1HXCOMM Use DEFHEADING() to define headings in both help text and rST. 2HXCOMM Text between SRST and ERST is copied to the rST version and 3HXCOMM discarded from C version. 4HXCOMM DEF(option, HAS_ARG/0, opt_enum, opt_help, arch_mask) is used to 5HXCOMM construct option structures, enums and help message for specified 6HXCOMM architectures. 7HXCOMM HXCOMM can be used for comments, discarded from both rST and C. 8 9DEFHEADING(Standard options:) 10 11DEF("help", 0, QEMU_OPTION_h, 12 "-h or -help display this help and exit\n", QEMU_ARCH_ALL) 13SRST 14``-h`` 15 Display help and exit 16ERST 17 18DEF("version", 0, QEMU_OPTION_version, 19 "-version display version information and exit\n", QEMU_ARCH_ALL) 20SRST 21``-version`` 22 Display version information and exit 23ERST 24 25DEF("machine", HAS_ARG, QEMU_OPTION_machine, \ 26 "-machine [type=]name[,prop[=value][,...]]\n" 27 " selects emulated machine ('-machine help' for list)\n" 28 " property accel=accel1[:accel2[:...]] selects accelerator\n" 29 " supported accelerators are kvm, xen, hax, hvf, nvmm, whpx or tcg (default: tcg)\n" 30 " vmport=on|off|auto controls emulation of vmport (default: auto)\n" 31 " dump-guest-core=on|off include guest memory in a core dump (default=on)\n" 32 " mem-merge=on|off controls memory merge support (default: on)\n" 33 " aes-key-wrap=on|off controls support for AES key wrapping (default=on)\n" 34 " dea-key-wrap=on|off controls support for DEA key wrapping (default=on)\n" 35 " suppress-vmdesc=on|off disables self-describing migration (default=off)\n" 36 " nvdimm=on|off controls NVDIMM support (default=off)\n" 37 " memory-encryption=@var{} memory encryption object to use (default=none)\n" 38 " hmat=on|off controls ACPI HMAT support (default=off)\n" 39 " memory-backend='backend-id' specifies explicitly provided backend for main RAM (default=none)\n", 40 QEMU_ARCH_ALL) 41SRST 42``-machine [type=]name[,prop=value[,...]]`` 43 Select the emulated machine by name. Use ``-machine help`` to list 44 available machines. 45 46 For architectures which aim to support live migration compatibility 47 across releases, each release will introduce a new versioned machine 48 type. For example, the 2.8.0 release introduced machine types 49 "pc-i440fx-2.8" and "pc-q35-2.8" for the x86\_64/i686 architectures. 50 51 To allow live migration of guests from QEMU version 2.8.0, to QEMU 52 version 2.9.0, the 2.9.0 version must support the "pc-i440fx-2.8" 53 and "pc-q35-2.8" machines too. To allow users live migrating VMs to 54 skip multiple intermediate releases when upgrading, new releases of 55 QEMU will support machine types from many previous versions. 56 57 Supported machine properties are: 58 59 ``accel=accels1[:accels2[:...]]`` 60 This is used to enable an accelerator. Depending on the target 61 architecture, kvm, xen, hax, hvf, nvmm, whpx or tcg can be available. 62 By default, tcg is used. If there is more than one accelerator 63 specified, the next one is used if the previous one fails to 64 initialize. 65 66 ``vmport=on|off|auto`` 67 Enables emulation of VMWare IO port, for vmmouse etc. auto says 68 to select the value based on accel. For accel=xen the default is 69 off otherwise the default is on. 70 71 ``dump-guest-core=on|off`` 72 Include guest memory in a core dump. The default is on. 73 74 ``mem-merge=on|off`` 75 Enables or disables memory merge support. This feature, when 76 supported by the host, de-duplicates identical memory pages 77 among VMs instances (enabled by default). 78 79 ``aes-key-wrap=on|off`` 80 Enables or disables AES key wrapping support on s390-ccw hosts. 81 This feature controls whether AES wrapping keys will be created 82 to allow execution of AES cryptographic functions. The default 83 is on. 84 85 ``dea-key-wrap=on|off`` 86 Enables or disables DEA key wrapping support on s390-ccw hosts. 87 This feature controls whether DEA wrapping keys will be created 88 to allow execution of DEA cryptographic functions. The default 89 is on. 90 91 ``nvdimm=on|off`` 92 Enables or disables NVDIMM support. The default is off. 93 94 ``memory-encryption=`` 95 Memory encryption object to use. The default is none. 96 97 ``hmat=on|off`` 98 Enables or disables ACPI Heterogeneous Memory Attribute Table 99 (HMAT) support. The default is off. 100 101 ``memory-backend='id'`` 102 An alternative to legacy ``-mem-path`` and ``mem-prealloc`` options. 103 Allows to use a memory backend as main RAM. 104 105 For example: 106 :: 107 108 -object memory-backend-file,id=pc.ram,size=512M,mem-path=/hugetlbfs,prealloc=on,share=on 109 -machine memory-backend=pc.ram 110 -m 512M 111 112 Migration compatibility note: 113 114 * as backend id one shall use value of 'default-ram-id', advertised by 115 machine type (available via ``query-machines`` QMP command), if migration 116 to/from old QEMU (<5.0) is expected. 117 * for machine types 4.0 and older, user shall 118 use ``x-use-canonical-path-for-ramblock-id=off`` backend option 119 if migration to/from old QEMU (<5.0) is expected. 120 121 For example: 122 :: 123 124 -object memory-backend-ram,id=pc.ram,size=512M,x-use-canonical-path-for-ramblock-id=off 125 -machine memory-backend=pc.ram 126 -m 512M 127ERST 128 129HXCOMM Deprecated by -machine 130DEF("M", HAS_ARG, QEMU_OPTION_M, "", QEMU_ARCH_ALL) 131 132DEF("cpu", HAS_ARG, QEMU_OPTION_cpu, 133 "-cpu cpu select CPU ('-cpu help' for list)\n", QEMU_ARCH_ALL) 134SRST 135``-cpu model`` 136 Select CPU model (``-cpu help`` for list and additional feature 137 selection) 138ERST 139 140DEF("accel", HAS_ARG, QEMU_OPTION_accel, 141 "-accel [accel=]accelerator[,prop[=value][,...]]\n" 142 " select accelerator (kvm, xen, hax, hvf, nvmm, whpx or tcg; use 'help' for a list)\n" 143 " igd-passthru=on|off (enable Xen integrated Intel graphics passthrough, default=off)\n" 144 " kernel-irqchip=on|off|split controls accelerated irqchip support (default=on)\n" 145 " kvm-shadow-mem=size of KVM shadow MMU in bytes\n" 146 " split-wx=on|off (enable TCG split w^x mapping)\n" 147 " tb-size=n (TCG translation block cache size)\n" 148 " dirty-ring-size=n (KVM dirty ring GFN count, default 0)\n" 149 " thread=single|multi (enable multi-threaded TCG)\n", QEMU_ARCH_ALL) 150SRST 151``-accel name[,prop=value[,...]]`` 152 This is used to enable an accelerator. Depending on the target 153 architecture, kvm, xen, hax, hvf, nvmm, whpx or tcg can be available. By 154 default, tcg is used. If there is more than one accelerator 155 specified, the next one is used if the previous one fails to 156 initialize. 157 158 ``igd-passthru=on|off`` 159 When Xen is in use, this option controls whether Intel 160 integrated graphics devices can be passed through to the guest 161 (default=off) 162 163 ``kernel-irqchip=on|off|split`` 164 Controls KVM in-kernel irqchip support. The default is full 165 acceleration of the interrupt controllers. On x86, split irqchip 166 reduces the kernel attack surface, at a performance cost for 167 non-MSI interrupts. Disabling the in-kernel irqchip completely 168 is not recommended except for debugging purposes. 169 170 ``kvm-shadow-mem=size`` 171 Defines the size of the KVM shadow MMU. 172 173 ``split-wx=on|off`` 174 Controls the use of split w^x mapping for the TCG code generation 175 buffer. Some operating systems require this to be enabled, and in 176 such a case this will default on. On other operating systems, this 177 will default off, but one may enable this for testing or debugging. 178 179 ``tb-size=n`` 180 Controls the size (in MiB) of the TCG translation block cache. 181 182 ``thread=single|multi`` 183 Controls number of TCG threads. When the TCG is multi-threaded 184 there will be one thread per vCPU therefore taking advantage of 185 additional host cores. The default is to enable multi-threading 186 where both the back-end and front-ends support it and no 187 incompatible TCG features have been enabled (e.g. 188 icount/replay). 189 190 ``dirty-ring-size=n`` 191 When the KVM accelerator is used, it controls the size of the per-vCPU 192 dirty page ring buffer (number of entries for each vCPU). It should 193 be a value that is power of two, and it should be 1024 or bigger (but 194 still less than the maximum value that the kernel supports). 4096 195 could be a good initial value if you have no idea which is the best. 196 Set this value to 0 to disable the feature. By default, this feature 197 is disabled (dirty-ring-size=0). When enabled, KVM will instead 198 record dirty pages in a bitmap. 199 200ERST 201 202DEF("smp", HAS_ARG, QEMU_OPTION_smp, 203 "-smp [[cpus=]n][,maxcpus=cpus][,sockets=sockets][,dies=dies][,cores=cores][,threads=threads]\n" 204 " set the number of CPUs to 'n' [default=1]\n" 205 " maxcpus= maximum number of total CPUs, including\n" 206 " offline CPUs for hotplug, etc\n" 207 " sockets= number of discrete sockets in the system\n" 208 " dies= number of CPU dies on one socket (for PC only)\n" 209 " cores= number of CPU cores on one socket (for PC, it's on one die)\n" 210 " threads= number of threads on one CPU core\n", 211 QEMU_ARCH_ALL) 212SRST 213``-smp [[cpus=]n][,maxcpus=maxcpus][,sockets=sockets][,dies=dies][,cores=cores][,threads=threads]`` 214 Simulate a SMP system with '\ ``n``\ ' CPUs initially present on 215 the machine type board. On boards supporting CPU hotplug, the optional 216 '\ ``maxcpus``\ ' parameter can be set to enable further CPUs to be 217 added at runtime. If omitted the maximum number of CPUs will be 218 set to match the initial CPU count. Both parameters are subject to 219 an upper limit that is determined by the specific machine type chosen. 220 221 To control reporting of CPU topology information, the number of sockets, 222 dies per socket, cores per die, and threads per core can be specified. 223 The sum `` sockets * cores * dies * threads `` must be equal to the 224 maximum CPU count. CPU targets may only support a subset of the topology 225 parameters. Where a CPU target does not support use of a particular 226 topology parameter, its value should be assumed to be 1 for the purpose 227 of computing the CPU maximum count. 228 229 Either the initial CPU count, or at least one of the topology parameters 230 must be specified. Values for any omitted parameters will be computed 231 from those which are given. Historically preference was given to the 232 coarsest topology parameters when computing missing values (ie sockets 233 preferred over cores, which were preferred over threads), however, this 234 behaviour is considered liable to change. 235ERST 236 237DEF("numa", HAS_ARG, QEMU_OPTION_numa, 238 "-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n" 239 "-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=node]\n" 240 "-numa dist,src=source,dst=destination,val=distance\n" 241 "-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]\n" 242 "-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" 243 "-numa hmat-cache,node-id=node,size=size,level=level[,associativity=none|direct|complex][,policy=none|write-back|write-through][,line=size]\n", 244 QEMU_ARCH_ALL) 245SRST 246``-numa node[,mem=size][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]`` 247 \ 248``-numa node[,memdev=id][,cpus=firstcpu[-lastcpu]][,nodeid=node][,initiator=initiator]`` 249 \ 250``-numa dist,src=source,dst=destination,val=distance`` 251 \ 252``-numa cpu,node-id=node[,socket-id=x][,core-id=y][,thread-id=z]`` 253 \ 254``-numa hmat-lb,initiator=node,target=node,hierarchy=hierarchy,data-type=tpye[,latency=lat][,bandwidth=bw]`` 255 \ 256``-numa hmat-cache,node-id=node,size=size,level=level[,associativity=str][,policy=str][,line=size]`` 257 Define a NUMA node and assign RAM and VCPUs to it. Set the NUMA 258 distance from a source node to a destination node. Set the ACPI 259 Heterogeneous Memory Attributes for the given nodes. 260 261 Legacy VCPU assignment uses '\ ``cpus``\ ' option where firstcpu and 262 lastcpu are CPU indexes. Each '\ ``cpus``\ ' option represent a 263 contiguous range of CPU indexes (or a single VCPU if lastcpu is 264 omitted). A non-contiguous set of VCPUs can be represented by 265 providing multiple '\ ``cpus``\ ' options. If '\ ``cpus``\ ' is 266 omitted on all nodes, VCPUs are automatically split between them. 267 268 For example, the following option assigns VCPUs 0, 1, 2 and 5 to a 269 NUMA node: 270 271 :: 272 273 -numa node,cpus=0-2,cpus=5 274 275 '\ ``cpu``\ ' option is a new alternative to '\ ``cpus``\ ' option 276 which uses '\ ``socket-id|core-id|thread-id``\ ' properties to 277 assign CPU objects to a node using topology layout properties of 278 CPU. The set of properties is machine specific, and depends on used 279 machine type/'\ ``smp``\ ' options. It could be queried with 280 '\ ``hotpluggable-cpus``\ ' monitor command. '\ ``node-id``\ ' 281 property specifies node to which CPU object will be assigned, it's 282 required for node to be declared with '\ ``node``\ ' option before 283 it's used with '\ ``cpu``\ ' option. 284 285 For example: 286 287 :: 288 289 -M pc \ 290 -smp 1,sockets=2,maxcpus=2 \ 291 -numa node,nodeid=0 -numa node,nodeid=1 \ 292 -numa cpu,node-id=0,socket-id=0 -numa cpu,node-id=1,socket-id=1 293 294 Legacy '\ ``mem``\ ' assigns a given RAM amount to a node (not supported 295 for 5.1 and newer machine types). '\ ``memdev``\ ' assigns RAM from 296 a given memory backend device to a node. If '\ ``mem``\ ' and 297 '\ ``memdev``\ ' are omitted in all nodes, RAM is split equally between them. 298 299 300 '\ ``mem``\ ' and '\ ``memdev``\ ' are mutually exclusive. 301 Furthermore, if one node uses '\ ``memdev``\ ', all of them have to 302 use it. 303 304 '\ ``initiator``\ ' is an additional option that points to an 305 initiator NUMA node that has best performance (the lowest latency or 306 largest bandwidth) to this NUMA node. Note that this option can be 307 set only when the machine property 'hmat' is set to 'on'. 308 309 Following example creates a machine with 2 NUMA nodes, node 0 has 310 CPU. node 1 has only memory, and its initiator is node 0. Note that 311 because node 0 has CPU, by default the initiator of node 0 is itself 312 and must be itself. 313 314 :: 315 316 -machine hmat=on \ 317 -m 2G,slots=2,maxmem=4G \ 318 -object memory-backend-ram,size=1G,id=m0 \ 319 -object memory-backend-ram,size=1G,id=m1 \ 320 -numa node,nodeid=0,memdev=m0 \ 321 -numa node,nodeid=1,memdev=m1,initiator=0 \ 322 -smp 2,sockets=2,maxcpus=2 \ 323 -numa cpu,node-id=0,socket-id=0 \ 324 -numa cpu,node-id=0,socket-id=1 325 326 source and destination are NUMA node IDs. distance is the NUMA 327 distance from source to destination. The distance from a node to 328 itself is always 10. If any pair of nodes is given a distance, then 329 all pairs must be given distances. Although, when distances are only 330 given in one direction for each pair of nodes, then the distances in 331 the opposite directions are assumed to be the same. If, however, an 332 asymmetrical pair of distances is given for even one node pair, then 333 all node pairs must be provided distance values for both directions, 334 even when they are symmetrical. When a node is unreachable from 335 another node, set the pair's distance to 255. 336 337 Note that the -``numa`` option doesn't allocate any of the specified 338 resources, it just assigns existing resources to NUMA nodes. This 339 means that one still has to use the ``-m``, ``-smp`` options to 340 allocate RAM and VCPUs respectively. 341 342 Use '\ ``hmat-lb``\ ' to set System Locality Latency and Bandwidth 343 Information between initiator and target NUMA nodes in ACPI 344 Heterogeneous Attribute Memory Table (HMAT). Initiator NUMA node can 345 create memory requests, usually it has one or more processors. 346 Target NUMA node contains addressable memory. 347 348 In '\ ``hmat-lb``\ ' option, node are NUMA node IDs. hierarchy is 349 the memory hierarchy of the target NUMA node: if hierarchy is 350 'memory', the structure represents the memory performance; if 351 hierarchy is 'first-level\|second-level\|third-level', this 352 structure represents aggregated performance of memory side caches 353 for each domain. type of 'data-type' is type of data represented by 354 this structure instance: if 'hierarchy' is 'memory', 'data-type' is 355 'access\|read\|write' latency or 'access\|read\|write' bandwidth of 356 the target memory; if 'hierarchy' is 357 'first-level\|second-level\|third-level', 'data-type' is 358 'access\|read\|write' hit latency or 'access\|read\|write' hit 359 bandwidth of the target memory side cache. 360 361 lat is latency value in nanoseconds. bw is bandwidth value, the 362 possible value and units are NUM[M\|G\|T], mean that the bandwidth 363 value are NUM byte per second (or MB/s, GB/s or TB/s depending on 364 used suffix). Note that if latency or bandwidth value is 0, means 365 the corresponding latency or bandwidth information is not provided. 366 367 In '\ ``hmat-cache``\ ' option, node-id is the NUMA-id of the memory 368 belongs. size is the size of memory side cache in bytes. level is 369 the cache level described in this structure, note that the cache 370 level 0 should not be used with '\ ``hmat-cache``\ ' option. 371 associativity is the cache associativity, the possible value is 372 'none/direct(direct-mapped)/complex(complex cache indexing)'. policy 373 is the write policy. line is the cache Line size in bytes. 