosts, ``/dev/parportN`` can be used
4275 to use hardware devices connected on the corresponding host parallel
4278 This option can be used several times to simulate up to 3 parallel
4281 Use ``-parallel none`` to disable all parallel ports.
4284 DEF("monitor", HAS_ARG, QEMU_OPTION_monitor, \
4285 "-monitor dev redirect the monitor to char device 'dev'\n",
4289 Redirect the monitor to host device dev (same devices as the serial
4290 port). The default device is ``vc`` in graphical mode and ``stdio``
4291 in non graphical mode. Use ``-monitor none`` to disable the default
4294 DEF("qmp", HAS_ARG, QEMU_OPTION_qmp, \
4295 "-qmp dev like -monitor but opens in 'control' mode\n",
4299 Like ``-monitor`` but opens in 'control' mode. For example, to make
4300 QMP available on localhost port 4444::
4302 -qmp tcp:localhost:4444,server=on,wait=off
4304 Not all options are configurable via this syntax; for maximum
4305 flexibility use the ``-mon`` option and an accompanying ``-chardev``.
4308 DEF("qmp-pretty", HAS_ARG, QEMU_OPTION_qmp_pretty, \
4309 "-qmp-pretty dev like -qmp but uses pretty JSON formatting\n",
4313 Like ``-qmp`` but uses pretty JSON formatting.
4316 DEF("mon", HAS_ARG, QEMU_OPTION_mon, \
4317 "-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]\n", QEMU_ARCH_ALL)
4319 ``-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]``
4320 Set up a monitor connected to the chardev ``name``.
4321 QEMU supports two monitors: the Human Monitor Protocol
4322 (HMP; for human interaction), and the QEMU Monitor Protocol
4323 (QMP; a JSON RPC-style protocol).
4324 The default is HMP; ``mode=control`` selects QMP instead.
4325 ``pretty`` is only valid when ``mode=control``,
4326 turning on JSON pretty printing to ease
4327 human reading and debugging.
4331 -chardev socket,id=mon1,host=localhost,port=4444,server=on,wait=off \
4332 -mon chardev=mon1,mode=control,pretty=on
4334 enables the QMP monitor on localhost port 4444 with pretty-printing.
4337 DEF("debugcon", HAS_ARG, QEMU_OPTION_debugcon, \
4338 "-debugcon dev redirect the debug console to char device 'dev'\n",
4342 Redirect the debug console to host device dev (same devices as the
4343 serial port). The debug console is an I/O port which is typically
4344 port 0xe9; writing to that I/O port sends output to this device. The
4345 default device is ``vc`` in graphical mode and ``stdio`` in non
4349 DEF("pidfile", HAS_ARG, QEMU_OPTION_pidfile, \
4350 "-pidfile file write PID to 'file'\n", QEMU_ARCH_ALL)
4353 Store the QEMU process PID in file. It is useful if you launch QEMU
4357 DEF("singlestep", 0, QEMU_OPTION_singlestep, \
4358 "-singlestep deprecated synonym for -accel tcg,one-insn-per-tb=on\n", QEMU_ARCH_ALL)
4361 This is a deprecated synonym for the TCG accelerator property
4362 ``one-insn-per-tb``.
4365 DEF("preconfig", 0, QEMU_OPTION_preconfig, \
4366 "--preconfig pause QEMU before machine is initialized (experimental)\n",
4370 Pause QEMU for interactive configuration before the machine is
4371 created, which allows querying and configuring properties that will
4372 affect machine initialization. Use QMP command 'x-exit-preconfig' to
4373 exit the preconfig state and move to the next state (i.e. run guest
4374 if -S isn't used or pause the second time if -S is used). This
4375 option is experimental.
4378 DEF("S", 0, QEMU_OPTION_S, \
4379 "-S freeze CPU at startup (use 'c' to start execution)\n",
4383 Do not start CPU at startup (you must type 'c' in the monitor).
4386 DEF("overcommit", HAS_ARG, QEMU_OPTION_overcommit,
4387 "-overcommit [mem-lock=on|off][cpu-pm=on|off]\n"
4388 " run qemu with overcommit hints\n"
4389 " mem-lock=on|off controls memory lock support (default: off)\n"
4390 " cpu-pm=on|off controls cpu power management (default: off)\n",
4393 ``-overcommit mem-lock=on|off``
4395 ``-overcommit cpu-pm=on|off``
4396 Run qemu with hints about host resource overcommit. The default is
4397 to assume that host overcommits all resources.
4399 Locking qemu and guest memory can be enabled via ``mem-lock=on``
4400 (disabled by default). This works when host memory is not
4401 overcommitted and reduces the worst-case latency for guest.
4403 Guest ability to manage power state of host cpus (increasing latency
4404 for other processes on the same host cpu, but decreasing latency for
4405 guest) can be enabled via ``cpu-pm=on`` (disabled by default). This
4406 works best when host CPU is not overcommitted. When used, host
4407 estimates of CPU cycle and power utilization will be incorrect, not
4408 taking into account guest idle time.
4411 DEF("gdb", HAS_ARG, QEMU_OPTION_gdb, \
4412 "-gdb dev accept gdb connection on 'dev'. (QEMU defaults to starting\n"
4413 " the guest without waiting for gdb to connect; use -S too\n"
4414 " if you want it to not start execution.)\n",
4418 Accept a gdb connection on device dev (see the :ref:`GDB usage` chapter
4419 in the System Emulation Users Guide). Note that this option does not pause QEMU
4420 execution -- if you want QEMU to not start the guest until you
4421 connect with gdb and issue a ``continue`` command, you will need to
4422 also pass the ``-S`` option to QEMU.
4424 The most usual configuration is to listen on a local TCP socket::
4428 but you can specify other backends; UDP, pseudo TTY, or even stdio
4429 are all reasonable use cases. For example, a stdio connection
4430 allows you to start QEMU from within gdb and establish the
4431 connection via a pipe:
4435 (gdb) target remote | exec |qemu_system| -gdb stdio ...
4438 DEF("s", 0, QEMU_OPTION_s, \
4439 "-s shorthand for -gdb tcp::" DEFAULT_GDBSTUB_PORT "\n",
4443 Shorthand for -gdb tcp::1234, i.e. open a gdbserver on TCP port 1234
4444 (see the :ref:`GDB usage` chapter in the System Emulation Users Guide).
4447 DEF("d", HAS_ARG, QEMU_OPTION_d, \
4448 "-d item1,... enable logging of specified items (use '-d help' for a list of log items)\n",
4452 Enable logging of specified items. Use '-d help' for a list of log
4456 DEF("D", HAS_ARG, QEMU_OPTION_D, \
4457 "-D logfile output log to logfile (default stderr)\n",
4461 Output log in logfile instead of to stderr
4464 DEF("dfilter", HAS_ARG, QEMU_OPTION_DFILTER, \
4465 "-dfilter range,.. filter debug output to range of addresses (useful for -d cpu,exec,etc..)\n",
4468 ``-dfilter range1[,...]``
4469 Filter debug output to that relevant to a range of target addresses.
4470 The filter spec can be either start+size, start-size or start..end
4471 where start end and size are the addresses and sizes required. For
4476 -dfilter 0x8000..0x8fff,0xffffffc000080000+0x200,0xffffffc000060000-0x1000
4478 Will dump output for any code in the 0x1000 sized block starting at
4479 0x8000 and the 0x200 sized block starting at 0xffffffc000080000 and
4480 another 0x1000 sized block starting at 0xffffffc00005f000.