374 375 For example, the following options describe 2 NUMA nodes. Node 0 has 376 2 cpus and a ram, node 1 has only a ram. The processors in node 0 377 access memory in node 0 with access-latency 5 nanoseconds, 378 access-bandwidth is 200 MB/s; The processors in NUMA node 0 access 379 memory in NUMA node 1 with access-latency 10 nanoseconds, 380 access-bandwidth is 100 MB/s. And for memory side cache information, 381 NUMA node 0 and 1 both have 1 level memory cache, size is 10KB, 382 policy is write-back, the cache Line size is 8 bytes: 383 384 :: 385 386 -machine hmat=on \ 387 -m 2G \ 388 -object memory-backend-ram,size=1G,id=m0 \ 389 -object memory-backend-ram,size=1G,id=m1 \ 390 -smp 2 \ 391 -numa node,nodeid=0,memdev=m0 \ 392 -numa node,nodeid=1,memdev=m1,initiator=0 \ 393 -numa cpu,node-id=0,socket-id=0 \ 394 -numa cpu,node-id=0,socket-id=1 \ 395 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-latency,latency=5 \ 396 -numa hmat-lb,initiator=0,target=0,hierarchy=memory,data-type=access-bandwidth,bandwidth=200M \ 397 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-latency,latency=10 \ 398 -numa hmat-lb,initiator=0,target=1,hierarchy=memory,data-type=access-bandwidth,bandwidth=100M \ 399 -numa hmat-cache,node-id=0,size=10K,level=1,associativity=direct,policy=write-back,line=8 \ 400 -numa hmat-cache,node-id=1,size=10K,level=1,associativity=direct,policy=write-back,line=8 401ERST 402 403DEF("add-fd", HAS_ARG, QEMU_OPTION_add_fd, 404 "-add-fd fd=fd,set=set[,opaque=opaque]\n" 405 " Add 'fd' to fd 'set'\n", QEMU_ARCH_ALL) 406SRST 407``-add-fd fd=fd,set=set[,opaque=opaque]`` 408 Add a file descriptor to an fd set. Valid options are: 409 410 ``fd=fd`` 411 This option defines the file descriptor of which a duplicate is 412 added to fd set. The file descriptor cannot be stdin, stdout, or 413 stderr. 414 415 ``set=set`` 416 This option defines the ID of the fd set to add the file 417 descriptor to. 418 419 ``opaque=opaque`` 420 This option defines a free-form string that can be used to 421 describe fd. 422 423 You can open an image using pre-opened file descriptors from an fd 424 set: 425 426 .. parsed-literal:: 427 428 |qemu_system| \\ 429 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\ 430 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\ 431 -drive file=/dev/fdset/2,index=0,media=disk 432ERST 433 434DEF("set", HAS_ARG, QEMU_OPTION_set, 435 "-set group.id.arg=value\n" 436 " set <arg> parameter for item <id> of type <group>\n" 437 " i.e. -set drive.$id.file=/path/to/image\n", QEMU_ARCH_ALL) 438SRST 439``-set group.id.arg=value`` 440 Set parameter arg for item id of type group 441ERST 442 443DEF("global", HAS_ARG, QEMU_OPTION_global, 444 "-global driver.property=value\n" 445 "-global driver=driver,property=property,value=value\n" 446 " set a global default for a driver property\n", 447 QEMU_ARCH_ALL) 448SRST 449``-global driver.prop=value`` 450 \ 451``-global driver=driver,property=property,value=value`` 452 Set default value of driver's property prop to value, e.g.: 453 454 .. parsed-literal:: 455 456 |qemu_system_x86| -global ide-hd.physical_block_size=4096 disk-image.img 457 458 In particular, you can use this to set driver properties for devices 459 which are created automatically by the machine model. To create a 460 device which is not created automatically and set properties on it, 461 use -``device``. 462 463 -global driver.prop=value is shorthand for -global 464 driver=driver,property=prop,value=value. The longhand syntax works 465 even when driver contains a dot. 466ERST 467 468DEF("boot", HAS_ARG, QEMU_OPTION_boot, 469 "-boot [order=drives][,once=drives][,menu=on|off]\n" 470 " [,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_time][,strict=on|off]\n" 471 " 'drives': floppy (a), hard disk (c), CD-ROM (d), network (n)\n" 472 " 'sp_name': the file's name that would be passed to bios as logo picture, if menu=on\n" 473 " 'sp_time': the period that splash picture last if menu=on, unit is ms\n" 474 " 'rb_timeout': the timeout before guest reboot when boot failed, unit is ms\n", 475 QEMU_ARCH_ALL) 476SRST 477``-boot [order=drives][,once=drives][,menu=on|off][,splash=sp_name][,splash-time=sp_time][,reboot-timeout=rb_timeout][,strict=on|off]`` 478 Specify boot order drives as a string of drive letters. Valid drive 479 letters depend on the target architecture. The x86 PC uses: a, b 480 (floppy 1 and 2), c (first hard disk), d (first CD-ROM), n-p 481 (Etherboot from network adapter 1-4), hard disk boot is the default. 482 To apply a particular boot order only on the first startup, specify 483 it via ``once``. Note that the ``order`` or ``once`` parameter 484 should not be used together with the ``bootindex`` property of 485 devices, since the firmware implementations normally do not support 486 both at the same time. 487 488 Interactive boot menus/prompts can be enabled via ``menu=on`` as far 489 as firmware/BIOS supports them. The default is non-interactive boot. 490 491 A splash picture could be passed to bios, enabling user to show it 492 as logo, when option splash=sp\_name is given and menu=on, If 493 firmware/BIOS supports them. Currently Seabios for X86 system 494 support it. limitation: The splash file could be a jpeg file or a 495 BMP file in 24 BPP format(true color). The resolution should be 496 supported by the SVGA mode, so the recommended is 320x240, 640x480, 497 800x640. 498 499 A timeout could be passed to bios, guest will pause for rb\_timeout 500 ms when boot failed, then reboot. If rb\_timeout is '-1', guest will 501 not reboot, qemu passes '-1' to bios by default. Currently Seabios 502 for X86 system support it. 503 504 Do strict boot via ``strict=on`` as far as firmware/BIOS supports 505 it. This only effects when boot priority is changed by bootindex 506 options. The default is non-strict boot. 507 508 .. parsed-literal:: 509 510 # try to boot from network first, then from hard disk 511 |qemu_system_x86| -boot order=nc 512 # boot from CD-ROM first, switch back to default order after reboot 513 |qemu_system_x86| -boot once=d 514 # boot with a splash picture for 5 seconds. 515 |qemu_system_x86| -boot menu=on,splash=/root/boot.bmp,splash-time=5000 516 517 Note: The legacy format '-boot drives' is still supported but its 518 use is discouraged as it may be removed from future versions. 519ERST 520 521DEF("m", HAS_ARG, QEMU_OPTION_m, 522 "-m [size=]megs[,slots=n,maxmem=size]\n" 523 " configure guest RAM\n" 524 " size: initial amount of guest memory\n" 525 " slots: number of hotplug slots (default: none)\n" 526 " maxmem: maximum amount of guest memory (default: none)\n" 527 "NOTE: Some architectures might enforce a specific granularity\n", 528 QEMU_ARCH_ALL) 529SRST 530``-m [size=]megs[,slots=n,maxmem=size]`` 531 Sets guest startup RAM size to megs megabytes. Default is 128 MiB. 532 Optionally, a suffix of "M" or "G" can be used to signify a value in 533 megabytes or gigabytes respectively. Optional pair slots, maxmem 534 could be used to set amount of hotpluggable memory slots and maximum 535 amount of memory. Note that maxmem must be aligned to the page size. 536 537 For example, the following command-line sets the guest startup RAM 538 size to 1GB, creates 3 slots to hotplug additional memory and sets 539 the maximum memory the guest can reach to 4GB: 540 541 .. parsed-literal:: 542 543 |qemu_system| -m 1G,slots=3,maxmem=4G 544 545 If slots and maxmem are not specified, memory hotplug won't be 546 enabled and the guest startup RAM will never increase. 547ERST 548 549DEF("mem-path", HAS_ARG, QEMU_OPTION_mempath, 550 "-mem-path FILE provide backing storage for guest RAM\n", QEMU_ARCH_ALL) 551SRST 552``-mem-path path`` 553 Allocate guest RAM from a temporarily created file in path. 554ERST 555 556DEF("mem-prealloc", 0, QEMU_OPTION_mem_prealloc, 557 "-mem-prealloc preallocate guest memory (use with -mem-path)\n", 558 QEMU_ARCH_ALL) 559SRST 560``-mem-prealloc`` 561 Preallocate memory when using -mem-path. 562ERST 563 564DEF("k", HAS_ARG, QEMU_OPTION_k, 565 "-k language use keyboard layout (for example 'fr' for French)\n", 566 QEMU_ARCH_ALL) 567SRST 568``-k language`` 569 Use keyboard layout language (for example ``fr`` for French). This 570 option is only needed where it is not easy to get raw PC keycodes 571 (e.g. on Macs, with some X11 servers or with a VNC or curses 572 display). You don't normally need to use it on PC/Linux or 573 PC/Windows hosts. 574 575 The available layouts are: 576 577 :: 578 579 ar de-ch es fo fr-ca hu ja mk no pt-br sv 580 da en-gb et fr fr-ch is lt nl pl ru th 581 de en-us fi fr-be hr it lv nl-be pt sl tr 582 583 The default is ``en-us``. 584ERST 585 586 587HXCOMM Deprecated by -audiodev 588DEF("audio-help", 0, QEMU_OPTION_audio_help, 589 "-audio-help show -audiodev equivalent of the currently specified audio settings\n", 590 QEMU_ARCH_ALL) 591SRST 592``-audio-help`` 593 Will show the -audiodev equivalent of the currently specified 594 (deprecated) environment variables. 595ERST 596 597DEF("audiodev", HAS_ARG, QEMU_OPTION_audiodev, 598 "-audiodev [driver=]driver,id=id[,prop[=value][,...]]\n" 599 " specifies the audio backend to use\n" 600 " id= identifier of the backend\n" 601 " timer-period= timer period in microseconds\n" 602 " in|out.mixing-engine= use mixing engine to mix streams inside QEMU\n" 603 " in|out.fixed-settings= use fixed settings for host audio\n" 604 " in|out.frequency= frequency to use with fixed settings\n" 605 " in|out.channels= number of channels to use with fixed settings\n" 606 " in|out.format= sample format to use with fixed settings\n" 607 " valid values: s8, s16, s32, u8, u16, u32, f32\n" 608 " in|out.voices= number of voices to use\n" 609 " in|out.buffer-length= length of buffer in microseconds\n" 610 "-audiodev none,id=id,[,prop[=value][,...]]\n" 611 " dummy driver that discards all output\n" 612#ifdef CONFIG_AUDIO_ALSA 613 "-audiodev alsa,id=id[,prop[=value][,...]]\n" 614 " in|out.dev= name of the audio device to use\n" 615 " in|out.period-length= length of period in microseconds\n" 616 " in|out.try-poll= attempt to use poll mode\n" 617 " threshold= threshold (in microseconds) when playback starts\n" 618#endif 619#ifdef CONFIG_AUDIO_COREAUDIO 620 "-audiodev coreaudio,id=id[,prop[=value][,...]]\n" 621 " in|out.buffer-count= number of buffers\n" 622#endif 623#ifdef CONFIG_AUDIO_DSOUND 624 "-audiodev dsound,id=id[,prop[=value][,...]]\n" 625 " latency= add extra latency to playback in microseconds\n" 626#endif 627#ifdef CONFIG_AUDIO_OSS 628 "-audiodev oss,id=id[,prop[=value][,...]]\n" 629 " in|out.dev= path of the audio device to use\n" 630 " in|out.buffer-count= number of buffers\n" 631 " in|out.try-poll= attempt to use poll mode\n" 632 " try-mmap= try using memory mapped access\n" 633 " exclusive= open device in exclusive mode\n" 634 " dsp-policy= set timing policy (0..10), -1 to use fragment mode\n" 635#endif 636#ifdef CONFIG_AUDIO_PA 637 "-audiodev pa,id=id[,prop[=value][,...]]\n" 638 " server= PulseAudio server address\n" 639 " in|out.name= source/sink device name\n" 640 " in|out.latency= desired latency in microseconds\n" 641#endif 642#ifdef CONFIG_AUDIO_SDL 643 "-audiodev sdl,id=id[,prop[=value][,...]]\n" 644 " in|out.buffer-count= number of buffers\n" 645#endif 646#ifdef CONFIG_SPICE 647 "-audiodev spice,id=id[,prop[=value][,...]]\n" 648#endif 649 "-audiodev wav,id=id[,prop[=value][,...]]\n" 650 " path= path of wav file to record\n", 651 QEMU_ARCH_ALL) 652SRST 653``-audiodev [driver=]driver,id=id[,prop[=value][,...]]`` 654 Adds a new audio backend driver identified by id. There are global 655 and driver specific properties. Some values can be set differently 656 for input and output, they're marked with ``in|out.``. You can set 657 the input's property with ``in.prop`` and the output's property with 658 ``out.prop``. For example: 659 660 :: 661 662 -audiodev alsa,id=example,in.frequency=44110,out.frequency=8000 663 -audiodev alsa,id=example,out.channels=1 # leaves in.channels unspecified 664 665 NOTE: parameter validation is known to be incomplete, in many cases 666 specifying an invalid option causes QEMU to print an error message 667 and continue emulation without sound. 668 669 Valid global options are: 670 671 ``id=identifier`` 672 Identifies the audio backend. 673 674 ``timer-period=period`` 675 Sets the timer period used by the audio subsystem in 676 microseconds. Default is 10000 (10 ms). 677 678 ``in|out.mixing-engine=on|off`` 679 Use QEMU's mixing engine to mix all streams inside QEMU and 680 convert audio formats when not supported by the backend. When 681 off, fixed-settings must be off too. Note that disabling this 682 option means that the selected backend must support multiple 683 streams and the audio formats used by the virtual cards, 684 otherwise you'll get no sound. It's not recommended to disable 685 this option unless you want to use 5.1 or 7.1 audio, as mixing 686 engine only supports mono and stereo audio. Default is on. 687 688 ``in|out.fixed-settings=on|off`` 689 Use fixed settings for host audio. When off, it will change 690 based on how the guest opens the sound card. In this case you 691 must not specify frequency, channels or format. Default is on. 692 693 ``in|out.frequency=frequency`` 694 Specify the frequency to use when using fixed-settings. Default 695 is 44100Hz. 696 697 ``in|out.channels=channels`` 698 Specify the number of channels to use when using fixed-settings. 699 Default is 2 (stereo). 700 701 ``in|out.format=format`` 702 Specify the sample format to use when using fixed-settings. 703 Valid values are: ``s8``, ``s16``, ``s32``, ``u8``, ``u16``, 704 ``u32``, ``f32``. Default is ``s16``. 705 706 ``in|out.voices=voices`` 707 Specify the number of voices to use. Default is 1. 708 709 ``in|out.buffer-length=usecs`` 710 Sets the size of the buffer in microseconds. 711 712``-audiodev none,id=id[,prop[=value][,...]]`` 713 Creates a dummy backend that discards all outputs. This backend has 714 no backend specific properties. 715 716``-audiodev alsa,id=id[,prop[=value][,...]]`` 717 Creates backend using the ALSA. This backend is only available on 718 Linux. 719 720 ALSA specific options are: 721 722 ``in|out.dev=device`` 723 Specify the ALSA device to use for input and/or output. Default 724 is ``default``. 725 726 ``in|out.period-length=usecs`` 727 Sets the period length in microseconds. 728 729 ``in|out.try-poll=on|off`` 730 Attempt to use poll mode with the device. Default is on. 731 732 ``threshold=threshold`` 733 Threshold (in microseconds) when playback starts. Default is 0. 734 735``-audiodev coreaudio,id=id[,prop[=value][,...]]`` 736 Creates a backend using Apple's Core Audio. This backend is only 737 available on Mac OS and only supports playback. 738 739 Core Audio specific options are: 740 741 ``in|out.buffer-count=count`` 742 Sets the count of the buffers. 743 744``-audiodev dsound,id=id[,prop[=value][,...]]`` 745 Creates a backend using Microsoft's DirectSound. This backend is 746 only available on Windows and only supports playback. 747 748 DirectSound specific options are: 749 750 ``latency=usecs`` 751 Add extra usecs microseconds latency to playback. Default is 752 10000 (10 ms). 753 754``-audiodev oss,id=id[,prop[=value][,...]]`` 755 Creates a backend using OSS. This backend is available on most 756 Unix-like systems. 757 758 OSS specific options are: 759 760 ``in|out.dev=device`` 761 Specify the file name of the OSS device to use. Default is 762 ``/dev/dsp``. 763 764 ``in|out.buffer-count=count`` 765 Sets the count of the buffers. 766 767 ``in|out.try-poll=on|of`` 768 Attempt to use poll mode with the device. Default is on. 769 770 ``try-mmap=on|off`` 771 Try using memory mapped device access. Default is off. 772 773 ``exclusive=on|off`` 774 Open the device in exclusive mode (vmix won't work in this 775 case). Default is off. 776 777 ``dsp-policy=policy`` 778 Sets the timing policy (between 0 and 10, where smaller number 779 means smaller latency but higher CPU usage). Use -1 to use 780 buffer sizes specified by ``buffer`` and ``buffer-count``. This 781 option is ignored if you do not have OSS 4. Default is 5. 782 783``-audiodev pa,id=id[,prop[=value][,...]]`` 784 Creates a backend using PulseAudio. This backend is available on 785 most systems. 786 787 PulseAudio specific options are: 788 789 ``server=server`` 790 Sets the PulseAudio server to connect to. 791 792 ``in|out.name=sink`` 793 Use the specified source/sink for recording/playback. 794 795 ``in|out.latency=usecs`` 796 Desired latency in microseconds. The PulseAudio server will try 797 to honor this value but actual latencies may be lower or higher. 798 799``-audiodev sdl,id=id[,prop[=value][,...]]`` 800 Creates a backend using SDL. This backend is available on most 801 systems, but you should use your platform's native backend if 802 possible. 803 804 SDL specific options are: 805 806 ``in|out.buffer-count=count`` 807 Sets the count of the buffers. 808 809``-audiodev spice,id=id[,prop[=value][,...]]`` 810 Creates a backend that sends audio through SPICE. This backend 811 requires ``-spice`` and automatically selected in that case, so 812 usually you can ignore this option. This backend has no backend 813 specific properties. 814 815``-audiodev wav,id=id[,prop[=value][,...]]`` 816 Creates a backend that writes audio to a WAV file. 817 818 Backend specific options are: 819 820 ``path=path`` 821 Write recorded audio into the specified file. Default is 822 ``qemu.wav``. 