4483 DEF("seed", HAS_ARG, QEMU_OPTION_seed, \
4484 "-seed number seed the pseudo-random number generator\n",
4488 Force the guest to use a deterministic pseudo-random number
4489 generator, seeded with number. This does not affect crypto routines
4493 DEF("L", HAS_ARG, QEMU_OPTION_L, \
4494 "-L path set the directory for the BIOS, VGA BIOS and keymaps\n",
4498 Set the directory for the BIOS, VGA BIOS and keymaps.
4500 To list all the data directories, use ``-L help``.
4503 DEF("enable-kvm", 0, QEMU_OPTION_enable_kvm, \
4504 "-enable-kvm enable KVM full virtualization support\n",
4505 QEMU_ARCH_ARM | QEMU_ARCH_I386 | QEMU_ARCH_MIPS | QEMU_ARCH_PPC |
4506 QEMU_ARCH_RISCV | QEMU_ARCH_S390X)
4509 Enable KVM full virtualization support. This option is only
4510 available if KVM support is enabled when compiling.
4513 DEF("xen-domid", HAS_ARG, QEMU_OPTION_xen_domid,
4514 "-xen-domid id specify xen guest domain id\n",
4515 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4516 DEF("xen-attach", 0, QEMU_OPTION_xen_attach,
4517 "-xen-attach attach to existing xen domain\n"
4518 " libxl will use this when starting QEMU\n",
4519 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4520 DEF("xen-domid-restrict", 0, QEMU_OPTION_xen_domid_restrict,
4521 "-xen-domid-restrict restrict set of available xen operations\n"
4522 " to specified domain id. (Does not affect\n"
4523 " xenpv machine type).\n",
4524 QEMU_ARCH_ARM | QEMU_ARCH_I386)
4527 Specify xen guest domain id (XEN only).
4530 Attach to existing xen domain. libxl will use this when starting
4531 QEMU (XEN only). Restrict set of available xen operations to
4532 specified domain id (XEN only).
4535 DEF("no-reboot", 0, QEMU_OPTION_no_reboot, \
4536 "-no-reboot exit instead of rebooting\n", QEMU_ARCH_ALL)
4539 Exit instead of rebooting.
4542 DEF("no-shutdown", 0, QEMU_OPTION_no_shutdown, \
4543 "-no-shutdown stop before shutdown\n", QEMU_ARCH_ALL)
4546 Don't exit QEMU on guest shutdown, but instead only stop the
4547 emulation. This allows for instance switching to monitor to commit
4548 changes to the disk image.
4551 DEF("action", HAS_ARG, QEMU_OPTION_action,
4552 "-action reboot=reset|shutdown\n"
4553 " action when guest reboots [default=reset]\n"
4554 "-action shutdown=poweroff|pause\n"
4555 " action when guest shuts down [default=poweroff]\n"
4556 "-action panic=pause|shutdown|exit-failure|none\n"
4557 " action when guest panics [default=shutdown]\n"
4558 "-action watchdog=reset|shutdown|poweroff|inject-nmi|pause|debug|none\n"
4559 " action when watchdog fires [default=reset]\n",
4562 ``-action event=action``
4563 The action parameter serves to modify QEMU's default behavior when
4564 certain guest events occur. It provides a generic method for specifying the
4565 same behaviors that are modified by the ``-no-reboot`` and ``-no-shutdown``
4570 ``-action panic=none``
4571 ``-action reboot=shutdown,shutdown=pause``
4572 ``-device i6300esb -action watchdog=pause``
4576 DEF("loadvm", HAS_ARG, QEMU_OPTION_loadvm, \
4577 "-loadvm [tag|id]\n" \
4578 " start right away with a saved state (loadvm in monitor)\n",
4582 Start right away with a saved state (``loadvm`` in monitor)
4586 DEF("daemonize", 0, QEMU_OPTION_daemonize, \
4587 "-daemonize daemonize QEMU after initializing\n", QEMU_ARCH_ALL)
4591 Daemonize the QEMU process after initialization. QEMU will not
4592 detach from standard IO until it is ready to receive connections on
4593 any of its devices. This option is a useful way for external
4594 programs to launch QEMU without having to cope with initialization
4598 DEF("option-rom", HAS_ARG, QEMU_OPTION_option_rom, \
4599 "-option-rom rom load a file, rom, into the option ROM space\n",
4602 ``-option-rom file``
4603 Load the contents of file as an option ROM. This option is useful to
4604 load things like EtherBoot.
4607 DEF("rtc", HAS_ARG, QEMU_OPTION_rtc, \
4608 "-rtc [base=utc|localtime|<datetime>][,clock=host|rt|vm][,driftfix=none|slew]\n" \
4609 " set the RTC base and clock, enable drift fix for clock ticks (x86 only)\n",
4613 ``-rtc [base=utc|localtime|datetime][,clock=host|rt|vm][,driftfix=none|slew]``
4614 Specify ``base`` as ``utc`` or ``localtime`` to let the RTC start at
4615 the current UTC or local time, respectively. ``localtime`` is
4616 required for correct date in MS-DOS or Windows. To start at a
4617 specific point in time, provide datetime in the format
4618 ``2006-06-17T16:01:21`` or ``2006-06-17``. The default base is UTC.
4620 By default the RTC is driven by the host system time. This allows
4621 using of the RTC as accurate reference clock inside the guest,
4622 specifically if the host time is smoothly following an accurate
4623 external reference clock, e.g. via NTP. If you want to isolate the
4624 guest time from the host, you can set ``clock`` to ``rt`` instead,
4625 which provides a host monotonic clock if host support it. To even
4626 prevent the RTC from progressing during suspension, you can set
4627 ``clock`` to ``vm`` (virtual clock). '\ ``clock=vm``\ ' is
4628 recommended especially in icount mode in order to preserve
4629 determinism; however, note that in icount mode the speed of the
4630 virtual clock is variable and can in general differ from the host
4633 Enable ``driftfix`` (i386 targets only) if you experience time drift
4634 problems, specifically with Windows' ACPI HAL. This option will try
4635 to figure out how many timer interrupts were not processed by the
4636 Windows guest and will re-inject them.
4639 DEF("icount", HAS_ARG, QEMU_OPTION_icount, \
4640 "-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=<filename>[,rrsnapshot=<snapshot>]]\n" \
4641 " enable virtual instruction counter with 2^N clock ticks per\n" \
4642 " instruction, enable aligning the host and virtual clocks\n" \
4643 " or disable real time cpu sleeping, and optionally enable\n" \
4644 " record-and-replay mode\n", QEMU_ARCH_ALL)
4646 ``-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=filename[,rrsnapshot=snapshot]]``
4647 Enable virtual instruction counter. The virtual cpu will execute one
4648 instruction every 2^N ns of virtual time. If ``auto`` is specified
4649 then the virtual cpu speed will be automatically adjusted to keep
4650 virtual time within a few seconds of real time.
4652 Note that while this option can give deterministic behavior, it does
4653 not provide cycle accurate emulation. Modern CPUs contain
4654 superscalar out of order cores with complex cache hierarchies. The
4655 number of instructions executed often has little or no correlation
4656 with actual performance.
4658 When the virtual cpu is sleeping, the virtual time will advance at
4659 default speed unless ``sleep=on`` is specified. With
4660 ``sleep=on``, the virtual time will jump to the next timer
4661 deadline instantly whenever the virtual cpu goes to sleep mode and
4662 will not advance if no timer is enabled. This behavior gives
4663 deterministic execution times from the guest point of view.