823ERST 824 825DEF("soundhw", HAS_ARG, QEMU_OPTION_soundhw, 826 "-soundhw c1,... enable audio support\n" 827 " and only specified sound cards (comma separated list)\n" 828 " use '-soundhw help' to get the list of supported cards\n" 829 " use '-soundhw all' to enable all of them\n", QEMU_ARCH_ALL) 830SRST 831``-soundhw card1[,card2,...] or -soundhw all`` 832 Enable audio and selected sound hardware. Use 'help' to print all 833 available sound hardware. For example: 834 835 .. parsed-literal:: 836 837 |qemu_system_x86| -soundhw sb16,adlib disk.img 838 |qemu_system_x86| -soundhw es1370 disk.img 839 |qemu_system_x86| -soundhw ac97 disk.img 840 |qemu_system_x86| -soundhw hda disk.img 841 |qemu_system_x86| -soundhw all disk.img 842 |qemu_system_x86| -soundhw help 843 844 Note that Linux's i810\_audio OSS kernel (for AC97) module might 845 require manually specifying clocking. 846 847 :: 848 849 modprobe i810_audio clocking=48000 850ERST 851 852DEF("device", HAS_ARG, QEMU_OPTION_device, 853 "-device driver[,prop[=value][,...]]\n" 854 " add device (based on driver)\n" 855 " prop=value,... sets driver properties\n" 856 " use '-device help' to print all possible drivers\n" 857 " use '-device driver,help' to print all possible properties\n", 858 QEMU_ARCH_ALL) 859SRST 860``-device driver[,prop[=value][,...]]`` 861 Add device driver. prop=value sets driver properties. Valid 862 properties depend on the driver. To get help on possible drivers and 863 properties, use ``-device help`` and ``-device driver,help``. 864 865 Some drivers are: 866 867``-device ipmi-bmc-sim,id=id[,prop[=value][,...]]`` 868 Add an IPMI BMC. This is a simulation of a hardware management 869 interface processor that normally sits on a system. It provides a 870 watchdog and the ability to reset and power control the system. You 871 need to connect this to an IPMI interface to make it useful 872 873 The IPMI slave address to use for the BMC. The default is 0x20. This 874 address is the BMC's address on the I2C network of management 875 controllers. If you don't know what this means, it is safe to ignore 876 it. 877 878 ``id=id`` 879 The BMC id for interfaces to use this device. 880 881 ``slave_addr=val`` 882 Define slave address to use for the BMC. The default is 0x20. 883 884 ``sdrfile=file`` 885 file containing raw Sensor Data Records (SDR) data. The default 886 is none. 887 888 ``fruareasize=val`` 889 size of a Field Replaceable Unit (FRU) area. The default is 890 1024. 891 892 ``frudatafile=file`` 893 file containing raw Field Replaceable Unit (FRU) inventory data. 894 The default is none. 895 896 ``guid=uuid`` 897 value for the GUID for the BMC, in standard UUID format. If this 898 is set, get "Get GUID" command to the BMC will return it. 899 Otherwise "Get GUID" will return an error. 900 901``-device ipmi-bmc-extern,id=id,chardev=id[,slave_addr=val]`` 902 Add a connection to an external IPMI BMC simulator. Instead of 903 locally emulating the BMC like the above item, instead connect to an 904 external entity that provides the IPMI services. 905 906 A connection is made to an external BMC simulator. If you do this, 907 it is strongly recommended that you use the "reconnect=" chardev 908 option to reconnect to the simulator if the connection is lost. Note 909 that if this is not used carefully, it can be a security issue, as 910 the interface has the ability to send resets, NMIs, and power off 911 the VM. It's best if QEMU makes a connection to an external 912 simulator running on a secure port on localhost, so neither the 913 simulator nor QEMU is exposed to any outside network. 914 915 See the "lanserv/README.vm" file in the OpenIPMI library for more 916 details on the external interface. 917 918``-device isa-ipmi-kcs,bmc=id[,ioport=val][,irq=val]`` 919 Add a KCS IPMI interafce on the ISA bus. This also adds a 920 corresponding ACPI and SMBIOS entries, if appropriate. 921 922 ``bmc=id`` 923 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern 924 above. 925 926 ``ioport=val`` 927 Define the I/O address of the interface. The default is 0xca0 928 for KCS. 929 930 ``irq=val`` 931 Define the interrupt to use. The default is 5. To disable 932 interrupts, set this to 0. 933 934``-device isa-ipmi-bt,bmc=id[,ioport=val][,irq=val]`` 935 Like the KCS interface, but defines a BT interface. The default port 936 is 0xe4 and the default interrupt is 5. 937 938``-device pci-ipmi-kcs,bmc=id`` 939 Add a KCS IPMI interafce on the PCI bus. 940 941 ``bmc=id`` 942 The BMC to connect to, one of ipmi-bmc-sim or ipmi-bmc-extern above. 943 944``-device pci-ipmi-bt,bmc=id`` 945 Like the KCS interface, but defines a BT interface on the PCI bus. 946 947``-device intel-iommu[,option=...]`` 948 This is only supported by ``-machine q35``, which will enable Intel VT-d 949 emulation within the guest. It supports below options: 950 951 ``intremap=on|off`` (default: auto) 952 This enables interrupt remapping feature. It's required to enable 953 complete x2apic. Currently it only supports kvm kernel-irqchip modes 954 ``off`` or ``split``, while full kernel-irqchip is not yet supported. 955 The default value is "auto", which will be decided by the mode of 956 kernel-irqchip. 957 958 ``caching-mode=on|off`` (default: off) 959 This enables caching mode for the VT-d emulated device. When 960 caching-mode is enabled, each guest DMA buffer mapping will generate an 961 IOTLB invalidation from the guest IOMMU driver to the vIOMMU device in 962 a synchronous way. It is required for ``-device vfio-pci`` to work 963 with the VT-d device, because host assigned devices requires to setup 964 the DMA mapping on the host before guest DMA starts. 965 966 ``device-iotlb=on|off`` (default: off) 967 This enables device-iotlb capability for the emulated VT-d device. So 968 far virtio/vhost should be the only real user for this parameter, 969 paired with ats=on configured for the device. 970 971 ``aw-bits=39|48`` (default: 39) 972 This decides the address width of IOVA address space. The address 973 space has 39 bits width for 3-level IOMMU page tables, and 48 bits for 974 4-level IOMMU page tables. 975 976 Please also refer to the wiki page for general scenarios of VT-d 977 emulation in QEMU: https://wiki.qemu.org/Features/VT-d. 978 979ERST 980 981DEF("name", HAS_ARG, QEMU_OPTION_name, 982 "-name string1[,process=string2][,debug-threads=on|off]\n" 983 " set the name of the guest\n" 984 " string1 sets the window title and string2 the process name\n" 985 " When debug-threads is enabled, individual threads are given a separate name\n" 986 " NOTE: The thread names are for debugging and not a stable API.\n", 987 QEMU_ARCH_ALL) 988SRST 989``-name name`` 990 Sets the name of the guest. This name will be displayed in the SDL 991 window caption. The name will also be used for the VNC server. Also 992 optionally set the top visible process name in Linux. Naming of 993 individual threads can also be enabled on Linux to aid debugging. 994ERST 995 996DEF("uuid", HAS_ARG, QEMU_OPTION_uuid, 997 "-uuid %08x-%04x-%04x-%04x-%012x\n" 998 " specify machine UUID\n", QEMU_ARCH_ALL) 999SRST 1000``-uuid uuid`` 1001 Set system UUID. 1002ERST 1003 1004DEFHEADING() 1005 1006DEFHEADING(Block device options:) 1007 1008DEF("fda", HAS_ARG, QEMU_OPTION_fda, 1009 "-fda/-fdb file use 'file' as floppy disk 0/1 image\n", QEMU_ARCH_ALL) 1010DEF("fdb", HAS_ARG, QEMU_OPTION_fdb, "", QEMU_ARCH_ALL) 1011SRST 1012``-fda file`` 1013 \ 1014``-fdb file`` 1015 Use file as floppy disk 0/1 image (see the :ref:`disk images` chapter in 1016 the System Emulation Users Guide). 1017ERST 1018 1019DEF("hda", HAS_ARG, QEMU_OPTION_hda, 1020 "-hda/-hdb file use 'file' as IDE hard disk 0/1 image\n", QEMU_ARCH_ALL) 1021DEF("hdb", HAS_ARG, QEMU_OPTION_hdb, "", QEMU_ARCH_ALL) 1022DEF("hdc", HAS_ARG, QEMU_OPTION_hdc, 1023 "-hdc/-hdd file use 'file' as IDE hard disk 2/3 image\n", QEMU_ARCH_ALL) 1024DEF("hdd", HAS_ARG, QEMU_OPTION_hdd, "", QEMU_ARCH_ALL) 1025SRST 1026``-hda file`` 1027 \ 1028``-hdb file`` 1029 \ 1030``-hdc file`` 1031 \ 1032``-hdd file`` 1033 Use file as hard disk 0, 1, 2 or 3 image (see the :ref:`disk images` 1034 chapter in the System Emulation Users Guide). 1035ERST 1036 1037DEF("cdrom", HAS_ARG, QEMU_OPTION_cdrom, 1038 "-cdrom file use 'file' as IDE cdrom image (cdrom is ide1 master)\n", 1039 QEMU_ARCH_ALL) 1040SRST 1041``-cdrom file`` 1042 Use file as CD-ROM image (you cannot use ``-hdc`` and ``-cdrom`` at 1043 the same time). You can use the host CD-ROM by using ``/dev/cdrom`` 1044 as filename. 1045ERST 1046 1047DEF("blockdev", HAS_ARG, QEMU_OPTION_blockdev, 1048 "-blockdev [driver=]driver[,node-name=N][,discard=ignore|unmap]\n" 1049 " [,cache.direct=on|off][,cache.no-flush=on|off]\n" 1050 " [,read-only=on|off][,auto-read-only=on|off]\n" 1051 " [,force-share=on|off][,detect-zeroes=on|off|unmap]\n" 1052 " [,driver specific parameters...]\n" 1053 " configure a block backend\n", QEMU_ARCH_ALL) 1054SRST 1055``-blockdev option[,option[,option[,...]]]`` 1056 Define a new block driver node. Some of the options apply to all 1057 block drivers, other options are only accepted for a specific block 1058 driver. See below for a list of generic options and options for the 1059 most common block drivers. 1060 1061 Options that expect a reference to another node (e.g. ``file``) can 1062 be given in two ways. Either you specify the node name of an already 1063 existing node (file=node-name), or you define a new node inline, 1064 adding options for the referenced node after a dot 1065 (file.filename=path,file.aio=native). 1066 1067 A block driver node created with ``-blockdev`` can be used for a 1068 guest device by specifying its node name for the ``drive`` property 1069 in a ``-device`` argument that defines a block device. 1070 1071 ``Valid options for any block driver node:`` 1072 ``driver`` 1073 Specifies the block driver to use for the given node. 1074 1075 ``node-name`` 1076 This defines the name of the block driver node by which it 1077 will be referenced later. The name must be unique, i.e. it 1078 must not match the name of a different block driver node, or 1079 (if you use ``-drive`` as well) the ID of a drive. 1080 1081 If no node name is specified, it is automatically generated. 1082 The generated node name is not intended to be predictable 1083 and changes between QEMU invocations. For the top level, an 1084 explicit node name must be specified. 1085 1086 ``read-only`` 1087 Open the node read-only. Guest write attempts will fail. 1088 1089 Note that some block drivers support only read-only access, 1090 either generally or in certain configurations. In this case, 1091 the default value ``read-only=off`` does not work and the 1092 option must be specified explicitly. 1093 1094 ``auto-read-only`` 1095 If ``auto-read-only=on`` is set, QEMU may fall back to 1096 read-only usage even when ``read-only=off`` is requested, or 1097 even switch between modes as needed, e.g. depending on 1098 whether the image file is writable or whether a writing user 1099 is attached to the node. 1100 1101 ``force-share`` 1102 Override the image locking system of QEMU by forcing the 1103 node to utilize weaker shared access for permissions where 1104 it would normally request exclusive access. When there is 1105 the potential for multiple instances to have the same file 1106 open (whether this invocation of QEMU is the first or the 1107 second instance), both instances must permit shared access 1108 for the second instance to succeed at opening the file. 1109 1110 Enabling ``force-share=on`` requires ``read-only=on``. 1111 1112 ``cache.direct`` 1113 The host page cache can be avoided with ``cache.direct=on``. 1114 This will attempt to do disk IO directly to the guest's 1115 memory. QEMU may still perform an internal copy of the data. 1116 1117 ``cache.no-flush`` 1118 In case you don't care about data integrity over host 1119 failures, you can use ``cache.no-flush=on``. This option 1120 tells QEMU that it never needs to write any data to the disk 1121 but can instead keep things in cache. If anything goes 1122 wrong, like your host losing power, the disk storage getting 1123 disconnected accidentally, etc. your image will most 1124 probably be rendered unusable. 1125 1126 ``discard=discard`` 1127 discard is one of "ignore" (or "off") or "unmap" (or "on") 1128 and controls whether ``discard`` (also known as ``trim`` or 1129 ``unmap``) requests are ignored or passed to the filesystem. 1130 Some machine types may not support discard requests. 1131 1132 ``detect-zeroes=detect-zeroes`` 1133 detect-zeroes is "off", "on" or "unmap" and enables the 1134 automatic conversion of plain zero writes by the OS to 1135 driver specific optimized zero write commands. You may even 1136 choose "unmap" if discard is set to "unmap" to allow a zero 1137 write to be converted to an ``unmap`` operation. 1138 1139 ``Driver-specific options for file`` 1140 This is the protocol-level block driver for accessing regular 1141 files. 1142 1143 ``filename`` 1144 The path to the image file in the local filesystem 1145 1146 ``aio`` 1147 Specifies the AIO backend (threads/native/io_uring, 1148 default: threads) 1149 1150 ``locking`` 1151 Specifies whether the image file is protected with Linux OFD 1152 / POSIX locks. The default is to use the Linux Open File 1153 Descriptor API if available, otherwise no lock is applied. 1154 (auto/on/off, default: auto) 1155 1156 Example: 1157 1158 :: 1159 1160 -blockdev driver=file,node-name=disk,filename=disk.img 1161 1162 ``Driver-specific options for raw`` 1163 This is the image format block driver for raw images. It is 1164 usually stacked on top of a protocol level block driver such as 1165 ``file``. 1166 1167 ``file`` 1168 Reference to or definition of the data source block driver 1169 node (e.g. a ``file`` driver node) 1170 1171 Example 1: 1172 1173 :: 1174 1175 -blockdev driver=file,node-name=disk_file,filename=disk.img 1176 -blockdev driver=raw,node-name=disk,file=disk_file 1177 1178 Example 2: 1179 1180 :: 1181 1182 -blockdev driver=raw,node-name=disk,file.driver=file,file.filename=disk.img 1183 1184 ``Driver-specific options for qcow2`` 1185 This is the image format block driver for qcow2 images. It is 1186 usually stacked on top of a protocol level block driver such as 1187 ``file``. 1188 1189 ``file`` 1190 Reference to or definition of the data source block driver 1191 node (e.g. a ``file`` driver node) 1192 1193 ``backing`` 1194 Reference to or definition of the backing file block device 1195 (default is taken from the image file). It is allowed to 1196 pass ``null`` here in order to disable the default backing 1197 file. 1198 1199 ``lazy-refcounts`` 1200 Whether to enable the lazy refcounts feature (on/off; 1201 default is taken from the image file) 1202 1203 ``cache-size`` 1204 The maximum total size of the L2 table and refcount block 1205 caches in bytes (default: the sum of l2-cache-size and 1206 refcount-cache-size) 1207 1208 ``l2-cache-size`` 1209 The maximum size of the L2 table cache in bytes (default: if 1210 cache-size is not specified - 32M on Linux platforms, and 8M 1211 on non-Linux platforms; otherwise, as large as possible 1212 within the cache-size, while permitting the requested or the 1213 minimal refcount cache size) 1214 1215 ``refcount-cache-size`` 1216 The maximum size of the refcount block cache in bytes 1217 (default: 4 times the cluster size; or if cache-size is 1218 specified, the part of it which is not used for the L2 1219 cache) 1220 1221 ``cache-clean-interval`` 1222 Clean unused entries in the L2 and refcount caches. The 1223 interval is in seconds. The default value is 600 on 1224 supporting platforms, and 0 on other platforms. Setting it 1225 to 0 disables this feature. 1226 1227 ``pass-discard-request`` 1228 Whether discard requests to the qcow2 device should be 1229 forwarded to the data source (on/off; default: on if 1230 discard=unmap is specified, off otherwise) 1231 1232 ``pass-discard-snapshot`` 1233 Whether discard requests for the data source should be 1234 issued when a snapshot operation (e.g. deleting a snapshot) 1235 frees clusters in the qcow2 file (on/off; default: on) 1236 1237 ``pass-discard-other`` 1238 Whether discard requests for the data source should be 1239 issued on other occasions where a cluster gets freed 1240 (on/off; default: off) 1241 1242 ``overlap-check`` 1243 Which overlap checks to perform for writes to the image 1244 (none/constant/cached/all; default: cached). For details or 1245 finer granularity control refer to the QAPI documentation of 1246 ``blockdev-add``. 1247 1248 Example 1: 1249 1250 :: 1251 1252 -blockdev driver=file,node-name=my_file,filename=/tmp/disk.qcow2 1253 -blockdev driver=qcow2,node-name=hda,file=my_file,overlap-check=none,cache-size=16777216 1254 1255 Example 2: 1256 1257 :: 1258 1259 -blockdev driver=qcow2,node-name=disk,file.driver=http,file.filename=http://example.com/image.qcow2 1260 1261 ``Driver-specific options for other drivers`` 1262 Please refer to the QAPI documentation of the ``blockdev-add`` 1263 QMP command. 1264ERST 1265 1266DEF("drive", HAS_ARG, QEMU_OPTION_drive, 1267 "-drive [file=file][,if=type][,bus=n][,unit=m][,media=d][,index=i]\n" 1268 " [,cache=writethrough|writeback|none|directsync|unsafe][,format=f]\n" 1269 " [,snapshot=on|off][,rerror=ignore|stop|report]\n" 1270 " [,werror=ignore|stop|report|enospc][,id=name]\n" 1271 " [,aio=threads|native|io_uring]\n" 1272 " [,readonly=on|off][,copy-on-read=on|off]\n" 1273 " [,discard=ignore|unmap][,detect-zeroes=on|off|unmap]\n" 1274 " [[,bps=b]|[[,bps_rd=r][,bps_wr=w]]]\n" 1275 " [[,iops=i]|[[,iops_rd=r][,iops_wr=w]]]\n" 1276 " [[,bps_max=bm]|[[,bps_rd_max=rm][,bps_wr_max=wm]]]\n" 1277 " [[,iops_max=im]|[[,iops_rd_max=irm][,iops_wr_max=iwm]]]\n" 1278 " [[,iops_size=is]]\n" 1279 " [[,group=g]]\n" 1280 " use 'file' as a drive image\n", QEMU_ARCH_ALL) 1281SRST 1282``-drive option[,option[,option[,...]]]`` 1283 Define a new drive. This includes creating a block driver node (the 1284 backend) as well as a guest device, and is mostly a shortcut for 1285 defining the corresponding ``-blockdev`` and ``-device`` options. 1286 1287 ``-drive`` accepts all options that are accepted by ``-blockdev``. 1288 In addition, it knows the following options: 1289 1290 ``file=file`` 1291 This option defines which disk image (see the :ref:`disk images` 1292 chapter in the System Emulation Users Guide) to use with this drive. 1293 If the filename contains comma, you must double it (for instance, 1294 "file=my,,file" to use file "my,file"). 1295 1296 Special files such as iSCSI devices can be specified using 1297 protocol specific URLs. See the section for "Device URL Syntax" 1298 for more information. 1299 1300 ``if=interface`` 1301 This option defines on which type on interface the drive is 1302 connected. Available types are: ide, scsi, sd, mtd, floppy, 1303 pflash, virtio, none. 1304 1305 ``bus=bus,unit=unit`` 1306 These options define where is connected the drive by defining 1307 the bus number and the unit id. 1308 1309 ``index=index`` 1310 This option defines where is connected the drive by using an 1311 index in the list of available connectors of a given interface 1312 type. 1313 1314 ``media=media`` 1315 This option defines the type of the media: disk or cdrom. 1316 1317 ``snapshot=snapshot`` 1318 snapshot is "on" or "off" and controls snapshot mode for the 1319 given drive (see ``-snapshot``). 1320 1321 ``cache=cache`` 1322 cache is "none", "writeback", "unsafe", "directsync" or 1323 "writethrough" and controls how the host cache is used to access 1324 block data. This is a shortcut that sets the ``cache.direct`` 1325 and ``cache.no-flush`` options (as in ``-blockdev``), and 1326 additionally ``cache.writeback``, which provides a default for 1327 the ``write-cache`` option of block guest devices (as in 1328 ``-device``). The modes correspond to the following settings: 1329 1330 ============= =============== ============ ============== 1331 \ cache.writeback cache.direct cache.no-flush 1332 ============= =============== ============ ============== 1333 writeback on off off 1334 none on on off 1335 writethrough off off off 1336 directsync off on off 1337 unsafe on off on 1338 ============= =============== ============ ============== 1339 1340 The default mode is ``cache=writeback``. 1341 1342 ``aio=aio`` 1343 aio is "threads", "native", or "io_uring" and selects between pthread 1344 based disk I/O, native Linux AIO, or Linux io_uring API. 1345 1346 ``format=format`` 1347 Specify which disk format will be used rather than detecting the 1348 format. Can be used to specify format=raw to avoid interpreting 1349 an untrusted format header. 1350 1351 ``werror=action,rerror=action`` 1352 Specify which action to take on write and read errors. Valid 1353 actions are: "ignore" (ignore the error and try to continue), 1354 "stop" (pause QEMU), "report" (report the error to the guest), 1355 "enospc" (pause QEMU only if the host disk is full; report the 1356 error to the guest otherwise). The default setting is 1357 ``werror=enospc`` and ``rerror=report``. 1358 1359 ``copy-on-read=copy-on-read`` 1360 copy-on-read is "on" or "off" and enables whether to copy read 1361 backing file sectors into the image file. 1362 1363 ``bps=b,bps_rd=r,bps_wr=w`` 1364 Specify bandwidth throttling limits in bytes per second, either 1365 for all request types or for reads or writes only. Small values 1366 can lead to timeouts or hangs inside the guest. A safe minimum 1367 for disks is 2 MB/s. 1368 1369 ``bps_max=bm,bps_rd_max=rm,bps_wr_max=wm`` 1370 Specify bursts in bytes per second, either for all request types 1371 or for reads or writes only. Bursts allow the guest I/O to spike 1372 above the limit temporarily. 1373 1374 ``iops=i,iops_rd=r,iops_wr=w`` 1375 Specify request rate limits in requests per second, either for 1376 all request types or for reads or writes only. 1377 1378 ``iops_max=bm,iops_rd_max=rm,iops_wr_max=wm`` 1379 Specify bursts in requests per second, either for all request 1380 types or for reads or writes only. Bursts allow the guest I/O to 1381 spike above the limit temporarily. 1382 1383 ``iops_size=is`` 1384 Let every is bytes of a request count as a new request for iops 1385 throttling purposes. Use this option to prevent guests from 1386 circumventing iops limits by sending fewer but larger requests. 1387 1388 ``group=g`` 1389 Join a throttling quota group with given name g. All drives that 1390 are members of the same group are accounted for together. Use 1391 this option to prevent guests from circumventing throttling 1392 limits by using many small disks instead of a single larger 1393 disk. 1394 1395 By default, the ``cache.writeback=on`` mode is used. It will report 1396 data writes as completed as soon as the data is present in the host 1397 page cache. This is safe as long as your guest OS makes sure to 1398 correctly flush disk caches where needed. If your guest OS does not 1399 handle volatile disk write caches correctly and your host crashes or 1400 loses power, then the guest may experience data corruption. 1401 1402 For such guests, you should consider using ``cache.writeback=off``. 1403 This means that the host page cache will be used to read and write 1404 data, but write notification will be sent to the guest only after 1405 QEMU has made sure to flush each write to the disk. Be aware that 1406 this has a major impact on performance. 1407 1408 When using the ``-snapshot`` option, unsafe caching is always used. 1409 1410 Copy-on-read avoids accessing the same backing file sectors 1411 repeatedly and is useful when the backing file is over a slow 1412 network. By default copy-on-read is off. 1413 1414 Instead of ``-cdrom`` you can use: 1415 1416 .. parsed-literal:: 1417 1418 |qemu_system| -drive file=file,index=2,media=cdrom 1419 1420 Instead of ``-hda``, ``-hdb``, ``-hdc``, ``-hdd``, you can use: 1421 1422 .. parsed-literal:: 1423 1424 |qemu_system| -drive file=file,index=0,media=disk 1425 |qemu_system| -drive file=file,index=1,media=disk 1426 |qemu_system| -drive file=file,index=2,media=disk 1427 |qemu_system| -drive file=file,index=3,media=disk 1428 1429 You can open an image using pre-opened file descriptors from an fd 1430 set: 1431 1432 .. parsed-literal:: 1433 1434 |qemu_system| \\ 1435 -add-fd fd=3,set=2,opaque="rdwr:/path/to/file" \\ 1436 -add-fd fd=4,set=2,opaque="rdonly:/path/to/file" \\ 1437 -drive file=/dev/fdset/2,index=0,media=disk 1438 1439 You can connect a CDROM to the slave of ide0: 1440 1441 .. parsed-literal:: 1442 1443 |qemu_system_x86| -drive file=file,if=ide,index=1,media=cdrom 1444 1445 If you don't specify the "file=" argument, you define an empty 1446 drive: 1447 1448 .. parsed-literal:: 1449 1450 |qemu_system_x86| -drive if=ide,index=1,media=cdrom 1451 1452 Instead of ``-fda``, ``-fdb``, you can use: 1453 1454 .. parsed-literal:: 1455 1456 |qemu_system_x86| -drive file=file,index=0,if=floppy 1457 |qemu_system_x86| -drive file=file,index=1,if=floppy 1458 1459 By default, interface is "ide" and index is automatically 1460 incremented: 1461 1462 .. parsed-literal:: 1463 1464 |qemu_system_x86| -drive file=a -drive file=b" 1465 1466 is interpreted like: 1467 1468 .. parsed-literal:: 1469 1470 |qemu_system_x86| -hda a -hdb b 1471ERST 1472 1473DEF("mtdblock", HAS_ARG, QEMU_OPTION_mtdblock, 1474 "-mtdblock file use 'file' as on-board Flash memory image\n", 1475 QEMU_ARCH_ALL) 1476SRST 1477``-mtdblock file`` 1478 Use file as on-board Flash memory image. 1479ERST 1480 1481DEF("sd", HAS_ARG, QEMU_OPTION_sd, 1482 "-sd file use 'file' as SecureDigital card image\n", QEMU_ARCH_ALL) 1483SRST 1484``-sd file`` 1485 Use file as SecureDigital card image. 1486ERST 1487 1488DEF("pflash", HAS_ARG, QEMU_OPTION_pflash, 1489 "-pflash file use 'file' as a parallel flash image\n", QEMU_ARCH_ALL) 1490SRST 1491``-pflash file`` 1492 Use file as a parallel flash image. 1493ERST 1494 1495DEF("snapshot", 0, QEMU_OPTION_snapshot, 1496 "-snapshot write to temporary files instead of disk image files\n", 1497 QEMU_ARCH_ALL) 1498SRST 1499``-snapshot`` 1500 Write to temporary files instead of disk image files. In this case, 1501 the raw disk image you use is not written back. You can however 1502 force the write back by pressing C-a s (see the :ref:`disk images` 1503 chapter in the System Emulation Users Guide). 1504ERST 1505 1506DEF("fsdev", HAS_ARG, QEMU_OPTION_fsdev, 1507 "-fsdev local,id=id,path=path,security_model=mapped-xattr|mapped-file|passthrough|none\n" 1508 " [,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode]\n" 1509 " [[,throttling.bps-total=b]|[[,throttling.bps-read=r][,throttling.bps-write=w]]]\n" 1510 " [[,throttling.iops-total=i]|[[,throttling.iops-read=r][,throttling.iops-write=w]]]\n" 1511 " [[,throttling.bps-total-max=bm]|[[,throttling.bps-read-max=rm][,throttling.bps-write-max=wm]]]\n" 1512 " [[,throttling.iops-total-max=im]|[[,throttling.iops-read-max=irm][,throttling.iops-write-max=iwm]]]\n" 1513 " [[,throttling.iops-size=is]]\n" 1514 "-fsdev proxy,id=id,socket=socket[,writeout=immediate][,readonly=on]\n" 1515 "-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=immediate][,readonly=on]\n" 1516 "-fsdev synth,id=id\n", 1517 QEMU_ARCH_ALL) 1518 1519SRST 1520``-fsdev local,id=id,path=path,security_model=security_model [,writeout=writeout][,readonly=on][,fmode=fmode][,dmode=dmode] [,throttling.option=value[,throttling.option=value[,...]]]`` 1521 \ 1522``-fsdev proxy,id=id,socket=socket[,writeout=writeout][,readonly=on]`` 1523 \ 1524``-fsdev proxy,id=id,sock_fd=sock_fd[,writeout=writeout][,readonly=on]`` 1525 \ 1526``-fsdev synth,id=id[,readonly=on]`` 1527 Define a new file system device. Valid options are: 1528 1529 ``local`` 1530 Accesses to the filesystem are done by QEMU. 1531 1532 ``proxy`` 1533 Accesses to the filesystem are done by virtfs-proxy-helper(1). 1534 1535 ``synth`` 1536 Synthetic filesystem, only used by QTests. 1537 1538 ``id=id`` 1539 Specifies identifier for this device. 1540 1541 ``path=path`` 1542 Specifies the export path for the file system device. Files 1543 under this path will be available to the 9p client on the guest. 1544 1545 ``security_model=security_model`` 1546 Specifies the security model to be used for this export path. 1547 Supported security models are "passthrough", "mapped-xattr", 1548 "mapped-file" and "none". In "passthrough" security model, files 1549 are stored using the same credentials as they are created on the 1550 guest. This requires QEMU to run as root. In "mapped-xattr" 1551 security model, some of the file attributes like uid, gid, mode 1552 bits and link target are stored as file attributes. For 1553 "mapped-file" these attributes are stored in the hidden 1554 .virtfs\_metadata directory. Directories exported by this 1555 security model cannot interact with other unix tools. "none" 1556 security model is same as passthrough except the sever won't 1557 report failures if it fails to set file attributes like 1558 ownership. Security model is mandatory only for local fsdriver. 1559 Other fsdrivers (like proxy) don't take security model as a 1560 parameter. 1561 1562 ``writeout=writeout`` 1563 This is an optional argument. The only supported value is 1564 "immediate". This means that host page cache will be used to 1565 read and write data but write notification will be sent to the 1566 guest only when the data has been reported as written by the 1567 storage subsystem. 1568 1569 ``readonly=on`` 1570 Enables exporting 9p share as a readonly mount for guests. By 1571 default read-write access is given. 1572 1573 ``socket=socket`` 1574 Enables proxy filesystem driver to use passed socket file for 1575 communicating with virtfs-proxy-helper(1). 1576 1577 ``sock_fd=sock_fd`` 1578 Enables proxy filesystem driver to use passed socket descriptor 1579 for communicating with virtfs-proxy-helper(1). Usually a helper 1580 like libvirt will create socketpair and pass one of the fds as 1581 sock\_fd. 1582 1583 ``fmode=fmode`` 1584 Specifies the default mode for newly created files on the host. 1585 Works only with security models "mapped-xattr" and 1586 "mapped-file". 1587 1588 ``dmode=dmode`` 1589 Specifies the default mode for newly created directories on the 1590 host. Works only with security models "mapped-xattr" and 1591 "mapped-file". 1592 1593 ``throttling.bps-total=b,throttling.bps-read=r,throttling.bps-write=w`` 1594 Specify bandwidth throttling limits in bytes per second, either 1595 for all request types or for reads or writes only. 1596 1597 ``throttling.bps-total-max=bm,bps-read-max=rm,bps-write-max=wm`` 1598 Specify bursts in bytes per second, either for all request types 1599 or for reads or writes only. Bursts allow the guest I/O to spike 1600 above the limit temporarily. 1601 1602 ``throttling.iops-total=i,throttling.iops-read=r, throttling.iops-write=w`` 1603 Specify request rate limits in requests per second, either for 1604 all request types or for reads or writes only. 1605 1606 ``throttling.iops-total-max=im,throttling.iops-read-max=irm, throttling.iops-write-max=iwm`` 1607 Specify bursts in requests per second, either for all request 1608 types or for reads or writes only. Bursts allow the guest I/O to 1609 spike above the limit temporarily. 1610 1611 ``throttling.iops-size=is`` 1612 Let every is bytes of a request count as a new request for iops 1613 throttling purposes. 1614 1615 -fsdev option is used along with -device driver "virtio-9p-...". 1616 1617``-device virtio-9p-type,fsdev=id,mount_tag=mount_tag`` 1618 Options for virtio-9p-... driver are: 1619 1620 ``type`` 1621 Specifies the variant to be used. Supported values are "pci", 1622 "ccw" or "device", depending on the machine type. 1623 1624 ``fsdev=id`` 1625 Specifies the id value specified along with -fsdev option. 1626 1627 ``mount_tag=mount_tag`` 1628 Specifies the tag name to be used by the guest to mount this 1629 export point. 1630ERST 1631 1632DEF("virtfs", HAS_ARG, QEMU_OPTION_virtfs, 1633 "-virtfs local,path=path,mount_tag=tag,security_model=mapped-xattr|mapped-file|passthrough|none\n" 1634 " [,id=id][,writeout=immediate][,readonly=on][,fmode=fmode][,dmode=dmode][,multidevs=remap|forbid|warn]\n" 1635 "-virtfs proxy,mount_tag=tag,socket=socket[,id=id][,writeout=immediate][,readonly=on]\n" 1636 "-virtfs proxy,mount_tag=tag,sock_fd=sock_fd[,id=id][,writeout=immediate][,readonly=on]\n" 1637 "-virtfs synth,mount_tag=tag[,id=id][,readonly=on]\n", 1638 QEMU_ARCH_ALL) 1639 1640SRST 1641``-virtfs local,path=path,mount_tag=mount_tag ,security_model=security_model[,writeout=writeout][,readonly=on] [,fmode=fmode][,dmode=dmode][,multidevs=multidevs]`` 1642 \ 1643``-virtfs proxy,socket=socket,mount_tag=mount_tag [,writeout=writeout][,readonly=on]`` 1644 \ 1645``-virtfs proxy,sock_fd=sock_fd,mount_tag=mount_tag [,writeout=writeout][,readonly=on]`` 1646 \ 1647``-virtfs synth,mount_tag=mount_tag`` 1648 Define a new virtual filesystem device and expose it to the guest using 1649 a virtio-9p-device (a.k.a. 9pfs), which essentially means that a certain 1650 directory on host is made directly accessible by guest as a pass-through 1651 file system by using the 9P network protocol for communication between 1652 host and guests, if desired even accessible, shared by several guests 1653 simultaniously. 1654 1655 Note that ``-virtfs`` is actually just a convenience shortcut for its 1656 generalized form ``-fsdev -device virtio-9p-pci``. 1657 1658 The general form of pass-through file system options are: 1659 1660 ``local`` 1661 Accesses to the filesystem are done by QEMU. 1662 1663 ``proxy`` 1664 Accesses to the filesystem are done by virtfs-proxy-helper(1). 1665 1666 ``synth`` 1667 Synthetic filesystem, only used by QTests. 1668 1669 ``id=id`` 1670 Specifies identifier for the filesystem device 1671 1672 ``path=path`` 1673 Specifies the export path for the file system device. Files 1674 under this path will be available to the 9p client on the guest. 1675 1676 ``security_model=security_model`` 1677 Specifies the security model to be used for this export path. 1678 Supported security models are "passthrough", "mapped-xattr", 1679 "mapped-file" and "none". In "passthrough" security model, files 1680 are stored using the same credentials as they are created on the 1681 guest. This requires QEMU to run as root. In "mapped-xattr" 1682 security model, some of the file attributes like uid, gid, mode 1683 bits and link target are stored as file attributes. For 1684 "mapped-file" these attributes are stored in the hidden 1685 .virtfs\_metadata directory. Directories exported by this 1686 security model cannot interact with other unix tools. "none" 1687 security model is same as passthrough except the sever won't 1688 report failures if it fails to set file attributes like 1689 ownership. Security model is mandatory only for local fsdriver. 