4664 The default if icount is enabled is ``sleep=off``.
4665 ``sleep=on`` cannot be used together with either ``shift=auto``
4668 ``align=on`` will activate the delay algorithm which will try to
4669 synchronise the host clock and the virtual clock. The goal is to
4670 have a guest running at the real frequency imposed by the shift
4671 option. Whenever the guest clock is behind the host clock and if
4672 ``align=on`` is specified then we print a message to the user to
4673 inform about the delay. Currently this option does not work when
4674 ``shift`` is ``auto``. Note: The sync algorithm will work for those
4675 shift values for which the guest clock runs ahead of the host clock.
4676 Typically this happens when the shift value is high (how high
4677 depends on the host machine). The default if icount is enabled
4680 When the ``rr`` option is specified deterministic record/replay is
4681 enabled. The ``rrfile=`` option must also be provided to
4682 specify the path to the replay log. In record mode data is written
4683 to this file, and in replay mode it is read back.
4684 If the ``rrsnapshot`` option is given then it specifies a VM snapshot
4685 name. In record mode, a new VM snapshot with the given name is created
4686 at the start of execution recording. In replay mode this option
4687 specifies the snapshot name used to load the initial VM state.
4690 DEF("watchdog-action", HAS_ARG, QEMU_OPTION_watchdog_action, \
4691 "-watchdog-action reset|shutdown|poweroff|inject-nmi|pause|debug|none\n" \
4692 " action when watchdog fires [default=reset]\n",
4695 ``-watchdog-action action``
4696 The action controls what QEMU will do when the watchdog timer
4697 expires. The default is ``reset`` (forcefully reset the guest).
4698 Other possible actions are: ``shutdown`` (attempt to gracefully
4699 shutdown the guest), ``poweroff`` (forcefully poweroff the guest),
4700 ``inject-nmi`` (inject a NMI into the guest), ``pause`` (pause the
4701 guest), ``debug`` (print a debug message and continue), or ``none``
4704 Note that the ``shutdown`` action requires that the guest responds
4705 to ACPI signals, which it may not be able to do in the sort of
4706 situations where the watchdog would have expired, and thus
4707 ``-watchdog-action shutdown`` is not recommended for production use.
4711 ``-device i6300esb -watchdog-action pause``
4715 DEF("echr", HAS_ARG, QEMU_OPTION_echr, \
4716 "-echr chr set terminal escape character instead of ctrl-a\n",
4719 ``-echr numeric_ascii_value``
4720 Change the escape character used for switching to the monitor when
4721 using monitor and serial sharing. The default is ``0x01`` when using
4722 the ``-nographic`` option. ``0x01`` is equal to pressing
4723 ``Control-a``. You can select a different character from the ascii
4724 control keys where 1 through 26 map to Control-a through Control-z.
4725 For instance you could use the either of the following to change the
4726 escape character to Control-t.
4728 ``-echr 0x14``; \ ``-echr 20``
4732 DEF("incoming", HAS_ARG, QEMU_OPTION_incoming, \
4733 "-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]\n" \
4734 "-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]\n" \
4735 "-incoming unix:socketpath\n" \
4736 " prepare for incoming migration, listen on\n" \
4737 " specified protocol and socket address\n" \
4738 "-incoming fd:fd\n" \
4739 "-incoming file:filename[,offset=offset]\n" \
4740 "-incoming exec:cmdline\n" \
4741 " accept incoming migration on given file descriptor\n" \
4742 " or from given external command\n" \
4743 "-incoming defer\n" \
4744 " wait for the URI to be specified via migrate_incoming\n",
4747 ``-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]``
4749 ``-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]``
4750 Prepare for incoming migration, listen on a given tcp port.
4752 ``-incoming unix:socketpath``
4753 Prepare for incoming migration, listen on a given unix socket.
4756 Accept incoming migration from a given file descriptor.
4758 ``-incoming file:filename[,offset=offset]``
4759 Accept incoming migration from a given file starting at offset.
4760 offset allows the common size suffixes, or a 0x prefix, but not both.
4762 ``-incoming exec:cmdline``
4763 Accept incoming migration as an output from specified external
4767 Wait for the URI to be specified via migrate\_incoming. The monitor
4768 can be used to change settings (such as migration parameters) prior
4769 to issuing the migrate\_incoming to allow the migration to begin.
4772 DEF("only-migratable", 0, QEMU_OPTION_only_migratable, \
4773 "-only-migratable allow only migratable devices\n", QEMU_ARCH_ALL)
4775 ``-only-migratable``
4776 Only allow migratable devices. Devices will not be allowed to enter
4777 an unmigratable state.
4780 DEF("nodefaults", 0, QEMU_OPTION_nodefaults, \
4781 "-nodefaults don't create default devices\n", QEMU_ARCH_ALL)
4784 Don't create default devices. Normally, QEMU sets the default
4785 devices like serial port, parallel port, virtual console, monitor
4786 device, VGA adapter, floppy and CD-ROM drive and others. The
4787 ``-nodefaults`` option will disable all those default devices.
4791 DEF("chroot", HAS_ARG, QEMU_OPTION_chroot, \
4792 "-chroot dir chroot to dir just before starting the VM (deprecated)\n",
4797 Deprecated, use '-run-with chroot=...' instead.
4798 Immediately before starting guest execution, chroot to the specified
4799 directory. Especially useful in combination with -runas.
4803 DEF("runas", HAS_ARG, QEMU_OPTION_runas, \
4804 "-runas user change to user id user just before starting the VM\n" \
4805 " user can be numeric uid:gid instead\n",
4810 Immediately before starting guest execution, drop root privileges,
4811 switching to the specified user.
4814 DEF("prom-env", HAS_ARG, QEMU_OPTION_prom_env,
4815 "-prom-env variable=value\n"
4816 " set OpenBIOS nvram variables\n",
4817 QEMU_ARCH_PPC | QEMU_ARCH_SPARC)
4819 ``-prom-env variable=value``
4820 Set OpenBIOS nvram variable to given value (PPC, SPARC only).
4824 qemu-system-sparc -prom-env 'auto-boot?=false' \
4825 -prom-env 'boot-device=sd(0,2,0):d' -prom-env 'boot-args=linux single'
4829 qemu-system-ppc -prom-env 'auto-boot?=false' \
4830 -prom-env 'boot-device=hd:2,\yaboot' \
4831 -prom-env 'boot-args=conf=hd:2,\yaboot.conf'
4833 DEF("semihosting", 0, QEMU_OPTION_semihosting,
4834 "-semihosting semihosting mode\n",
4835 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4836 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4839 Enable :ref:`Semihosting` mode (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V only).
4842 Note that this allows guest direct access to the host filesystem, so
4843 should only be used with a trusted guest OS.
4845 See the -semihosting-config option documentation for further
4846 information about the facilities this enables.
4848 DEF("semihosting-config", HAS_ARG, QEMU_OPTION_semihosting_config,
4849 "-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,userspace=on|off][,arg=str[,...]]\n" \
4850 " semihosting configuration\n",
4851 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA |
4852 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV)
4854 ``-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,userspace=on|off][,arg=str[,...]]``
4855 Enable and configure :ref:`Semihosting` (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V
4859 Note that this allows guest direct access to the host filesystem, so
4860 should only be used with a trusted guest OS.
4862 ``target=native|gdb|auto``
4863 Defines where the semihosting calls will be addressed, to QEMU
4864 (``native``) or to GDB (``gdb``). The default is ``auto``, which
4865 means ``gdb`` during debug sessions and ``native`` otherwise.