1690 Other fsdrivers (like proxy) don't take security model as a 1691 parameter. 1692 1693 ``writeout=writeout`` 1694 This is an optional argument. The only supported value is 1695 "immediate". This means that host page cache will be used to 1696 read and write data but write notification will be sent to the 1697 guest only when the data has been reported as written by the 1698 storage subsystem. 1699 1700 ``readonly=on`` 1701 Enables exporting 9p share as a readonly mount for guests. By 1702 default read-write access is given. 1703 1704 ``socket=socket`` 1705 Enables proxy filesystem driver to use passed socket file for 1706 communicating with virtfs-proxy-helper(1). Usually a helper like 1707 libvirt will create socketpair and pass one of the fds as 1708 sock\_fd. 1709 1710 ``sock_fd`` 1711 Enables proxy filesystem driver to use passed 'sock\_fd' as the 1712 socket descriptor for interfacing with virtfs-proxy-helper(1). 1713 1714 ``fmode=fmode`` 1715 Specifies the default mode for newly created files on the host. 1716 Works only with security models "mapped-xattr" and 1717 "mapped-file". 1718 1719 ``dmode=dmode`` 1720 Specifies the default mode for newly created directories on the 1721 host. Works only with security models "mapped-xattr" and 1722 "mapped-file". 1723 1724 ``mount_tag=mount_tag`` 1725 Specifies the tag name to be used by the guest to mount this 1726 export point. 1727 1728 ``multidevs=multidevs`` 1729 Specifies how to deal with multiple devices being shared with a 1730 9p export. Supported behaviours are either "remap", "forbid" or 1731 "warn". The latter is the default behaviour on which virtfs 9p 1732 expects only one device to be shared with the same export, and 1733 if more than one device is shared and accessed via the same 9p 1734 export then only a warning message is logged (once) by qemu on 1735 host side. In order to avoid file ID collisions on guest you 1736 should either create a separate virtfs export for each device to 1737 be shared with guests (recommended way) or you might use "remap" 1738 instead which allows you to share multiple devices with only one 1739 export instead, which is achieved by remapping the original 1740 inode numbers from host to guest in a way that would prevent 1741 such collisions. Remapping inodes in such use cases is required 1742 because the original device IDs from host are never passed and 1743 exposed on guest. Instead all files of an export shared with 1744 virtfs always share the same device id on guest. So two files 1745 with identical inode numbers but from actually different devices 1746 on host would otherwise cause a file ID collision and hence 1747 potential misbehaviours on guest. "forbid" on the other hand 1748 assumes like "warn" that only one device is shared by the same 1749 export, however it will not only log a warning message but also 1750 deny access to additional devices on guest. Note though that 1751 "forbid" does currently not block all possible file access 1752 operations (e.g. readdir() would still return entries from other 1753 devices). 1754ERST 1755 1756DEF("iscsi", HAS_ARG, QEMU_OPTION_iscsi, 1757 "-iscsi [user=user][,password=password]\n" 1758 " [,header-digest=CRC32C|CR32C-NONE|NONE-CRC32C|NONE\n" 1759 " [,initiator-name=initiator-iqn][,id=target-iqn]\n" 1760 " [,timeout=timeout]\n" 1761 " iSCSI session parameters\n", QEMU_ARCH_ALL) 1762 1763SRST 1764``-iscsi`` 1765 Configure iSCSI session parameters. 1766ERST 1767 1768DEFHEADING() 1769 1770DEFHEADING(USB convenience options:) 1771 1772DEF("usb", 0, QEMU_OPTION_usb, 1773 "-usb enable on-board USB host controller (if not enabled by default)\n", 1774 QEMU_ARCH_ALL) 1775SRST 1776``-usb`` 1777 Enable USB emulation on machine types with an on-board USB host 1778 controller (if not enabled by default). Note that on-board USB host 1779 controllers may not support USB 3.0. In this case 1780 ``-device qemu-xhci`` can be used instead on machines with PCI. 1781ERST 1782 1783DEF("usbdevice", HAS_ARG, QEMU_OPTION_usbdevice, 1784 "-usbdevice name add the host or guest USB device 'name'\n", 1785 QEMU_ARCH_ALL) 1786SRST 1787``-usbdevice devname`` 1788 Add the USB device devname, and enable an on-board USB controller 1789 if possible and necessary (just like it can be done via 1790 ``-machine usb=on``). Note that this option is mainly intended for 1791 the user's convenience only. More fine-grained control can be 1792 achieved by selecting a USB host controller (if necessary) and the 1793 desired USB device via the ``-device`` option instead. For example, 1794 instead of using ``-usbdevice mouse`` it is possible to use 1795 ``-device qemu-xhci -device usb-mouse`` to connect the USB mouse 1796 to a USB 3.0 controller instead (at least on machines that support 1797 PCI and do not have an USB controller enabled by default yet). 1798 For more details, see the chapter about 1799 :ref:`Connecting USB devices` in the System Emulation Users Guide. 1800 Possible devices for devname are: 1801 1802 ``braille`` 1803 Braille device. This will use BrlAPI to display the braille 1804 output on a real or fake device (i.e. it also creates a 1805 corresponding ``braille`` chardev automatically beside the 1806 ``usb-braille`` USB device). 1807 1808 ``keyboard`` 1809 Standard USB keyboard. Will override the PS/2 keyboard (if present). 1810 1811 ``mouse`` 1812 Virtual Mouse. This will override the PS/2 mouse emulation when 1813 activated. 1814 1815 ``tablet`` 1816 Pointer device that uses absolute coordinates (like a 1817 touchscreen). This means QEMU is able to report the mouse 1818 position without having to grab the mouse. Also overrides the 1819 PS/2 mouse emulation when activated. 1820 1821 ``wacom-tablet`` 1822 Wacom PenPartner USB tablet. 1823 1824 1825ERST 1826 1827DEFHEADING() 1828 1829DEFHEADING(Display options:) 1830 1831DEF("display", HAS_ARG, QEMU_OPTION_display, 1832#if defined(CONFIG_SPICE) 1833 "-display spice-app[,gl=on|off]\n" 1834#endif 1835#if defined(CONFIG_SDL) 1836 "-display sdl[,alt_grab=on|off][,ctrl_grab=on|off][,gl=on|core|es|off]\n" 1837 " [,grab-mod=<mod>][,show-cursor=on|off][,window-close=on|off]\n" 1838#endif 1839#if defined(CONFIG_GTK) 1840 "-display gtk[,full-screen=on|off][,gl=on|off][,grab-on-hover=on|off]\n" 1841 " [,show-cursor=on|off][,window-close=on|off]\n" 1842#endif 1843#if defined(CONFIG_VNC) 1844 "-display vnc=<display>[,<optargs>]\n" 1845#endif 1846#if defined(CONFIG_CURSES) 1847 "-display curses[,charset=<encoding>]\n" 1848#endif 1849#if defined(CONFIG_OPENGL) 1850 "-display egl-headless[,rendernode=<file>]\n" 1851#endif 1852 "-display none\n" 1853 " select display backend type\n" 1854 " The default display is equivalent to\n " 1855#if defined(CONFIG_GTK) 1856 "\"-display gtk\"\n" 1857#elif defined(CONFIG_SDL) 1858 "\"-display sdl\"\n" 1859#elif defined(CONFIG_COCOA) 1860 "\"-display cocoa\"\n" 1861#elif defined(CONFIG_VNC) 1862 "\"-vnc localhost:0,to=99,id=default\"\n" 1863#else 1864 "\"-display none\"\n" 1865#endif 1866 , QEMU_ARCH_ALL) 1867SRST 1868``-display type`` 1869 Select type of display to use. This option is a replacement for the 1870 old style -sdl/-curses/... options. Use ``-display help`` to list 1871 the available display types. Valid values for type are 1872 1873 ``spice-app[,gl=on|off]`` 1874 Start QEMU as a Spice server and launch the default Spice client 1875 application. The Spice server will redirect the serial consoles 1876 and QEMU monitors. (Since 4.0) 1877 1878 ``sdl`` 1879 Display video output via SDL (usually in a separate graphics 1880 window; see the SDL documentation for other possibilities). 1881 Valid parameters are: 1882 1883 ``grab-mod=<mods>`` : Used to select the modifier keys for toggling 1884 the mouse grabbing in conjunction with the "g" key. `<mods>` can be 1885 either `lshift-lctrl-lalt` or `rctrl`. 1886 1887 ``alt_grab=on|off`` : Use Control+Alt+Shift-g to toggle mouse grabbing. 1888 This parameter is deprecated - use ``grab-mod`` instead. 1889 1890 ``ctrl_grab=on|off`` : Use Right-Control-g to toggle mouse grabbing. 1891 This parameter is deprecated - use ``grab-mod`` instead. 1892 1893 ``gl=on|off|core|es`` : Use OpenGL for displaying 1894 1895 ``show-cursor=on|off`` : Force showing the mouse cursor 1896 1897 ``window-close=on|off`` : Allow to quit qemu with window close button 1898 1899 ``gtk`` 1900 Display video output in a GTK window. This interface provides 1901 drop-down menus and other UI elements to configure and control 1902 the VM during runtime. Valid parameters are: 1903 1904 ``full-screen=on|off`` : Start in fullscreen mode 1905 1906 ``gl=on|off`` : Use OpenGL for displaying 1907 1908 ``grab-on-hover=on|off`` : Grab keyboard input on mouse hover 1909 1910 ``show-cursor=on|off`` : Force showing the mouse cursor 1911 1912 ``window-close=on|off`` : Allow to quit qemu with window close button 1913 1914 ``curses[,charset=<encoding>]`` 1915 Display video output via curses. For graphics device models 1916 which support a text mode, QEMU can display this output using a 1917 curses/ncurses interface. Nothing is displayed when the graphics 1918 device is in graphical mode or if the graphics device does not 1919 support a text mode. Generally only the VGA device models 1920 support text mode. The font charset used by the guest can be 1921 specified with the ``charset`` option, for example 1922 ``charset=CP850`` for IBM CP850 encoding. The default is 1923 ``CP437``. 1924 1925 ``egl-headless[,rendernode=<file>]`` 1926 Offload all OpenGL operations to a local DRI device. For any 1927 graphical display, this display needs to be paired with either 1928 VNC or SPICE displays. 1929 1930 ``vnc=<display>`` 1931 Start a VNC server on display <display> 1932 1933 ``none`` 1934 Do not display video output. The guest will still see an 1935 emulated graphics card, but its output will not be displayed to 1936 the QEMU user. This option differs from the -nographic option in 1937 that it only affects what is done with video output; -nographic 1938 also changes the destination of the serial and parallel port 1939 data. 1940ERST 1941 1942DEF("nographic", 0, QEMU_OPTION_nographic, 1943 "-nographic disable graphical output and redirect serial I/Os to console\n", 1944 QEMU_ARCH_ALL) 1945SRST 1946``-nographic`` 1947 Normally, if QEMU is compiled with graphical window support, it 1948 displays output such as guest graphics, guest console, and the QEMU 1949 monitor in a window. With this option, you can totally disable 1950 graphical output so that QEMU is a simple command line application. 1951 The emulated serial port is redirected on the console and muxed with 1952 the monitor (unless redirected elsewhere explicitly). Therefore, you 1953 can still use QEMU to debug a Linux kernel with a serial console. 1954 Use C-a h for help on switching between the console and monitor. 1955ERST 1956 1957DEF("curses", 0, QEMU_OPTION_curses, 1958 "-curses shorthand for -display curses\n", 1959 QEMU_ARCH_ALL) 1960SRST 1961``-curses`` 1962 Normally, if QEMU is compiled with graphical window support, it 1963 displays output such as guest graphics, guest console, and the QEMU 1964 monitor in a window. With this option, QEMU can display the VGA 1965 output when in text mode using a curses/ncurses interface. Nothing 1966 is displayed in graphical mode. 1967ERST 1968 1969DEF("alt-grab", 0, QEMU_OPTION_alt_grab, 1970 "-alt-grab use Ctrl-Alt-Shift to grab mouse (instead of Ctrl-Alt)\n", 1971 QEMU_ARCH_ALL) 1972SRST 1973``-alt-grab`` 1974 Use Ctrl-Alt-Shift to grab mouse (instead of Ctrl-Alt). Note that 1975 this also affects the special keys (for fullscreen, monitor-mode 1976 switching, etc). This option is deprecated - please use 1977 ``-display sdl,grab-mod=lshift-lctrl-lalt`` instead. 1978ERST 1979 1980DEF("ctrl-grab", 0, QEMU_OPTION_ctrl_grab, 1981 "-ctrl-grab use Right-Ctrl to grab mouse (instead of Ctrl-Alt)\n", 1982 QEMU_ARCH_ALL) 1983SRST 1984``-ctrl-grab`` 1985 Use Right-Ctrl to grab mouse (instead of Ctrl-Alt). Note that this 1986 also affects the special keys (for fullscreen, monitor-mode 1987 switching, etc). This option is deprecated - please use 1988 ``-display sdl,grab-mod=rctrl`` instead. 1989ERST 1990 1991DEF("no-quit", 0, QEMU_OPTION_no_quit, 1992 "-no-quit disable SDL/GTK window close capability (deprecated)\n", QEMU_ARCH_ALL) 1993SRST 1994``-no-quit`` 1995 Disable window close capability (SDL and GTK only). This option is 1996 deprecated, please use ``-display ...,window-close=off`` instead. 1997ERST 1998 1999DEF("sdl", 0, QEMU_OPTION_sdl, 2000 "-sdl shorthand for -display sdl\n", QEMU_ARCH_ALL) 2001SRST 2002``-sdl`` 2003 Enable SDL. 2004ERST 2005 2006DEF("spice", HAS_ARG, QEMU_OPTION_spice, 2007 "-spice [port=port][,tls-port=secured-port][,x509-dir=<dir>]\n" 2008 " [,x509-key-file=<file>][,x509-key-password=<file>]\n" 2009 " [,x509-cert-file=<file>][,x509-cacert-file=<file>]\n" 2010 " [,x509-dh-key-file=<file>][,addr=addr]\n" 2011 " [,ipv4=on|off][,ipv6=on|off][,unix=on|off]\n" 2012 " [,tls-ciphers=<list>]\n" 2013 " [,tls-channel=[main|display|cursor|inputs|record|playback]]\n" 2014 " [,plaintext-channel=[main|display|cursor|inputs|record|playback]]\n" 2015 " [,sasl=on|off][,disable-ticketing=on|off]\n" 2016 " [,password=<string>][,password-secret=<secret-id>]\n" 2017 " [,image-compression=[auto_glz|auto_lz|quic|glz|lz|off]]\n" 2018 " [,jpeg-wan-compression=[auto|never|always]]\n" 2019 " [,zlib-glz-wan-compression=[auto|never|always]]\n" 2020 " [,streaming-video=[off|all|filter]][,disable-copy-paste=on|off]\n" 2021 " [,disable-agent-file-xfer=on|off][,agent-mouse=[on|off]]\n" 2022 " [,playback-compression=[on|off]][,seamless-migration=[on|off]]\n" 2023 " [,gl=[on|off]][,rendernode=<file>]\n" 2024 " enable spice\n" 2025 " at least one of {port, tls-port} is mandatory\n", 2026 QEMU_ARCH_ALL) 2027SRST 2028``-spice option[,option[,...]]`` 2029 Enable the spice remote desktop protocol. Valid options are 2030 2031 ``port=<nr>`` 2032 Set the TCP port spice is listening on for plaintext channels. 2033 2034 ``addr=<addr>`` 2035 Set the IP address spice is listening on. Default is any 2036 address. 2037 2038 ``ipv4=on|off``; \ ``ipv6=on|off``; \ ``unix=on|off`` 2039 Force using the specified IP version. 2040 2041 ``password=<string>`` 2042 Set the password you need to authenticate. 2043 2044 This option is deprecated and insecure because it leaves the 2045 password visible in the process listing. Use ``password-secret`` 2046 instead. 2047 2048 ``password-secret=<secret-id>`` 2049 Set the ID of the ``secret`` object containing the password 2050 you need to authenticate. 2051 2052 ``sasl=on|off`` 2053 Require that the client use SASL to authenticate with the spice. 2054 The exact choice of authentication method used is controlled 2055 from the system / user's SASL configuration file for the 'qemu' 2056 service. This is typically found in /etc/sasl2/qemu.conf. If 2057 running QEMU as an unprivileged user, an environment variable 2058 SASL\_CONF\_PATH can be used to make it search alternate 2059 locations for the service config. While some SASL auth methods 2060 can also provide data encryption (eg GSSAPI), it is recommended 2061 that SASL always be combined with the 'tls' and 'x509' settings 2062 to enable use of SSL and server certificates. This ensures a 2063 data encryption preventing compromise of authentication 2064 credentials. 2065 2066 ``disable-ticketing=on|off`` 2067 Allow client connects without authentication. 2068 2069 ``disable-copy-paste=on|off`` 2070 Disable copy paste between the client and the guest. 2071 2072 ``disable-agent-file-xfer=on|off`` 2073 Disable spice-vdagent based file-xfer between the client and the 2074 guest. 2075 2076 ``tls-port=<nr>`` 2077 Set the TCP port spice is listening on for encrypted channels. 2078 2079 ``x509-dir=<dir>`` 2080 Set the x509 file directory. Expects same filenames as -vnc 2081 $display,x509=$dir 2082 2083 ``x509-key-file=<file>``; \ ``x509-key-password=<file>``; \ ``x509-cert-file=<file>``; \ ``x509-cacert-file=<file>``; \ ``x509-dh-key-file=<file>`` 2084 The x509 file names can also be configured individually. 2085 2086 ``tls-ciphers=<list>`` 2087 Specify which ciphers to use. 2088 2089 ``tls-channel=[main|display|cursor|inputs|record|playback]``; \ ``plaintext-channel=[main|display|cursor|inputs|record|playback]`` 2090 Force specific channel to be used with or without TLS 2091 encryption. The options can be specified multiple times to 2092 configure multiple channels. The special name "default" can be 2093 used to set the default mode. For channels which are not 2094 explicitly forced into one mode the spice client is allowed to 2095 pick tls/plaintext as he pleases. 2096 2097 ``image-compression=[auto_glz|auto_lz|quic|glz|lz|off]`` 2098 Configure image compression (lossless). Default is auto\_glz. 2099 2100 ``jpeg-wan-compression=[auto|never|always]``; \ ``zlib-glz-wan-compression=[auto|never|always]`` 2101 Configure wan image compression (lossy for slow links). Default 2102 is auto. 2103 2104 ``streaming-video=[off|all|filter]`` 2105 Configure video stream detection. Default is off. 2106 2107 ``agent-mouse=[on|off]`` 2108 Enable/disable passing mouse events via vdagent. Default is on. 2109 2110 ``playback-compression=[on|off]`` 2111 Enable/disable audio stream compression (using celt 0.5.1). 2112 Default is on. 2113 2114 ``seamless-migration=[on|off]`` 2115 Enable/disable spice seamless migration. Default is off. 2116 2117 ``gl=[on|off]`` 2118 Enable/disable OpenGL context. Default is off. 2119 2120 ``rendernode=<file>`` 2121 DRM render node for OpenGL rendering. If not specified, it will 2122 pick the first available. (Since 2.9) 2123ERST 2124 2125DEF("portrait", 0, QEMU_OPTION_portrait, 2126 "-portrait rotate graphical output 90 deg left (only PXA LCD)\n", 2127 QEMU_ARCH_ALL) 2128SRST 2129``-portrait`` 2130 Rotate graphical output 90 deg left (only PXA LCD). 2131ERST 2132 2133DEF("rotate", HAS_ARG, QEMU_OPTION_rotate, 2134 "-rotate <deg> rotate graphical output some deg left (only PXA LCD)\n", 2135 QEMU_ARCH_ALL) 2136SRST 2137``-rotate deg`` 2138 Rotate graphical output some deg left (only PXA LCD). 2139ERST 2140 2141DEF("vga", HAS_ARG, QEMU_OPTION_vga, 2142 "-vga [std|cirrus|vmware|qxl|xenfb|tcx|cg3|virtio|none]\n" 2143 " select video card type\n", QEMU_ARCH_ALL) 2144SRST 2145``-vga type`` 2146 Select type of VGA card to emulate. Valid values for type are 2147 2148 ``cirrus`` 2149 Cirrus Logic GD5446 Video card. All Windows versions starting 2150 from Windows 95 should recognize and use this graphic card. For 2151 optimal performances, use 16 bit color depth in the guest and 2152 the host OS. (This card was the default before QEMU 2.2) 2153 2154 ``std`` 2155 Standard VGA card with Bochs VBE extensions. If your guest OS 2156 supports the VESA 2.0 VBE extensions (e.g. Windows XP) and if 2157 you want to use high resolution modes (>= 1280x1024x16) then you 2158 should use this option. (This card is the default since QEMU 2159 2.2) 2160 2161 ``vmware`` 2162 VMWare SVGA-II compatible adapter. Use it if you have 2163 sufficiently recent XFree86/XOrg server or Windows guest with a 2164 driver for this card. 2165 2166 ``qxl`` 2167 QXL paravirtual graphic card. It is VGA compatible (including 2168 VESA 2.0 VBE support). Works best with qxl guest drivers 2169 installed though. Recommended choice when using the spice 2170 protocol. 2171 2172 ``tcx`` 2173 (sun4m only) Sun TCX framebuffer. This is the default 2174 framebuffer for sun4m machines and offers both 8-bit and 24-bit 2175 colour depths at a fixed resolution of 1024x768. 2176 2177 ``cg3`` 2178 (sun4m only) Sun cgthree framebuffer. This is a simple 8-bit 2179 framebuffer for sun4m machines available in both 1024x768 2180 (OpenBIOS) and 1152x900 (OBP) resolutions aimed at people 2181 wishing to run older Solaris versions. 2182 2183 ``virtio`` 2184 Virtio VGA card. 2185 2186 ``none`` 2187 Disable VGA card. 2188ERST 2189 2190DEF("full-screen", 0, QEMU_OPTION_full_screen, 2191 "-full-screen start in full screen\n", QEMU_ARCH_ALL) 2192SRST 2193``-full-screen`` 2194 Start in full screen. 2195ERST 2196 2197DEF("g", HAS_ARG, QEMU_OPTION_g , 2198 "-g WxH[xDEPTH] Set the initial graphical resolution and depth\n", 2199 QEMU_ARCH_PPC | QEMU_ARCH_SPARC | QEMU_ARCH_M68K) 2200SRST 2201``-g`` *width*\ ``x``\ *height*\ ``[x``\ *depth*\ ``]`` 2202 Set the initial graphical resolution and depth (PPC, SPARC only). 2203 2204 For PPC the default is 800x600x32. 2205 2206 For SPARC with the TCX graphics device, the default is 1024x768x8 2207 with the option of 1024x768x24. For cgthree, the default is 2208 1024x768x8 with the option of 1152x900x8 for people who wish to use 2209 OBP. 2210ERST 2211 2212DEF("vnc", HAS_ARG, QEMU_OPTION_vnc , 2213 "-vnc <display> shorthand for -display vnc=<display>\n", QEMU_ARCH_ALL) 2214SRST 2215``-vnc display[,option[,option[,...]]]`` 2216 Normally, if QEMU is compiled with graphical window support, it 2217 displays output such as guest graphics, guest console, and the QEMU 2218 monitor in a window. With this option, you can have QEMU listen on 2219 VNC display display and redirect the VGA display over the VNC 2220 session. It is very useful to enable the usb tablet device when 2221 using this option (option ``-device usb-tablet``). When using the 2222 VNC display, you must use the ``-k`` parameter to set the keyboard 2223 layout if you are not using en-us. Valid syntax for the display is 2224 2225 ``to=L`` 2226 With this option, QEMU will try next available VNC displays, 2227 until the number L, if the origianlly defined "-vnc display" is 2228 not available, e.g. port 5900+display is already used by another 2229 application. By default, to=0. 2230 2231 ``host:d`` 2232 TCP connections will only be allowed from host on display d. By 2233 convention the TCP port is 5900+d. Optionally, host can be 2234 omitted in which case the server will accept connections from 2235 any host. 2236 2237 ``unix:path`` 2238 Connections will be allowed over UNIX domain sockets where path 2239 is the location of a unix socket to listen for connections on. 2240 2241 ``none`` 2242 VNC is initialized but not started. The monitor ``change`` 2243 command can be used to later start the VNC server. 2244 2245 Following the display value there may be one or more option flags 2246 separated by commas. Valid options are 2247 2248 ``reverse=on|off`` 2249 Connect to a listening VNC client via a "reverse" connection. 2250 The client is specified by the display. For reverse network 2251 connections (host:d,``reverse``), the d argument is a TCP port 2252 number, not a display number. 2253 2254 ``websocket=on|off`` 2255 Opens an additional TCP listening port dedicated to VNC 2256 Websocket connections. If a bare websocket option is given, the 2257 Websocket port is 5700+display. An alternative port can be 2258 specified with the syntax ``websocket``\ =port. 2259 2260 If host is specified connections will only be allowed from this 2261 host. It is possible to control the websocket listen address 2262 independently, using the syntax ``websocket``\ =host:port. 2263 2264 If no TLS credentials are provided, the websocket connection 2265 runs in unencrypted mode. If TLS credentials are provided, the 2266 websocket connection requires encrypted client connections. 2267 2268 ``password=on|off`` 2269 Require that password based authentication is used for client 2270 connections. 2271 2272 The password must be set separately using the ``set_password`` 2273 command in the :ref:`QEMU monitor`. The 2274 syntax to change your password is: 2275 ``set_password <protocol> <password>`` where <protocol> could be 2276 either "vnc" or "spice". 2277 2278 If you would like to change <protocol> password expiration, you 2279 should use ``expire_password <protocol> <expiration-time>`` 2280 where expiration time could be one of the following options: 2281 now, never, +seconds or UNIX time of expiration, e.g. +60 to 2282 make password expire in 60 seconds, or 1335196800 to make 2283 password expire on "Mon Apr 23 12:00:00 EDT 2012" (UNIX time for 2284 this date and time). 2285 2286 You can also use keywords "now" or "never" for the expiration 2287 time to allow <protocol> password to expire immediately or never 2288 expire. 2289 2290 ``password-secret=<secret-id>`` 2291 Require that password based authentication is used for client 2292 connections, using the password provided by the ``secret`` 2293 object identified by ``secret-id``. 2294 2295 ``tls-creds=ID`` 2296 Provides the ID of a set of TLS credentials to use to secure the 2297 VNC server. They will apply to both the normal VNC server socket 2298 and the websocket socket (if enabled). Setting TLS credentials 2299 will cause the VNC server socket to enable the VeNCrypt auth 2300 mechanism. The credentials should have been previously created 2301 using the ``-object tls-creds`` argument. 2302 2303 ``tls-authz=ID`` 2304 Provides the ID of the QAuthZ authorization object against which 2305 the client's x509 distinguished name will validated. This object 2306 is only resolved at time of use, so can be deleted and recreated 2307 on the fly while the VNC server is active. If missing, it will 2308 default to denying access. 2309 2310 ``sasl=on|off`` 2311 Require that the client use SASL to authenticate with the VNC 2312 server. The exact choice of authentication method used is 2313 controlled from the system / user's SASL configuration file for 2314 the 'qemu' service. This is typically found in 2315 /etc/sasl2/qemu.conf. If running QEMU as an unprivileged user, 2316 an environment variable SASL\_CONF\_PATH can be used to make it 2317 search alternate locations for the service config. While some 2318 SASL auth methods can also provide data encryption (eg GSSAPI), 2319 it is recommended that SASL always be combined with the 'tls' 2320 and 'x509' settings to enable use of SSL and server 2321 certificates. This ensures a data encryption preventing 2322 compromise of authentication credentials. See the 2323 :ref:`VNC security` section in the System Emulation Users Guide 2324 for details on using SASL authentication. 2325 2326 ``sasl-authz=ID`` 2327 Provides the ID of the QAuthZ authorization object against which 2328 the client's SASL username will validated. This object is only 2329 resolved at time of use, so can be deleted and recreated on the 2330 fly while the VNC server is active. If missing, it will default 2331 to denying access. 2332 2333 ``acl=on|off`` 2334 Legacy method for enabling authorization of clients against the 2335 x509 distinguished name and SASL username. It results in the 2336 creation of two ``authz-list`` objects with IDs of 2337 ``vnc.username`` and ``vnc.x509dname``. The rules for these 2338 objects must be configured with the HMP ACL commands. 2339 2340 This option is deprecated and should no longer be used. The new 2341 ``sasl-authz`` and ``tls-authz`` options are a replacement. 2342 2343 ``lossy=on|off`` 2344 Enable lossy compression methods (gradient, JPEG, ...). If this 2345 option is set, VNC client may receive lossy framebuffer updates 2346 depending on its encoding settings. Enabling this option can 2347 save a lot of bandwidth at the expense of quality. 2348 2349 ``non-adaptive=on|off`` 2350 Disable adaptive encodings. Adaptive encodings are enabled by 2351 default. An adaptive encoding will try to detect frequently 2352 updated screen regions, and send updates in these regions using 2353 a lossy encoding (like JPEG). This can be really helpful to save 2354 bandwidth when playing videos. Disabling adaptive encodings 2355 restores the original static behavior of encodings like Tight. 2356 2357 ``share=[allow-exclusive|force-shared|ignore]`` 2358 Set display sharing policy. 'allow-exclusive' allows clients to 2359 ask for exclusive access. As suggested by the rfb spec this is 2360 implemented by dropping other connections. Connecting multiple 2361 clients in parallel requires all clients asking for a shared 2362 session (vncviewer: -shared switch). This is the default. 2363 'force-shared' disables exclusive client access. Useful for 2364 shared desktop sessions, where you don't want someone forgetting 2365 specify -shared disconnect everybody else. 'ignore' completely 2366 ignores the shared flag and allows everybody connect 2367 unconditionally. Doesn't conform to the rfb spec but is 2368 traditional QEMU behavior. 2369 2370 ``key-delay-ms`` 2371 Set keyboard delay, for key down and key up events, in 2372 milliseconds. Default is 10. Keyboards are low-bandwidth 2373 devices, so this slowdown can help the device and guest to keep 2374 up and not lose events in case events are arriving in bulk. 2375 Possible causes for the latter are flaky network connections, or 2376 scripts for automated testing. 2377 2378 ``audiodev=audiodev`` 2379 Use the specified audiodev when the VNC client requests audio 2380 transmission. When not using an -audiodev argument, this option 2381 must be omitted, otherwise is must be present and specify a 2382 valid audiodev. 2383 2384 ``power-control=on|off`` 2385 Permit the remote client to issue shutdown, reboot or reset power 2386 control requests. 2387ERST 2388 2389ARCHHEADING(, QEMU_ARCH_I386) 2390 2391ARCHHEADING(i386 target only:, QEMU_ARCH_I386) 2392 2393DEF("win2k-hack", 0, QEMU_OPTION_win2k_hack, 2394 "-win2k-hack use it when installing Windows 2000 to avoid a disk full bug\n", 2395 QEMU_ARCH_I386) 2396SRST 2397``-win2k-hack`` 2398 Use it when installing Windows 2000 to avoid a disk full bug. After 2399 Windows 2000 is installed, you no longer need this option (this 2400 option slows down the IDE transfers). 2401ERST 2402 2403DEF("no-fd-bootchk", 0, QEMU_OPTION_no_fd_bootchk, 2404 "-no-fd-bootchk disable boot signature checking for floppy disks\n", 2405 QEMU_ARCH_I386) 2406SRST 2407``-no-fd-bootchk`` 2408 Disable boot signature checking for floppy disks in BIOS. May be 2409 needed to boot from old floppy disks. 2410ERST 2411 2412DEF("no-acpi", 0, QEMU_OPTION_no_acpi, 2413 "-no-acpi disable ACPI\n", QEMU_ARCH_I386 | QEMU_ARCH_ARM) 2414SRST 2415``-no-acpi`` 2416 Disable ACPI (Advanced Configuration and Power Interface) support. 2417 Use it if your guest OS complains about ACPI problems (PC target 2418 machine only). 2419ERST 2420 2421DEF("no-hpet", 0, QEMU_OPTION_no_hpet, 2422 "-no-hpet disable HPET\n", QEMU_ARCH_I386) 2423SRST 2424``-no-hpet`` 2425 Disable HPET support. 2426ERST 2427 2428DEF("acpitable", HAS_ARG, QEMU_OPTION_acpitable, 2429 "-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" 2430 " ACPI table description\n", QEMU_ARCH_I386) 2431SRST 2432``-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]...]`` 2433 Add ACPI table with specified header fields and context from 2434 specified files. For file=, take whole ACPI table from the specified 2435 files, including all ACPI headers (possible overridden by other 2436 options). For data=, only data portion of the table is used, all 2437 header information is specified in the command line. If a SLIC table 2438 is supplied to QEMU, then the SLIC's oem\_id and oem\_table\_id 2439 fields will override the same in the RSDT and the FADT (a.k.a. 2440 FACP), in order to ensure the field matches required by the 2441 Microsoft SLIC spec and the ACPI spec. 2442ERST 2443 2444DEF("smbios", HAS_ARG, QEMU_OPTION_smbios, 2445 "-smbios file=binary\n" 2446 " load SMBIOS entry from binary file\n" 2447 "-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d]\n" 2448 " [,uefi=on|off]\n" 2449 " specify SMBIOS type 0 fields\n" 2450 "-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str]\n" 2451 " [,uuid=uuid][,sku=str][,family=str]\n" 2452 " specify SMBIOS type 1 fields\n" 2453 "-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str]\n" 2454 " [,asset=str][,location=str]\n" 2455 " specify SMBIOS type 2 fields\n" 2456 "-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str]\n" 2457 " [,sku=str]\n" 2458 " specify SMBIOS type 3 fields\n" 2459 "-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str]\n" 2460 " [,asset=str][,part=str][,max-speed=%d][,current-speed=%d]\n" 2461 " specify SMBIOS type 4 fields\n" 2462 "-smbios type=11[,value=str][,path=filename]\n" 2463 " specify SMBIOS type 11 fields\n" 2464 "-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str]\n" 2465 " [,asset=str][,part=str][,speed=%d]\n" 2466 " specify SMBIOS type 17 fields\n" 2467 "-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]\n" 2468 " specify SMBIOS type 41 fields\n", 2469 QEMU_ARCH_I386 | QEMU_ARCH_ARM) 2470SRST 2471``-smbios file=binary`` 2472 Load SMBIOS entry from binary file. 2473 2474``-smbios type=0[,vendor=str][,version=str][,date=str][,release=%d.%d][,uefi=on|off]`` 2475 Specify SMBIOS type 0 fields 2476 2477``-smbios type=1[,manufacturer=str][,product=str][,version=str][,serial=str][,uuid=uuid][,sku=str][,family=str]`` 2478 Specify SMBIOS type 1 fields 2479 2480``-smbios type=2[,manufacturer=str][,product=str][,version=str][,serial=str][,asset=str][,location=str]`` 2481 Specify SMBIOS type 2 fields 2482 2483``-smbios type=3[,manufacturer=str][,version=str][,serial=str][,asset=str][,sku=str]`` 2484 Specify SMBIOS type 3 fields 2485 2486``-smbios type=4[,sock_pfx=str][,manufacturer=str][,version=str][,serial=str][,asset=str][,part=str]`` 2487 Specify SMBIOS type 4 fields 2488 2489``-smbios type=11[,value=str][,path=filename]`` 2490 Specify SMBIOS type 11 fields 2491 2492 This argument can be repeated multiple times, and values are added in the order they are parsed. 2493 Applications intending to use OEM strings data are encouraged to use their application name as 2494 a prefix for the value string. This facilitates passing information for multiple applications 2495 concurrently. 2496 2497 The ``value=str`` syntax provides the string data inline, while the ``path=filename`` syntax 2498 loads data from a file on disk. Note that the file is not permitted to contain any NUL bytes. 2499 2500 Both the ``value`` and ``path`` options can be repeated multiple times and will be added to 2501 the SMBIOS table in the order in which they appear. 2502 2503 Note that on the x86 architecture, the total size of all SMBIOS tables is limited to 65535 2504 bytes. Thus the OEM strings data is not suitable for passing large amounts of data into the 2505 guest. Instead it should be used as a indicator to inform the guest where to locate the real 2506 data set, for example, by specifying the serial ID of a block device. 