4868 Send the output to a chardev backend output for native or auto
4869 output when not in gdb
4871 ``userspace=on|off``
4872 Allows code running in guest userspace to access the semihosting
4873 interface. The default is that only privileged guest code can
4874 make semihosting calls. Note that setting ``userspace=on`` should
4875 only be used if all guest code is trusted (for example, in
4876 bare-metal test case code).
4878 ``arg=str1,arg=str2,...``
4879 Allows the user to pass input arguments, and can be used
4880 multiple times to build up a list. The old-style
4881 ``-kernel``/``-append`` method of passing a command line is
4882 still supported for backward compatibility. If both the
4883 ``--semihosting-config arg`` and the ``-kernel``/``-append`` are
4884 specified, the former is passed to semihosting as it always
4887 DEF("old-param", 0, QEMU_OPTION_old_param,
4888 "-old-param old param mode\n", QEMU_ARCH_ARM)
4891 Old param mode (ARM only).
4894 DEF("sandbox", HAS_ARG, QEMU_OPTION_sandbox, \
4895 "-sandbox on[,obsolete=allow|deny][,elevateprivileges=allow|deny|children]\n" \
4896 " [,spawn=allow|deny][,resourcecontrol=allow|deny]\n" \
4897 " Enable seccomp mode 2 system call filter (default 'off').\n" \
4898 " use 'obsolete' to allow obsolete system calls that are provided\n" \
4899 " by the kernel, but typically no longer used by modern\n" \
4900 " C library implementations.\n" \
4901 " use 'elevateprivileges' to allow or deny the QEMU process ability\n" \
4902 " to elevate privileges using set*uid|gid system calls.\n" \
4903 " The value 'children' will deny set*uid|gid system calls for\n" \
4904 " main QEMU process but will allow forks and execves to run unprivileged\n" \
4905 " use 'spawn' to avoid QEMU to spawn new threads or processes by\n" \
4906 " blocking *fork and execve\n" \
4907 " use 'resourcecontrol' to disable process affinity and schedular priority\n",
4910 ``-sandbox arg[,obsolete=string][,elevateprivileges=string][,spawn=string][,resourcecontrol=string]``
4911 Enable Seccomp mode 2 system call filter. 'on' will enable syscall
4912 filtering and 'off' will disable it. The default is 'off'.
4915 Enable Obsolete system calls
4917 ``elevateprivileges=string``
4918 Disable set\*uid\|gid system calls
4921 Disable \*fork and execve
4923 ``resourcecontrol=string``
4924 Disable process affinity and schedular priority
4927 DEF("readconfig", HAS_ARG, QEMU_OPTION_readconfig,
4928 "-readconfig <file>\n"
4929 " read config file\n", QEMU_ARCH_ALL)
4931 ``-readconfig file``
4932 Read device configuration from file. This approach is useful when
4933 you want to spawn QEMU process with many command line options but
4934 you don't want to exceed the command line character limit.
4937 DEF("no-user-config", 0, QEMU_OPTION_nouserconfig,
4939 " do not load default user-provided config files at startup\n",
4943 The ``-no-user-config`` option makes QEMU not load any of the
4944 user-provided config files on sysconfdir.
4947 DEF("trace", HAS_ARG, QEMU_OPTION_trace,
4948 "-trace [[enable=]<pattern>][,events=<file>][,file=<file>]\n"
4949 " specify tracing options\n",
4952 ``-trace [[enable=]pattern][,events=file][,file=file]``
4953 .. include:: ../qemu-option-trace.rst.inc
4956 DEF("plugin", HAS_ARG, QEMU_OPTION_plugin,
4957 "-plugin [file=]<file>[,<argname>=<argvalue>]\n"
4961 ``-plugin file=file[,argname=argvalue]``
4965 Load the given plugin from a shared library file.
4967 ``argname=argvalue``
4968 Argument passed to the plugin. (Can be given multiple times.)
4972 DEF("qtest", HAS_ARG, QEMU_OPTION_qtest, "", QEMU_ARCH_ALL)
4973 DEF("qtest-log", HAS_ARG, QEMU_OPTION_qtest_log, "", QEMU_ARCH_ALL)
4976 DEF("async-teardown", 0, QEMU_OPTION_asyncteardown,
4977 "-async-teardown enable asynchronous teardown\n",
4981 This option is deprecated and should no longer be used. The new option
4982 ``-run-with async-teardown=on`` is a replacement.
4986 DEF("run-with", HAS_ARG, QEMU_OPTION_run_with,
4987 "-run-with [async-teardown=on|off][,chroot=dir]\n"
4988 " Set miscellaneous QEMU process lifecycle options:\n"
4989 " async-teardown=on enables asynchronous teardown (Linux only)\n"
4990 " chroot=dir chroot to dir just before starting the VM\n",
4993 ``-run-with [async-teardown=on|off][,chroot=dir]``
4994 Set QEMU process lifecycle options.
4996 ``async-teardown=on`` enables asynchronous teardown. A new process called
4997 "cleanup/<QEMU_PID>" will be created at startup sharing the address
4998 space with the main QEMU process, using clone. It will wait for the
4999 main QEMU process to terminate completely, and then exit. This allows
5000 QEMU to terminate very quickly even if the guest was huge, leaving the
5001 teardown of the address space to the cleanup process. Since the cleanup
5002 process shares the same cgroups as the main QEMU process, accounting is
5003 performed correctly. This only works if the cleanup process is not
5004 forcefully killed with SIGKILL before the main QEMU process has
5005 terminated completely.
5007 ``chroot=dir`` can be used for doing a chroot to the specified directory
5008 immediately before starting the guest execution. This is especially useful
5009 in combination with -runas.
5013 DEF("msg", HAS_ARG, QEMU_OPTION_msg,
5014 "-msg [timestamp[=on|off]][,guest-name=[on|off]]\n"
5015 " control error message format\n"
5016 " timestamp=on enables timestamps (default: off)\n"
5017 " guest-name=on enables guest name prefix but only if\n"
5018 " -name guest option is set (default: off)\n",
5021 ``-msg [timestamp[=on|off]][,guest-name[=on|off]]``
5022 Control error message format.
5024 ``timestamp=on|off``
5025 Prefix messages with a timestamp. Default is off.
5027 ``guest-name=on|off``
5028 Prefix messages with guest name but only if -name guest option is set
5029 otherwise the option is ignored. Default is off.
5032 DEF("dump-vmstate", HAS_ARG, QEMU_OPTION_dump_vmstate,
5033 "-dump-vmstate <file>\n"
5034 " Output vmstate information in JSON format to file.\n"
5035 " Use the scripts/vmstate-static-checker.py file to\n"
5036 " check for possible regressions in migration code\n"
5037 " by comparing two such vmstate dumps.\n",
5040 ``-dump-vmstate file``
5041 Dump json-encoded vmstate information for current machine type to
5045 DEF("enable-sync-profile", 0, QEMU_OPTION_enable_sync_profile,
5046 "-enable-sync-profile\n"
5047 " enable synchronization profiling\n",
5050 ``-enable-sync-profile``
5051 Enable synchronization profiling.
5054 #if defined(CONFIG_TCG) && defined(CONFIG_LINUX)
5055 DEF("perfmap", 0, QEMU_OPTION_perfmap,
5056 "-perfmap generate a /tmp/perf-${pid}.map file for perf\n",
5060 Generate a map file for Linux perf tools that will allow basic profiling
5061 information to be broken down into basic blocks.
5064 DEF("jitdump", 0, QEMU_OPTION_jitdump,
5065 "-jitdump generate a jit-${pid}.dump file for perf\n",
5069 Generate a dump file for Linux perf tools that maps basic blocks to symbol
5070 names, line numbers and JITted code.
5076 DEFHEADING(Generic object creation:)
5078 DEF("object", HAS_ARG, QEMU_OPTION_object,
5079 "-object TYPENAME[,PROP1=VALUE1,...]\n"
5080 " create a new object of type TYPENAME setting properties\n"
5081 " in the order they are specified. Note that the 'id'\n"
5082 " property must be set. These objects are placed in the\n"
5083 " '/objects' path.\n",
5086 ``-object typename[,prop1=value1,...]``
5087 Create a new object of type typename setting properties in the order
5088 they are specified. Note that the 'id' property must be set. These
5089 objects are placed in the '/objects' path.
5091 ``-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,offset=offset,readonly=on|off,rom=on|off|auto``
5092 Creates a memory file backend object, which can be used to back
5093 the guest RAM with huge pages.
5095 The ``id`` parameter is a unique ID that will be used to
5096 reference this memory region in other parameters, e.g. ``-numa``,
5097 ``-device nvdimm``, etc.
5099 The ``size`` option provides the size of the memory region, and
5100 accepts common suffixes, e.g. ``500M``.
5102 The ``mem-path`` provides the path to either a shared memory or
5103 huge page filesystem mount.
5105 The ``share`` boolean option determines whether the memory
5106 region is marked as private to QEMU, or shared. The latter
5107 allows a co-operating external process to access the QEMU memory
5110 The ``share`` is also required for pvrdma devices due to
5111 limitations in the RDMA API provided by Linux.
5113 Setting share=on might affect the ability to configure NUMA
5114 bindings for the memory backend under some circumstances, see
5115 Documentation/vm/numa\_memory\_policy.txt on the Linux kernel
5116 source tree for additional details.
5118 Setting the ``discard-data`` boolean option to on indicates that
5119 file contents can be destroyed when QEMU exits, to avoid
5120 unnecessarily flushing data to the backing file. Note that
5121 ``discard-data`` is only an optimization, and QEMU might not
5122 discard file contents if it aborts unexpectedly or is terminated
5125 The ``merge`` boolean option enables memory merge, also known as
5126 MADV\_MERGEABLE, so that Kernel Samepage Merging will consider
5127 the pages for memory deduplication.
5129 Setting the ``dump`` boolean option to off excludes the memory
5130 from core dumps. This feature is also known as MADV\_DONTDUMP.
5132 The ``prealloc`` boolean option enables memory preallocation.
5134 The ``host-nodes`` option binds the memory range to a list of
5137 The ``policy`` option sets the NUMA policy to one of the
5144 prefer the given host node list for allocation
5147 restrict memory allocation to the given host node list
5150 interleave memory allocations across the given host node
5153 The ``align`` option specifies the base address alignment when
5154 QEMU mmap(2) ``mem-path``, and accepts common suffixes, eg
5155 ``2M``. Some backend store specified by ``mem-path`` requires an
5156 alignment different than the default one used by QEMU, eg the
5157 device DAX /dev/dax0.0 requires 2M alignment rather than 4K. In
5158 such cases, users can specify the required alignment via this
5161 The ``offset`` option specifies the offset into the target file
5162 that the region starts at. You can use this parameter to back
5163 multiple regions with a single file.
5165 The ``pmem`` option specifies whether the backing file specified
5166 by ``mem-path`` is in host persistent memory that can be
5167 accessed using the SNIA NVM programming model (e.g. Intel
5168 NVDIMM). If ``pmem`` is set to 'on', QEMU will take necessary
5169 operations to guarantee the persistence of its own writes to
5170 ``mem-path`` (e.g. in vNVDIMM label emulation and live
5171 migration). Also, we will map the backend-file with MAP\_SYNC
5172 flag, which ensures the file metadata is in sync for
5173 ``mem-path`` in case of host crash or a power failure. MAP\_SYNC
5174 requires support from both the host kernel (since Linux kernel
5175 4.15) and the filesystem of ``mem-path`` mounted with DAX
5178 The ``readonly`` option specifies whether the backing file is opened
5179 read-only or read-write (default).
5181 The ``rom`` option specifies whether to create Read Only Memory
5182 (ROM) that cannot be modified by the VM. Any write attempts to such
5183 ROM will be denied. Most use cases want proper RAM instead of ROM.
5184 However, selected use cases, like R/O NVDIMMs, can benefit from
5185 ROM. If set to ``on``, create ROM; if set to ``off``, create
5186 writable RAM; if set to ``auto`` (default), the value of the
5187 ``readonly`` option is used. This option is primarily helpful when
5188 we want to have writable RAM in configurations that would
5189 traditionally create ROM before the ``rom`` option was introduced:
5190 VM templating, where we want to open a file readonly
5191 (``readonly=on``) and mark the memory to be private for QEMU
5192 (``share=off``). For this use case, we need writable RAM instead
5193 of ROM, and want to also set ``rom=off``.
5195 ``-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``
5196 Creates a memory backend object, which can be used to back the
5197 guest RAM. Memory backend objects offer more control than the
5198 ``-m`` option that is traditionally used to define guest RAM.
5199 Please refer to ``memory-backend-file`` for a description of the
5202 ``-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``
5203 Creates an anonymous memory file backend object, which allows
5204 QEMU to share the memory with an external process (e.g. when
5205 using vhost-user). The memory is allocated with memfd and
5206 optional sealing. (Linux only)
5208 The ``seal`` option creates a sealed-file, that will block
5209 further resizing the memory ('on' by default).
5211 The ``hugetlb`` option specify the file to be created resides in
5212 the hugetlbfs filesystem (since Linux 4.14). Used in conjunction
5213 with the ``hugetlb`` option, the ``hugetlbsize`` option specify
5214 the hugetlb page size on systems that support multiple hugetlb
5215 page sizes (it must be a power of 2 value supported by the
5218 In some versions of Linux, the ``hugetlb`` option is
5219 incompatible with the ``seal`` option (requires at least Linux
5222 Please refer to ``memory-backend-file`` for a description of the
5225 The ``share`` boolean option is on by default with memfd.
5227 ``-object rng-builtin,id=id``
5228 Creates a random number generator backend which obtains entropy
5229 from QEMU builtin functions. The ``id`` parameter is a unique ID
5230 that will be used to reference this entropy backend from the
5231 ``virtio-rng`` device. By default, the ``virtio-rng`` device
5232 uses this RNG backend.
5234 ``-object rng-random,id=id,filename=/dev/random``
5235 Creates a random number generator backend which obtains entropy
5236 from a device on the host. The ``id`` parameter is a unique ID
5237 that will be used to reference this entropy backend from the
5238 ``virtio-rng`` device. The ``filename`` parameter specifies
5239 which file to obtain entropy from and if omitted defaults to
5242 ``-object rng-egd,id=id,chardev=chardevid``
5243 Creates a random number generator backend which obtains entropy
5244 from an external daemon running on the host. The ``id``
5245 parameter is a unique ID that will be used to reference this
5246 entropy backend from the ``virtio-rng`` device. The ``chardev``
5247 parameter is the unique ID of a character device backend that
5248 provides the connection to the RNG daemon.
5250 ``-object tls-creds-anon,id=id,endpoint=endpoint,dir=/path/to/cred/dir,verify-peer=on|off``
5251 Creates a TLS anonymous credentials object, which can be used to
5252 provide TLS support on network backends. The ``id`` parameter is
5253 a unique ID which network backends will use to access the
5254 credentials. The ``endpoint`` is either ``server`` or ``client``
5255 depending on whether the QEMU network backend that uses the
5256 credentials will be acting as a client or as a server. If
5257 ``verify-peer`` is enabled (the default) then once the handshake
5258 is completed, the peer credentials will be verified, though this
5259 is a no-op for anonymous credentials.