2507 2508 An example passing three strings is 2509 2510 .. parsed-literal:: 2511 2512 -smbios type=11,value=cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/,\\ 2513 value=anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os,\\ 2514 path=/some/file/with/oemstringsdata.txt 2515 2516 In the guest OS this is visible with the ``dmidecode`` command 2517 2518 .. parsed-literal:: 2519 2520 $ dmidecode -t 11 2521 Handle 0x0E00, DMI type 11, 5 bytes 2522 OEM Strings 2523 String 1: cloud-init:ds=nocloud-net;s=http://10.10.0.1:8000/ 2524 String 2: anaconda:method=http://dl.fedoraproject.org/pub/fedora/linux/releases/25/x86_64/os 2525 String 3: myapp:some extra data 2526 2527 2528``-smbios type=17[,loc_pfx=str][,bank=str][,manufacturer=str][,serial=str][,asset=str][,part=str][,speed=%d]`` 2529 Specify SMBIOS type 17 fields 2530 2531``-smbios type=41[,designation=str][,kind=str][,instance=%d][,pcidev=str]`` 2532 Specify SMBIOS type 41 fields 2533 2534 This argument can be repeated multiple times. Its main use is to allow network interfaces be created 2535 as ``enoX`` on Linux, with X being the instance number, instead of the name depending on the interface 2536 position on the PCI bus. 2537 2538 Here is an example of use: 2539 2540 .. parsed-literal:: 2541 2542 -netdev user,id=internet \\ 2543 -device virtio-net-pci,mac=50:54:00:00:00:42,netdev=internet,id=internet-dev \\ 2544 -smbios type=41,designation='Onboard LAN',instance=1,kind=ethernet,pcidev=internet-dev 2545 2546 In the guest OS, the device should then appear as ``eno1``: 2547 2548 ..parsed-literal:: 2549 2550 $ ip -brief l 2551 lo UNKNOWN 00:00:00:00:00:00 <LOOPBACK,UP,LOWER_UP> 2552 eno1 UP 50:54:00:00:00:42 <BROADCAST,MULTICAST,UP,LOWER_UP> 2553 2554 Currently, the PCI device has to be attached to the root bus. 2555 2556ERST 2557 2558DEFHEADING() 2559 2560DEFHEADING(Network options:) 2561 2562DEF("netdev", HAS_ARG, QEMU_OPTION_netdev, 2563#ifdef CONFIG_SLIRP 2564 "-netdev user,id=str[,ipv4=on|off][,net=addr[/mask]][,host=addr]\n" 2565 " [,ipv6=on|off][,ipv6-net=addr[/int]][,ipv6-host=addr]\n" 2566 " [,restrict=on|off][,hostname=host][,dhcpstart=addr]\n" 2567 " [,dns=addr][,ipv6-dns=addr][,dnssearch=domain][,domainname=domain]\n" 2568 " [,tftp=dir][,tftp-server-name=name][,bootfile=f][,hostfwd=rule][,guestfwd=rule]" 2569#ifndef _WIN32 2570 "[,smb=dir[,smbserver=addr]]\n" 2571#endif 2572 " configure a user mode network backend with ID 'str',\n" 2573 " its DHCP server and optional services\n" 2574#endif 2575#ifdef _WIN32 2576 "-netdev tap,id=str,ifname=name\n" 2577 " configure a host TAP network backend with ID 'str'\n" 2578#else 2579 "-netdev tap,id=str[,fd=h][,fds=x:y:...:z][,ifname=name][,script=file][,downscript=dfile]\n" 2580 " [,br=bridge][,helper=helper][,sndbuf=nbytes][,vnet_hdr=on|off][,vhost=on|off]\n" 2581 " [,vhostfd=h][,vhostfds=x:y:...:z][,vhostforce=on|off][,queues=n]\n" 2582 " [,poll-us=n]\n" 2583 " configure a host TAP network backend with ID 'str'\n" 2584 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n" 2585 " use network scripts 'file' (default=" DEFAULT_NETWORK_SCRIPT ")\n" 2586 " to configure it and 'dfile' (default=" DEFAULT_NETWORK_DOWN_SCRIPT ")\n" 2587 " to deconfigure it\n" 2588 " use '[down]script=no' to disable script execution\n" 2589 " use network helper 'helper' (default=" DEFAULT_BRIDGE_HELPER ") to\n" 2590 " configure it\n" 2591 " use 'fd=h' to connect to an already opened TAP interface\n" 2592 " use 'fds=x:y:...:z' to connect to already opened multiqueue capable TAP interfaces\n" 2593 " use 'sndbuf=nbytes' to limit the size of the send buffer (the\n" 2594 " default is disabled 'sndbuf=0' to enable flow control set 'sndbuf=1048576')\n" 2595 " use vnet_hdr=off to avoid enabling the IFF_VNET_HDR tap flag\n" 2596 " use vnet_hdr=on to make the lack of IFF_VNET_HDR support an error condition\n" 2597 " use vhost=on to enable experimental in kernel accelerator\n" 2598 " (only has effect for virtio guests which use MSIX)\n" 2599 " use vhostforce=on to force vhost on for non-MSIX virtio guests\n" 2600 " use 'vhostfd=h' to connect to an already opened vhost net device\n" 2601 " use 'vhostfds=x:y:...:z to connect to multiple already opened vhost net devices\n" 2602 " use 'queues=n' to specify the number of queues to be created for multiqueue TAP\n" 2603 " use 'poll-us=n' to specify the maximum number of microseconds that could be\n" 2604 " spent on busy polling for vhost net\n" 2605 "-netdev bridge,id=str[,br=bridge][,helper=helper]\n" 2606 " configure a host TAP network backend with ID 'str' that is\n" 2607 " connected to a bridge (default=" DEFAULT_BRIDGE_INTERFACE ")\n" 2608 " using the program 'helper (default=" DEFAULT_BRIDGE_HELPER ")\n" 2609#endif 2610#ifdef __linux__ 2611 "-netdev l2tpv3,id=str,src=srcaddr,dst=dstaddr[,srcport=srcport][,dstport=dstport]\n" 2612 " [,rxsession=rxsession],txsession=txsession[,ipv6=on|off][,udp=on|off]\n" 2613 " [,cookie64=on|off][,counter][,pincounter][,txcookie=txcookie]\n" 2614 " [,rxcookie=rxcookie][,offset=offset]\n" 2615 " configure a network backend with ID 'str' connected to\n" 2616 " an Ethernet over L2TPv3 pseudowire.\n" 2617 " Linux kernel 3.3+ as well as most routers can talk\n" 2618 " L2TPv3. This transport allows connecting a VM to a VM,\n" 2619 " VM to a router and even VM to Host. It is a nearly-universal\n" 2620 " standard (RFC3931). Note - this implementation uses static\n" 2621 " pre-configured tunnels (same as the Linux kernel).\n" 2622 " use 'src=' to specify source address\n" 2623 " use 'dst=' to specify destination address\n" 2624 " use 'udp=on' to specify udp encapsulation\n" 2625 " use 'srcport=' to specify source udp port\n" 2626 " use 'dstport=' to specify destination udp port\n" 2627 " use 'ipv6=on' to force v6\n" 2628 " L2TPv3 uses cookies to prevent misconfiguration as\n" 2629 " well as a weak security measure\n" 2630 " use 'rxcookie=0x012345678' to specify a rxcookie\n" 2631 " use 'txcookie=0x012345678' to specify a txcookie\n" 2632 " use 'cookie64=on' to set cookie size to 64 bit, otherwise 32\n" 2633 " use 'counter=off' to force a 'cut-down' L2TPv3 with no counter\n" 2634 " use 'pincounter=on' to work around broken counter handling in peer\n" 2635 " use 'offset=X' to add an extra offset between header and data\n" 2636#endif 2637 "-netdev socket,id=str[,fd=h][,listen=[host]:port][,connect=host:port]\n" 2638 " configure a network backend to connect to another network\n" 2639 " using a socket connection\n" 2640 "-netdev socket,id=str[,fd=h][,mcast=maddr:port[,localaddr=addr]]\n" 2641 " configure a network backend to connect to a multicast maddr and port\n" 2642 " use 'localaddr=addr' to specify the host address to send packets from\n" 2643 "-netdev socket,id=str[,fd=h][,udp=host:port][,localaddr=host:port]\n" 2644 " configure a network backend to connect to another network\n" 2645 " using an UDP tunnel\n" 2646#ifdef CONFIG_VDE 2647 "-netdev vde,id=str[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]\n" 2648 " configure a network backend to connect to port 'n' of a vde switch\n" 2649 " running on host and listening for incoming connections on 'socketpath'.\n" 2650 " Use group 'groupname' and mode 'octalmode' to change default\n" 2651 " ownership and permissions for communication port.\n" 2652#endif 2653#ifdef CONFIG_NETMAP 2654 "-netdev netmap,id=str,ifname=name[,devname=nmname]\n" 2655 " attach to the existing netmap-enabled network interface 'name', or to a\n" 2656 " VALE port (created on the fly) called 'name' ('nmname' is name of the \n" 2657 " netmap device, defaults to '/dev/netmap')\n" 2658#endif 2659#ifdef CONFIG_POSIX 2660 "-netdev vhost-user,id=str,chardev=dev[,vhostforce=on|off]\n" 2661 " configure a vhost-user network, backed by a chardev 'dev'\n" 2662#endif 2663#ifdef __linux__ 2664 "-netdev vhost-vdpa,id=str,vhostdev=/path/to/dev\n" 2665 " configure a vhost-vdpa network,Establish a vhost-vdpa netdev\n" 2666#endif 2667 "-netdev hubport,id=str,hubid=n[,netdev=nd]\n" 2668 " configure a hub port on the hub with ID 'n'\n", QEMU_ARCH_ALL) 2669DEF("nic", HAS_ARG, QEMU_OPTION_nic, 2670 "-nic [tap|bridge|" 2671#ifdef CONFIG_SLIRP 2672 "user|" 2673#endif 2674#ifdef __linux__ 2675 "l2tpv3|" 2676#endif 2677#ifdef CONFIG_VDE 2678 "vde|" 2679#endif 2680#ifdef CONFIG_NETMAP 2681 "netmap|" 2682#endif 2683#ifdef CONFIG_POSIX 2684 "vhost-user|" 2685#endif 2686 "socket][,option][,...][mac=macaddr]\n" 2687 " initialize an on-board / default host NIC (using MAC address\n" 2688 " macaddr) and connect it to the given host network backend\n" 2689 "-nic none use it alone to have zero network devices (the default is to\n" 2690 " provided a 'user' network connection)\n", 2691 QEMU_ARCH_ALL) 2692DEF("net", HAS_ARG, QEMU_OPTION_net, 2693 "-net nic[,macaddr=mac][,model=type][,name=str][,addr=str][,vectors=v]\n" 2694 " configure or create an on-board (or machine default) NIC and\n" 2695 " connect it to hub 0 (please use -nic unless you need a hub)\n" 2696 "-net [" 2697#ifdef CONFIG_SLIRP 2698 "user|" 2699#endif 2700 "tap|" 2701 "bridge|" 2702#ifdef CONFIG_VDE 2703 "vde|" 2704#endif 2705#ifdef CONFIG_NETMAP 2706 "netmap|" 2707#endif 2708 "socket][,option][,option][,...]\n" 2709 " old way to initialize a host network interface\n" 2710 " (use the -netdev option if possible instead)\n", QEMU_ARCH_ALL) 2711SRST 2712``-nic [tap|bridge|user|l2tpv3|vde|netmap|vhost-user|socket][,...][,mac=macaddr][,model=mn]`` 2713 This option is a shortcut for configuring both the on-board 2714 (default) guest NIC hardware and the host network backend in one go. 2715 The host backend options are the same as with the corresponding 2716 ``-netdev`` options below. The guest NIC model can be set with 2717 ``model=modelname``. Use ``model=help`` to list the available device 2718 types. The hardware MAC address can be set with ``mac=macaddr``. 2719 2720 The following two example do exactly the same, to show how ``-nic`` 2721 can be used to shorten the command line length: 2722 2723 .. parsed-literal:: 2724 2725 |qemu_system| -netdev user,id=n1,ipv6=off -device e1000,netdev=n1,mac=52:54:98:76:54:32 2726 |qemu_system| -nic user,ipv6=off,model=e1000,mac=52:54:98:76:54:32 2727 2728``-nic none`` 2729 Indicate that no network devices should be configured. It is used to 2730 override the default configuration (default NIC with "user" host 2731 network backend) which is activated if no other networking options 2732 are provided. 2733 2734``-netdev user,id=id[,option][,option][,...]`` 2735 Configure user mode host network backend which requires no 2736 administrator privilege to run. Valid options are: 2737 2738 ``id=id`` 2739 Assign symbolic name for use in monitor commands. 2740 2741 ``ipv4=on|off and ipv6=on|off`` 2742 Specify that either IPv4 or IPv6 must be enabled. If neither is 2743 specified both protocols are enabled. 2744 2745 ``net=addr[/mask]`` 2746 Set IP network address the guest will see. Optionally specify 2747 the netmask, either in the form a.b.c.d or as number of valid 2748 top-most bits. Default is 10.0.2.0/24. 2749 2750 ``host=addr`` 2751 Specify the guest-visible address of the host. Default is the 2752 2nd IP in the guest network, i.e. x.x.x.2. 2753 2754 ``ipv6-net=addr[/int]`` 2755 Set IPv6 network address the guest will see (default is 2756 fec0::/64). The network prefix is given in the usual hexadecimal 2757 IPv6 address notation. The prefix size is optional, and is given 2758 as the number of valid top-most bits (default is 64). 2759 2760 ``ipv6-host=addr`` 2761 Specify the guest-visible IPv6 address of the host. Default is 2762 the 2nd IPv6 in the guest network, i.e. xxxx::2. 2763 2764 ``restrict=on|off`` 2765 If this option is enabled, the guest will be isolated, i.e. it 2766 will not be able to contact the host and no guest IP packets 2767 will be routed over the host to the outside. This option does 2768 not affect any explicitly set forwarding rules. 2769 2770 ``hostname=name`` 2771 Specifies the client hostname reported by the built-in DHCP 2772 server. 2773 2774 ``dhcpstart=addr`` 2775 Specify the first of the 16 IPs the built-in DHCP server can 2776 assign. Default is the 15th to 31st IP in the guest network, 2777 i.e. x.x.x.15 to x.x.x.31. 2778 2779 ``dns=addr`` 2780 Specify the guest-visible address of the virtual nameserver. The 2781 address must be different from the host address. Default is the 2782 3rd IP in the guest network, i.e. x.x.x.3. 2783 2784 ``ipv6-dns=addr`` 2785 Specify the guest-visible address of the IPv6 virtual 2786 nameserver. The address must be different from the host address. 2787 Default is the 3rd IP in the guest network, i.e. xxxx::3. 2788 2789 ``dnssearch=domain`` 2790 Provides an entry for the domain-search list sent by the 2791 built-in DHCP server. More than one domain suffix can be 2792 transmitted by specifying this option multiple times. If 2793 supported, this will cause the guest to automatically try to 2794 append the given domain suffix(es) in case a domain name can not 2795 be resolved. 2796 2797 Example: 2798 2799 .. parsed-literal:: 2800 2801 |qemu_system| -nic user,dnssearch=mgmt.example.org,dnssearch=example.org 2802 2803 ``domainname=domain`` 2804 Specifies the client domain name reported by the built-in DHCP 2805 server. 2806 2807 ``tftp=dir`` 2808 When using the user mode network stack, activate a built-in TFTP 2809 server. The files in dir will be exposed as the root of a TFTP 2810 server. The TFTP client on the guest must be configured in 2811 binary mode (use the command ``bin`` of the Unix TFTP client). 2812 2813 ``tftp-server-name=name`` 2814 In BOOTP reply, broadcast name as the "TFTP server name" 2815 (RFC2132 option 66). This can be used to advise the guest to 2816 load boot files or configurations from a different server than 2817 the host address. 2818 2819 ``bootfile=file`` 2820 When using the user mode network stack, broadcast file as the 2821 BOOTP filename. In conjunction with ``tftp``, this can be used 2822 to network boot a guest from a local directory. 2823 2824 Example (using pxelinux): 2825 2826 .. parsed-literal:: 2827 2828 |qemu_system| -hda linux.img -boot n -device e1000,netdev=n1 \\ 2829 -netdev user,id=n1,tftp=/path/to/tftp/files,bootfile=/pxelinux.0 2830 2831 ``smb=dir[,smbserver=addr]`` 2832 When using the user mode network stack, activate a built-in SMB 2833 server so that Windows OSes can access to the host files in 2834 ``dir`` transparently. The IP address of the SMB server can be 2835 set to addr. By default the 4th IP in the guest network is used, 2836 i.e. x.x.x.4. 2837 2838 In the guest Windows OS, the line: 2839 2840 :: 2841 2842 10.0.2.4 smbserver 2843 2844 must be added in the file ``C:\WINDOWS\LMHOSTS`` (for windows 2845 9x/Me) or ``C:\WINNT\SYSTEM32\DRIVERS\ETC\LMHOSTS`` (Windows 2846 NT/2000). 2847 2848 Then ``dir`` can be accessed in ``\\smbserver\qemu``. 2849 2850 Note that a SAMBA server must be installed on the host OS. 2851 2852 ``hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport`` 2853 Redirect incoming TCP or UDP connections to the host port 2854 hostport to the guest IP address guestaddr on guest port 2855 guestport. If guestaddr is not specified, its value is x.x.x.15 2856 (default first address given by the built-in DHCP server). By 2857 specifying hostaddr, the rule can be bound to a specific host 2858 interface. If no connection type is set, TCP is used. This 2859 option can be given multiple times. 2860 2861 For example, to redirect host X11 connection from screen 1 to 2862 guest screen 0, use the following: 2863 2864 .. parsed-literal:: 2865 2866 # on the host 2867 |qemu_system| -nic user,hostfwd=tcp:127.0.0.1:6001-:6000 2868 # this host xterm should open in the guest X11 server 2869 xterm -display :1 2870 2871 To redirect telnet connections from host port 5555 to telnet 2872 port on the guest, use the following: 2873 2874 .. parsed-literal:: 2875 2876 # on the host 2877 |qemu_system| -nic user,hostfwd=tcp::5555-:23 2878 telnet localhost 5555 2879 2880 Then when you use on the host ``telnet localhost 5555``, you 2881 connect to the guest telnet server. 2882 2883 ``guestfwd=[tcp]:server:port-dev``; \ ``guestfwd=[tcp]:server:port-cmd:command`` 2884 Forward guest TCP connections to the IP address server on port 2885 port to the character device dev or to a program executed by 2886 cmd:command which gets spawned for each connection. This option 2887 can be given multiple times. 2888 2889 You can either use a chardev directly and have that one used 2890 throughout QEMU's lifetime, like in the following example: 2891 2892 .. parsed-literal:: 2893 2894 # open 10.10.1.1:4321 on bootup, connect 10.0.2.100:1234 to it whenever 2895 # the guest accesses it 2896 |qemu_system| -nic user,guestfwd=tcp:10.0.2.100:1234-tcp:10.10.1.1:4321 2897 2898 Or you can execute a command on every TCP connection established 2899 by the guest, so that QEMU behaves similar to an inetd process 2900 for that virtual server: 2901 2902 .. parsed-literal:: 2903 2904 # call "netcat 10.10.1.1 4321" on every TCP connection to 10.0.2.100:1234 2905 # and connect the TCP stream to its stdin/stdout 2906 |qemu_system| -nic 'user,id=n1,guestfwd=tcp:10.0.2.100:1234-cmd:netcat 10.10.1.1 4321' 2907 2908``-netdev tap,id=id[,fd=h][,ifname=name][,script=file][,downscript=dfile][,br=bridge][,helper=helper]`` 2909 Configure a host TAP network backend with ID id. 2910 2911 Use the network script file to configure it and the network script 2912 dfile to deconfigure it. If name is not provided, the OS 2913 automatically provides one. The default network configure script is 2914 ``/etc/qemu-ifup`` and the default network deconfigure script is 2915 ``/etc/qemu-ifdown``. Use ``script=no`` or ``downscript=no`` to 2916 disable script execution. 