5261 The dir parameter tells QEMU where to find the credential files.
5262 For server endpoints, this directory may contain a file
5263 dh-params.pem providing diffie-hellman parameters to use for the
5264 TLS server. If the file is missing, QEMU will generate a set of
5265 DH parameters at startup. This is a computationally expensive
5266 operation that consumes random pool entropy, so it is
5267 recommended that a persistent set of parameters be generated
5270 ``-object tls-creds-psk,id=id,endpoint=endpoint,dir=/path/to/keys/dir[,username=username]``
5271 Creates a TLS Pre-Shared Keys (PSK) credentials object, which
5272 can be used to provide TLS support on network backends. The
5273 ``id`` parameter is a unique ID which network backends will use
5274 to access the credentials. The ``endpoint`` is either ``server``
5275 or ``client`` depending on whether the QEMU network backend that
5276 uses the credentials will be acting as a client or as a server.
5277 For clients only, ``username`` is the username which will be
5278 sent to the server. If omitted it defaults to "qemu".
5280 The dir parameter tells QEMU where to find the keys file. It is
5281 called "dir/keys.psk" and contains "username:key" pairs. This
5282 file can most easily be created using the GnuTLS ``psktool``
5285 For server endpoints, dir may also contain a file dh-params.pem
5286 providing diffie-hellman parameters to use for the TLS server.
5287 If the file is missing, QEMU will generate a set of DH
5288 parameters at startup. This is a computationally expensive
5289 operation that consumes random pool entropy, so it is
5290 recommended that a persistent set of parameters be generated up
5293 ``-object tls-creds-x509,id=id,endpoint=endpoint,dir=/path/to/cred/dir,priority=priority,verify-peer=on|off,passwordid=id``
5294 Creates a TLS anonymous credentials object, which can be used to
5295 provide TLS support on network backends. The ``id`` parameter is
5296 a unique ID which network backends will use to access the
5297 credentials. The ``endpoint`` is either ``server`` or ``client``
5298 depending on whether the QEMU network backend that uses the
5299 credentials will be acting as a client or as a server. If
5300 ``verify-peer`` is enabled (the default) then once the handshake
5301 is completed, the peer credentials will be verified. With x509
5302 certificates, this implies that the clients must be provided
5303 with valid client certificates too.
5305 The dir parameter tells QEMU where to find the credential files.
5306 For server endpoints, this directory may contain a file
5307 dh-params.pem providing diffie-hellman parameters to use for the
5308 TLS server. If the file is missing, QEMU will generate a set of
5309 DH parameters at startup. This is a computationally expensive
5310 operation that consumes random pool entropy, so it is
5311 recommended that a persistent set of parameters be generated
5314 For x509 certificate credentials the directory will contain
5315 further files providing the x509 certificates. The certificates
5316 must be stored in PEM format, in filenames ca-cert.pem,
5317 ca-crl.pem (optional), server-cert.pem (only servers),
5318 server-key.pem (only servers), client-cert.pem (only clients),
5319 and client-key.pem (only clients).
5321 For the server-key.pem and client-key.pem files which contain
5322 sensitive private keys, it is possible to use an encrypted
5323 version by providing the passwordid parameter. This provides the
5324 ID of a previously created ``secret`` object containing the
5325 password for decryption.
5327 The priority parameter allows to override the global default
5328 priority used by gnutls. This can be useful if the system
5329 administrator needs to use a weaker set of crypto priorities for
5330 QEMU without potentially forcing the weakness onto all
5331 applications. Or conversely if one wants wants a stronger
5332 default for QEMU than for all other applications, they can do
5333 this through this parameter. Its format is a gnutls priority
5334 string as described at
5335 https://gnutls.org/manual/html_node/Priority-Strings.html.
5337 ``-object tls-cipher-suites,id=id,priority=priority``
5338 Creates a TLS cipher suites object, which can be used to control
5339 the TLS cipher/protocol algorithms that applications are permitted
5342 The ``id`` parameter is a unique ID which frontends will use to
5343 access the ordered list of permitted TLS cipher suites from the
5346 The ``priority`` parameter allows to override the global default
5347 priority used by gnutls. This can be useful if the system
5348 administrator needs to use a weaker set of crypto priorities for
5349 QEMU without potentially forcing the weakness onto all
5350 applications. Or conversely if one wants wants a stronger
5351 default for QEMU than for all other applications, they can do
5352 this through this parameter. Its format is a gnutls priority
5353 string as described at
5354 https://gnutls.org/manual/html_node/Priority-Strings.html.
5356 An example of use of this object is to control UEFI HTTPS Boot.
5357 The tls-cipher-suites object exposes the ordered list of permitted
5358 TLS cipher suites from the host side to the guest firmware, via
5359 fw_cfg. The list is represented as an array of IANA_TLS_CIPHER
5360 objects. The firmware uses the IANA_TLS_CIPHER array for configuring
5363 In the following example, the priority at which the host-side policy
5364 is retrieved is given by the ``priority`` property.
5365 Given that QEMU uses GNUTLS, ``priority=@SYSTEM`` may be used to
5366 refer to /etc/crypto-policies/back-ends/gnutls.config.
5371 -object tls-cipher-suites,id=mysuite0,priority=@SYSTEM \\
5372 -fw_cfg name=etc/edk2/https/ciphers,gen_id=mysuite0
5374 ``-object filter-buffer,id=id,netdev=netdevid,interval=t[,queue=all|rx|tx][,status=on|off][,position=head|tail|id=<id>][,insert=behind|before]``
5375 Interval t can't be 0, this filter batches the packet delivery:
5376 all packets arriving in a given interval on netdev netdevid are
5377 delayed until the end of the interval. Interval is in
5378 microseconds. ``status`` is optional that indicate whether the
5379 netfilter is on (enabled) or off (disabled), the default status
5380 for netfilter will be 'on'.
5382 queue all\|rx\|tx is an option that can be applied to any
5385 ``all``: the filter is attached both to the receive and the
5386 transmit queue of the netdev (default).
5388 ``rx``: the filter is attached to the receive queue of the
5389 netdev, where it will receive packets sent to the netdev.
5391 ``tx``: the filter is attached to the transmit queue of the
5392 netdev, where it will receive packets sent by the netdev.
5394 position head\|tail\|id=<id> is an option to specify where the
5395 filter should be inserted in the filter list. It can be applied
5398 ``head``: the filter is inserted at the head of the filter list,
5399 before any existing filters.
5401 ``tail``: the filter is inserted at the tail of the filter list,
5402 behind any existing filters (default).
5404 ``id=<id>``: the filter is inserted before or behind the filter
5405 specified by <id>, see the insert option below.
5407 insert behind\|before is an option to specify where to insert
5408 the new filter relative to the one specified with
5409 position=id=<id>. It can be applied to any netfilter.
5411 ``before``: insert before the specified filter.
5413 ``behind``: insert behind the specified filter (default).
5415 ``-object filter-mirror,id=id,netdev=netdevid,outdev=chardevid,queue=all|rx|tx[,vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5416 filter-mirror on netdev netdevid,mirror net packet to
5417 chardevchardevid, if it has the vnet\_hdr\_support flag,
5418 filter-mirror will mirror packet with vnet\_hdr\_len.
5420 ``-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]``
5421 filter-redirector on netdev netdevid,redirect filter's net
5422 packet to chardev chardevid,and redirect indev's packet to
5423 filter.if it has the vnet\_hdr\_support flag, filter-redirector
5424 will redirect packet with vnet\_hdr\_len. Create a
5425 filter-redirector we need to differ outdev id from indev id, id
5426 can not be the same. we can just use indev or outdev, but at
5427 least one of indev or outdev need to be specified.
5429 ``-object filter-rewriter,id=id,netdev=netdevid,queue=all|rx|tx,[vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]``
5430 Filter-rewriter is a part of COLO project.It will rewrite tcp
5431 packet to secondary from primary to keep secondary tcp
5432 connection,and rewrite tcp packet to primary from secondary make
5433 tcp packet can be handled by client.if it has the
5434 vnet\_hdr\_support flag, we can parse packet with vnet header.
5436 usage: colo secondary: -object
5437 filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 -object
5438 filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 -object
5439 filter-rewriter,id=rew0,netdev=hn0,queue=all
5441 ``-object filter-dump,id=id,netdev=dev[,file=filename][,maxlen=len][,position=head|tail|id=<id>][,insert=behind|before]``
5442 Dump the network traffic on netdev dev to the file specified by
5443 filename. At most len bytes (64k by default) per packet are
5444 stored. The file format is libpcap, so it can be analyzed with
5445 tools such as tcpdump or Wireshark.
5447 ``-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}]``
5448 Colo-compare gets packet from primary\_in chardevid and
5449 secondary\_in, then compare whether the payload of primary packet
5450 and secondary packet are the same. If same, it will output
5451 primary packet to out\_dev, else it will notify COLO-framework to do
5452 checkpoint and send primary packet to out\_dev. In order to
5453 improve efficiency, we need to put the task of comparison in
5454 another iothread. If it has the vnet\_hdr\_support flag,
5455 colo compare will send/recv packet with vnet\_hdr\_len.
5456 The compare\_timeout=@var{ms} determines the maximum time of the
5457 colo-compare hold the packet. The expired\_scan\_cycle=@var{ms}
5458 is to set the period of scanning expired primary node network packets.
5459 The max\_queue\_size=@var{size} is to set the max compare queue
5460 size depend on user environment.
5461 If user want to use Xen COLO, need to add the notify\_dev to
5462 notify Xen colo-frame to do checkpoint.
5464 COLO-compare must be used with the help of filter-mirror,
5465 filter-redirector and filter-rewriter.
5472 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5473 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5474 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5475 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5476 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5477 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5478 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5479 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5480 -object iothread,id=iothread1
5481 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5482 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5483 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5484 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,iothread=iothread1
5487 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5488 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5489 -chardev socket,id=red0,host=3.3.3.3,port=9003
5490 -chardev socket,id=red1,host=3.3.3.3,port=9004
5491 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5492 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5498 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown
5499 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66
5500 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off
5501 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off
5502 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off
5503 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001
5504 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off
5505 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005
5506 -chardev socket,id=notify_way,host=3.3.3.3,port=9009,server=on,wait=off
5507 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0
5508 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out
5509 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0
5510 -object iothread,id=iothread1
5511 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,notify_dev=nofity_way,iothread=iothread1
5514 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown
5515 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66
5516 -chardev socket,id=red0,host=3.3.3.3,port=9003
5517 -chardev socket,id=red1,host=3.3.3.3,port=9004
5518 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0
5519 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1
5521 If you want to know the detail of above command line, you can
5522 read the colo-compare git log.
5524 ``-object cryptodev-backend-builtin,id=id[,queues=queues]``
5525 Creates a cryptodev backend which executes crypto operations from
5526 the QEMU cipher APIs. The id parameter is a unique ID that will
5527 be used to reference this cryptodev backend from the
5528 ``virtio-crypto`` device. The queues parameter is optional,
5529 which specify the queue number of cryptodev backend, the default
5536 -object cryptodev-backend-builtin,id=cryptodev0 \\
5537 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5540 ``-object cryptodev-vhost-user,id=id,chardev=chardevid[,queues=queues]``
5541 Creates a vhost-user cryptodev backend, backed by a chardev
5542 chardevid. The id parameter is a unique ID that will be used to
5543 reference this cryptodev backend from the ``virtio-crypto``
5544 device. The chardev should be a unix domain socket backed one.
5545 The vhost-user uses a specifically defined protocol to pass
5546 vhost ioctl replacement messages to an application on the other
5547 end of the socket. The queues parameter is optional, which
5548 specify the queue number of cryptodev backend for multiqueue
5549 vhost-user, the default of queues is 1.
5555 -chardev socket,id=chardev0,path=/path/to/socket \\
5556 -object cryptodev-vhost-user,id=cryptodev0,chardev=chardev0 \\
5557 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\
5560 ``-object secret,id=id,data=string,format=raw|base64[,keyid=secretid,iv=string]``
5562 ``-object secret,id=id,file=filename,format=raw|base64[,keyid=secretid,iv=string]``
5563 Defines a secret to store a password, encryption key, or some
5564 other sensitive data. The sensitive data can either be passed
5565 directly via the data parameter, or indirectly via the file
5566 parameter. Using the data parameter is insecure unless the
5567 sensitive data is encrypted.
5569 The sensitive data can be provided in raw format (the default),
5570 or base64. When encoded as JSON, the raw format only supports
5571 valid UTF-8 characters, so base64 is recommended for sending
5572 binary data. QEMU will convert from which ever format is
5573 provided to the format it needs internally. eg, an RBD password
5574 can be provided in raw format, even though it will be base64
5575 encoded when passed onto the RBD sever.
5577 For added protection, it is possible to encrypt the data
5578 associated with a secret using the AES-256-CBC cipher. Use of
5579 encryption is indicated by providing the keyid and iv
5580 parameters. The keyid parameter provides the ID of a previously
5581 defined secret that contains the AES-256 decryption key. This
5582 key should be 32-bytes long and be base64 encoded. The iv
5583 parameter provides the random initialization vector used for
5584 encryption of this particular secret and should be a base64
5585 encrypted string of the 16-byte IV.
5587 The simplest (insecure) usage is to provide the secret inline
5591 # |qemu_system| -object secret,id=sec0,data=letmein,format=raw
5593 The simplest secure usage is to provide the secret via a file
5595 # printf "letmein" > mypasswd.txt # QEMU\_SYSTEM\_MACRO -object
5596 secret,id=sec0,file=mypasswd.txt,format=raw
5598 For greater security, AES-256-CBC should be used. To illustrate
5599 usage, consider the openssl command line tool which can encrypt
5600 the data. Note that when encrypting, the plaintext must be
5601 padded to the cipher block size (32 bytes) using the standard
5602 PKCS#5/6 compatible padding algorithm.
5604 First a master key needs to be created in base64 encoding:
5608 # openssl rand -base64 32 > key.b64
5609 # KEY=$(base64 -d key.b64 | hexdump -v -e '/1 "%02X"')
5611 Each secret to be encrypted needs to have a random
5612 initialization vector generated. These do not need to be kept
5617 # openssl rand -base64 16 > iv.b64
5618 # IV=$(base64 -d iv.b64 | hexdump -v -e '/1 "%02X"')
5620 The secret to be defined can now be encrypted, in this case
5621 we're telling openssl to base64 encode the result, but it could
5622 be left as raw bytes if desired.
5626 # SECRET=$(printf "letmein" |
5627 openssl enc -aes-256-cbc -a -K $KEY -iv $IV)
5629 When launching QEMU, create a master secret pointing to
5630 ``key.b64`` and specify that to be used to decrypt the user
5631 password. Pass the contents of ``iv.b64`` to the second secret
5636 -object secret,id=secmaster0,format=base64,file=key.b64 \\
5637 -object secret,id=sec0,keyid=secmaster0,format=base64,\\
5638 data=$SECRET,iv=$(<iv.b64)
5640 ``-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]``
5641 Create a Secure Encrypted Virtualization (SEV) guest object,
5642 which can be used to provide the guest memory encryption support
5645 When memory encryption is enabled, one of the physical address
5646 bit (aka the C-bit) is utilized to mark if a memory page is
5647 protected. The ``cbitpos`` is used to provide the C-bit
5648 position. The C-bit position is Host family dependent hence user
5649 must provide this value. On EPYC, the value should be 47.
5651 When memory encryption is enabled, we loose certain bits in
5652 physical address space. The ``reduced-phys-bits`` is used to
5653 provide the number of bits we loose in physical address space.
5654 Similar to C-bit, the value is Host family dependent. On EPYC,
5655 a guest will lose a maximum of 1 bit, so the value should be 1.
5657 The ``sev-device`` provides the device file to use for
5658 communicating with the SEV firmware running inside AMD Secure
5659 Processor. The default device is '/dev/sev'. If hardware
5660 supports memory encryption then /dev/sev devices are created by
5663 The ``policy`` provides the guest policy to be enforced by the
5664 SEV firmware and restrict what configuration and operational
5665 commands can be performed on this guest by the hypervisor. The
5666 policy should be provided by the guest owner and is bound to the
5667 guest and cannot be changed throughout the lifetime of the
5668 guest. The default is 0.
5670 If guest ``policy`` allows sharing the key with another SEV
5671 guest then ``handle`` can be use to provide handle of the guest
5672 from which to share the key.
5674 The ``dh-cert-file`` and ``session-file`` provides the guest
5675 owner's Public Diffie-Hillman key defined in SEV spec. The PDH
5676 and session parameters are used for establishing a cryptographic
5677 session with the guest owner to negotiate keys used for
5678 attestation. The file must be encoded in base64.
5680 The ``kernel-hashes`` adds the hashes of given kernel/initrd/
5681 cmdline to a designated guest firmware page for measured Linux
5682 boot with -kernel. The default is off. (Since 6.2)
5684 e.g to launch a SEV guest
5688 # |qemu_system_x86| \\
5690 -object sev-guest,id=sev0,cbitpos=47,reduced-phys-bits=1 \\
5691 -machine ...,memory-encryption=sev0 \\
5694 ``-object authz-simple,id=id,identity=string``
5695 Create an authorization object that will control access to
5698 The ``identity`` parameter is identifies the user and its format
5699 depends on the network service that authorization object is
5700 associated with. For authorizing based on TLS x509 certificates,
5701 the identity must be the x509 distinguished name. Note that care
5702 must be taken to escape any commas in the distinguished name.
5704 An example authorization object to validate a x509 distinguished
5705 name would look like:
5711 -object 'authz-simple,id=auth0,identity=CN=laptop.example.com,,O=Example Org,,L=London,,ST=London,,C=GB' \\
5714 Note the use of quotes due to the x509 distinguished name
5715 containing whitespace, and escaping of ','.
5717 ``-object authz-listfile,id=id,filename=path,refresh=on|off``
5718 Create an authorization object that will control access to
5721 The ``filename`` parameter is the fully qualified path to a file
5722 containing the access control list rules in JSON format.
5724 An example set of rules that match against SASL usernames might
5731 { "match": "fred", "policy": "allow", "format": "exact" },
5732 { "match": "bob", "policy": "allow", "format": "exact" },
5733 { "match": "danb", "policy": "deny", "format": "glob" },
5734 { "match": "dan*", "policy": "allow", "format": "exact" },
5739 When checking access the object will iterate over all the rules
5740 and the first rule to match will have its ``policy`` value
5741 returned as the result. If no rules match, then the default
5742 ``policy`` value is returned.
5744 The rules can either be an exact string match, or they can use
5745 the simple UNIX glob pattern matching to allow wildcards to be
5748 If ``refresh`` is set to true the file will be monitored and
5749 automatically reloaded whenever its content changes.
5751 As with the ``authz-simple`` object, the format of the identity
5752 strings being matched depends on the network service, but is
5753 usually a TLS x509 distinguished name, or a SASL username.
5755 An example authorization object to validate a SASL username
5762 -object authz-simple,id=auth0,filename=/etc/qemu/vnc-sasl.acl,refresh=on \\
5765 ``-object authz-pam,id=id,service=string``
5766 Create an authorization object that will control access to
5769 The ``service`` parameter provides the name of a PAM service to
5770 use for authorization. It requires that a file
5771 ``/etc/pam.d/service`` exist to provide the configuration for
5772 the ``account`` subsystem.
5774 An example authorization object to validate a TLS x509
5775 distinguished name would look like:
5781 -object authz-pam,id=auth0,service=qemu-vnc \\
5784 There would then be a corresponding config file for PAM at
5785 ``/etc/pam.d/qemu-vnc`` that contains:
5789 account requisite pam_listfile.so item=user sense=allow \
5790 file=/etc/qemu/vnc.allow
5792 Finally the ``/etc/qemu/vnc.allow`` file would contain the list
5793 of x509 distinguished names that are permitted access
5797 CN=laptop.example.com,O=Example Home,L=London,ST=London,C=GB
5799 ``-object iothread,id=id,poll-max-ns=poll-max-ns,poll-grow=poll-grow,poll-shrink=poll-shrink,aio-max-batch=aio-max-batch``
5800 Creates a dedicated event loop thread that devices can be
5801 assigned to. This is known as an IOThread. By default device
5802 emulation happens in vCPU threads or the main event loop thread.
5803 This can become a scalability bottleneck. IOThreads allow device
5804 emulation and I/O to run on other host CPUs.
5806 The ``id`` parameter is a unique ID that will be used to
5807 reference this IOThread from ``-device ...,iothread=id``.
5808 Multiple devices can be assigned to an IOThread. Note that not
5809 all devices support an ``iothread`` parameter.
5811 The ``query-iothreads`` QMP command lists IOThreads and reports
5812 their thread IDs so that the user can configure host CPU
5815 IOThreads use an adaptive polling algorithm to reduce event loop
5816 latency. Instead of entering a blocking system call to monitor
5817 file descriptors and then pay the cost of being woken up when an
5818 event occurs, the polling algorithm spins waiting for events for
5819 a short time. The algorithm's default parameters are suitable
5820 for many cases but can be adjusted based on knowledge of the
5821 workload and/or host device latency.
5823 The ``poll-max-ns`` parameter is the maximum number of
5824 nanoseconds to busy wait for events. Polling can be disabled by
5825 setting this value to 0.
5827 The ``poll-grow`` parameter is the multiplier used to increase
5828 the polling time when the algorithm detects it is missing events
5829 due to not polling long enough.
5831 The ``poll-shrink`` parameter is the divisor used to decrease
5832 the polling time when the algorithm detects it is spending too
5833 long polling without encountering events.
5835 The ``aio-max-batch`` parameter is the maximum number of requests
5836 in a batch for the AIO engine, 0 means that the engine will use
5839 The IOThread parameters can be modified at run-time using the
5840 ``qom-set`` command (where ``iothread1`` is the IOThread's
5845 (qemu) qom-set /objects/iothread1 poll-max-ns 100000
5849 HXCOMM This is the last statement. Insert new options before this line!