2917 2918 If running QEMU as an unprivileged user, use the network helper 2919 to configure the TAP interface and attach it to the bridge. 2920 The default network helper executable is 2921 ``/path/to/qemu-bridge-helper`` and the default bridge device is 2922 ``br0``. 2923 2924 ``fd``\ =h can be used to specify the handle of an already opened 2925 host TAP interface. 2926 2927 Examples: 2928 2929 .. parsed-literal:: 2930 2931 #launch a QEMU instance with the default network script 2932 |qemu_system| linux.img -nic tap 2933 2934 .. parsed-literal:: 2935 2936 #launch a QEMU instance with two NICs, each one connected 2937 #to a TAP device 2938 |qemu_system| linux.img \\ 2939 -netdev tap,id=nd0,ifname=tap0 -device e1000,netdev=nd0 \\ 2940 -netdev tap,id=nd1,ifname=tap1 -device rtl8139,netdev=nd1 2941 2942 .. parsed-literal:: 2943 2944 #launch a QEMU instance with the default network helper to 2945 #connect a TAP device to bridge br0 2946 |qemu_system| linux.img -device virtio-net-pci,netdev=n1 \\ 2947 -netdev tap,id=n1,"helper=/path/to/qemu-bridge-helper" 2948 2949``-netdev bridge,id=id[,br=bridge][,helper=helper]`` 2950 Connect a host TAP network interface to a host bridge device. 2951 2952 Use the network helper helper to configure the TAP interface and 2953 attach it to the bridge. The default network helper executable is 2954 ``/path/to/qemu-bridge-helper`` and the default bridge device is 2955 ``br0``. 2956 2957 Examples: 2958 2959 .. parsed-literal:: 2960 2961 #launch a QEMU instance with the default network helper to 2962 #connect a TAP device to bridge br0 2963 |qemu_system| linux.img -netdev bridge,id=n1 -device virtio-net,netdev=n1 2964 2965 .. parsed-literal:: 2966 2967 #launch a QEMU instance with the default network helper to 2968 #connect a TAP device to bridge qemubr0 2969 |qemu_system| linux.img -netdev bridge,br=qemubr0,id=n1 -device virtio-net,netdev=n1 2970 2971``-netdev socket,id=id[,fd=h][,listen=[host]:port][,connect=host:port]`` 2972 This host network backend can be used to connect the guest's network 2973 to another QEMU virtual machine using a TCP socket connection. If 2974 ``listen`` is specified, QEMU waits for incoming connections on port 2975 (host is optional). ``connect`` is used to connect to another QEMU 2976 instance using the ``listen`` option. ``fd``\ =h specifies an 2977 already opened TCP socket. 2978 2979 Example: 2980 2981 .. parsed-literal:: 2982 2983 # launch a first QEMU instance 2984 |qemu_system| linux.img \\ 2985 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\ 2986 -netdev socket,id=n1,listen=:1234 2987 # connect the network of this instance to the network of the first instance 2988 |qemu_system| linux.img \\ 2989 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\ 2990 -netdev socket,id=n2,connect=127.0.0.1:1234 2991 2992``-netdev socket,id=id[,fd=h][,mcast=maddr:port[,localaddr=addr]]`` 2993 Configure a socket host network backend to share the guest's network 2994 traffic with another QEMU virtual machines using a UDP multicast 2995 socket, effectively making a bus for every QEMU with same multicast 2996 address maddr and port. NOTES: 2997 2998 1. Several QEMU can be running on different hosts and share same bus 2999 (assuming correct multicast setup for these hosts). 3000 3001 2. mcast support is compatible with User Mode Linux (argument 3002 ``ethN=mcast``), see http://user-mode-linux.sf.net. 3003 3004 3. Use ``fd=h`` to specify an already opened UDP multicast socket. 3005 3006 Example: 3007 3008 .. parsed-literal:: 3009 3010 # launch one QEMU instance 3011 |qemu_system| linux.img \\ 3012 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\ 3013 -netdev socket,id=n1,mcast=230.0.0.1:1234 3014 # launch another QEMU instance on same "bus" 3015 |qemu_system| linux.img \\ 3016 -device e1000,netdev=n2,mac=52:54:00:12:34:57 \\ 3017 -netdev socket,id=n2,mcast=230.0.0.1:1234 3018 # launch yet another QEMU instance on same "bus" 3019 |qemu_system| linux.img \\ 3020 -device e1000,netdev=n3,mac=52:54:00:12:34:58 \\ 3021 -netdev socket,id=n3,mcast=230.0.0.1:1234 3022 3023 Example (User Mode Linux compat.): 3024 3025 .. parsed-literal:: 3026 3027 # launch QEMU instance (note mcast address selected is UML's default) 3028 |qemu_system| linux.img \\ 3029 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\ 3030 -netdev socket,id=n1,mcast=239.192.168.1:1102 3031 # launch UML 3032 /path/to/linux ubd0=/path/to/root_fs eth0=mcast 3033 3034 Example (send packets from host's 1.2.3.4): 3035 3036 .. parsed-literal:: 3037 3038 |qemu_system| linux.img \\ 3039 -device e1000,netdev=n1,mac=52:54:00:12:34:56 \\ 3040 -netdev socket,id=n1,mcast=239.192.168.1:1102,localaddr=1.2.3.4 3041 3042``-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]`` 3043 Configure a L2TPv3 pseudowire host network backend. L2TPv3 (RFC3931) 3044 is a popular protocol to transport Ethernet (and other Layer 2) data 3045 frames between two systems. It is present in routers, firewalls and 3046 the Linux kernel (from version 3.3 onwards). 3047 3048 This transport allows a VM to communicate to another VM, router or 3049 firewall directly. 3050 3051 ``src=srcaddr`` 3052 source address (mandatory) 3053 3054 ``dst=dstaddr`` 3055 destination address (mandatory) 3056 3057 ``udp`` 3058 select udp encapsulation (default is ip). 3059 3060 ``srcport=srcport`` 3061 source udp port. 3062 3063 ``dstport=dstport`` 3064 destination udp port. 3065 3066 ``ipv6`` 3067 force v6, otherwise defaults to v4. 3068 3069 ``rxcookie=rxcookie``; \ ``txcookie=txcookie`` 3070 Cookies are a weak form of security in the l2tpv3 specification. 3071 Their function is mostly to prevent misconfiguration. By default 3072 they are 32 bit. 3073 3074 ``cookie64`` 3075 Set cookie size to 64 bit instead of the default 32 3076 3077 ``counter=off`` 3078 Force a 'cut-down' L2TPv3 with no counter as in 3079 draft-mkonstan-l2tpext-keyed-ipv6-tunnel-00 3080 3081 ``pincounter=on`` 3082 Work around broken counter handling in peer. This may also help 3083 on networks which have packet reorder. 3084 3085 ``offset=offset`` 3086 Add an extra offset between header and data 3087 3088 For example, to attach a VM running on host 4.3.2.1 via L2TPv3 to 3089 the bridge br-lan on the remote Linux host 1.2.3.4: 3090 3091 .. parsed-literal:: 3092 3093 # Setup tunnel on linux host using raw ip as encapsulation 3094 # on 1.2.3.4 3095 ip l2tp add tunnel remote 4.3.2.1 local 1.2.3.4 tunnel_id 1 peer_tunnel_id 1 \\ 3096 encap udp udp_sport 16384 udp_dport 16384 3097 ip l2tp add session tunnel_id 1 name vmtunnel0 session_id \\ 3098 0xFFFFFFFF peer_session_id 0xFFFFFFFF 3099 ifconfig vmtunnel0 mtu 1500 3100 ifconfig vmtunnel0 up 3101 brctl addif br-lan vmtunnel0 3102 3103 3104 # on 4.3.2.1 3105 # launch QEMU instance - if your network has reorder or is very lossy add ,pincounter 3106 3107 |qemu_system| linux.img -device e1000,netdev=n1 \\ 3108 -netdev l2tpv3,id=n1,src=4.2.3.1,dst=1.2.3.4,udp,srcport=16384,dstport=16384,rxsession=0xffffffff,txsession=0xffffffff,counter 3109 3110``-netdev vde,id=id[,sock=socketpath][,port=n][,group=groupname][,mode=octalmode]`` 3111 Configure VDE backend to connect to PORT n of a vde switch running 3112 on host and listening for incoming connections on socketpath. Use 3113 GROUP groupname and MODE octalmode to change default ownership and 3114 permissions for communication port. This option is only available if 3115 QEMU has been compiled with vde support enabled. 3116 3117 Example: 3118 3119 .. parsed-literal:: 3120 3121 # launch vde switch 3122 vde_switch -F -sock /tmp/myswitch 3123 # launch QEMU instance 3124 |qemu_system| linux.img -nic vde,sock=/tmp/myswitch 3125 3126``-netdev vhost-user,chardev=id[,vhostforce=on|off][,queues=n]`` 3127 Establish a vhost-user netdev, backed by a chardev id. The chardev 3128 should be a unix domain socket backed one. The vhost-user uses a 3129 specifically defined protocol to pass vhost ioctl replacement 3130 messages to an application on the other end of the socket. On 3131 non-MSIX guests, the feature can be forced with vhostforce. Use 3132 'queues=n' to specify the number of queues to be created for 3133 multiqueue vhost-user. 3134 3135 Example: 3136 3137 :: 3138 3139 qemu -m 512 -object memory-backend-file,id=mem,size=512M,mem-path=/hugetlbfs,share=on \ 3140 -numa node,memdev=mem \ 3141 -chardev socket,id=chr0,path=/path/to/socket \ 3142 -netdev type=vhost-user,id=net0,chardev=chr0 \ 3143 -device virtio-net-pci,netdev=net0 3144 3145``-netdev vhost-vdpa,vhostdev=/path/to/dev`` 3146 Establish a vhost-vdpa netdev. 3147 3148 vDPA device is a device that uses a datapath which complies with 3149 the virtio specifications with a vendor specific control path. 3150 vDPA devices can be both physically located on the hardware or 3151 emulated by software. 3152 3153``-netdev hubport,id=id,hubid=hubid[,netdev=nd]`` 3154 Create a hub port on the emulated hub with ID hubid. 3155 3156 The hubport netdev lets you connect a NIC to a QEMU emulated hub 3157 instead of a single netdev. Alternatively, you can also connect the 3158 hubport to another netdev with ID nd by using the ``netdev=nd`` 3159 option. 3160 3161``-net nic[,netdev=nd][,macaddr=mac][,model=type] [,name=name][,addr=addr][,vectors=v]`` 3162 Legacy option to configure or create an on-board (or machine 3163 default) Network Interface Card(NIC) and connect it either to the 3164 emulated hub with ID 0 (i.e. the default hub), or to the netdev nd. 3165 If model is omitted, then the default NIC model associated with the 3166 machine type is used. Note that the default NIC model may change in 3167 future QEMU releases, so it is highly recommended to always specify 3168 a model. Optionally, the MAC address can be changed to mac, the 3169 device address set to addr (PCI cards only), and a name can be 3170 assigned for use in monitor commands. Optionally, for PCI cards, you 3171 can specify the number v of MSI-X vectors that the card should have; 3172 this option currently only affects virtio cards; set v = 0 to 3173 disable MSI-X. If no ``-net`` option is specified, a single NIC is 3174 created. QEMU can emulate several different models of network card. 3175 Use ``-net nic,model=help`` for a list of available devices for your 3176 target. 3177 3178``-net user|tap|bridge|socket|l2tpv3|vde[,...][,name=name]`` 3179 Configure a host network backend (with the options corresponding to 3180 the same ``-netdev`` option) and connect it to the emulated hub 0 3181 (the default hub). Use name to specify the name of the hub port. 3182ERST 3183 3184DEFHEADING() 3185 3186DEFHEADING(Character device options:) 3187 3188DEF("chardev", HAS_ARG, QEMU_OPTION_chardev, 3189 "-chardev help\n" 3190 "-chardev null,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3191 "-chardev socket,id=id[,host=host],port=port[,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off][,reconnect=seconds]\n" 3192 " [,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds][,mux=on|off]\n" 3193 " [,logfile=PATH][,logappend=on|off][,tls-creds=ID][,tls-authz=ID] (tcp)\n" 3194 "-chardev socket,id=id,path=path[,server=on|off][,wait=on|off][,telnet=on|off][,websocket=on|off][,reconnect=seconds]\n" 3195 " [,mux=on|off][,logfile=PATH][,logappend=on|off][,abstract=on|off][,tight=on|off] (unix)\n" 3196 "-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr]\n" 3197 " [,localport=localport][,ipv4=on|off][,ipv6=on|off][,mux=on|off]\n" 3198 " [,logfile=PATH][,logappend=on|off]\n" 3199 "-chardev msmouse,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3200 "-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]\n" 3201 " [,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3202 "-chardev ringbuf,id=id[,size=size][,logfile=PATH][,logappend=on|off]\n" 3203 "-chardev file,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3204 "-chardev pipe,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3205#ifdef _WIN32 3206 "-chardev console,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3207 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3208#else 3209 "-chardev pty,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3210 "-chardev stdio,id=id[,mux=on|off][,signal=on|off][,logfile=PATH][,logappend=on|off]\n" 3211#endif 3212#ifdef CONFIG_BRLAPI 3213 "-chardev braille,id=id[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3214#endif 3215#if defined(__linux__) || defined(__sun__) || defined(__FreeBSD__) \ 3216 || defined(__NetBSD__) || defined(__OpenBSD__) || defined(__DragonFly__) 3217 "-chardev serial,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3218 "-chardev tty,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3219#endif 3220#if defined(__linux__) || defined(__FreeBSD__) || defined(__DragonFly__) 3221 "-chardev parallel,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3222 "-chardev parport,id=id,path=path[,mux=on|off][,logfile=PATH][,logappend=on|off]\n" 3223#endif 3224#if defined(CONFIG_SPICE) 3225 "-chardev spicevmc,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n" 3226 "-chardev spiceport,id=id,name=name[,debug=debug][,logfile=PATH][,logappend=on|off]\n" 3227#endif 3228 , QEMU_ARCH_ALL 3229) 3230 3231SRST 3232The general form of a character device option is: 3233 3234``-chardev backend,id=id[,mux=on|off][,options]`` 3235 Backend is one of: ``null``, ``socket``, ``udp``, ``msmouse``, 3236 ``vc``, ``ringbuf``, ``file``, ``pipe``, ``console``, ``serial``, 3237 ``pty``, ``stdio``, ``braille``, ``tty``, ``parallel``, ``parport``, 3238 ``spicevmc``, ``spiceport``. The specific backend will determine the 3239 applicable options. 3240 3241 Use ``-chardev help`` to print all available chardev backend types. 3242 3243 All devices must have an id, which can be any string up to 127 3244 characters long. It is used to uniquely identify this device in 3245 other command line directives. 3246 3247 A character device may be used in multiplexing mode by multiple 3248 front-ends. Specify ``mux=on`` to enable this mode. A multiplexer is 3249 a "1:N" device, and here the "1" end is your specified chardev 3250 backend, and the "N" end is the various parts of QEMU that can talk 3251 to a chardev. If you create a chardev with ``id=myid`` and 3252 ``mux=on``, QEMU will create a multiplexer with your specified ID, 3253 and you can then configure multiple front ends to use that chardev 3254 ID for their input/output. Up to four different front ends can be 3255 connected to a single multiplexed chardev. (Without multiplexing 3256 enabled, a chardev can only be used by a single front end.) For 3257 instance you could use this to allow a single stdio chardev to be 3258 used by two serial ports and the QEMU monitor: 3259 3260 :: 3261 3262 -chardev stdio,mux=on,id=char0 \ 3263 -mon chardev=char0,mode=readline \ 3264 -serial chardev:char0 \ 3265 -serial chardev:char0 3266 3267 You can have more than one multiplexer in a system configuration; 3268 for instance you could have a TCP port multiplexed between UART 0 3269 and UART 1, and stdio multiplexed between the QEMU monitor and a 3270 parallel port: 3271 3272 :: 3273 3274 -chardev stdio,mux=on,id=char0 \ 3275 -mon chardev=char0,mode=readline \ 3276 -parallel chardev:char0 \ 3277 -chardev tcp,...,mux=on,id=char1 \ 3278 -serial chardev:char1 \ 3279 -serial chardev:char1 3280 3281 When you're using a multiplexed character device, some escape 3282 sequences are interpreted in the input. See the chapter about 3283 :ref:`keys in the character backend multiplexer` in the 3284 System Emulation Users Guide for more details. 3285 3286 Note that some other command line options may implicitly create 3287 multiplexed character backends; for instance ``-serial mon:stdio`` 3288 creates a multiplexed stdio backend connected to the serial port and 3289 the QEMU monitor, and ``-nographic`` also multiplexes the console 3290 and the monitor to stdio. 3291 3292 There is currently no support for multiplexing in the other 3293 direction (where a single QEMU front end takes input and output from 3294 multiple chardevs). 3295 3296 Every backend supports the ``logfile`` option, which supplies the 3297 path to a file to record all data transmitted via the backend. The 3298 ``logappend`` option controls whether the log file will be truncated 3299 or appended to when opened. 3300 3301The available backends are: 3302 3303``-chardev null,id=id`` 3304 A void device. This device will not emit any data, and will drop any 3305 data it receives. The null backend does not take any options. 3306 3307``-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]`` 3308 Create a two-way stream socket, which can be either a TCP or a unix 3309 socket. A unix socket will be created if ``path`` is specified. 3310 Behaviour is undefined if TCP options are specified for a unix 3311 socket. 3312 3313 ``server=on|off`` specifies that the socket shall be a listening socket. 3314 3315 ``wait=on|off`` specifies that QEMU should not block waiting for a client 3316 to connect to a listening socket. 3317 3318 ``telnet=on|off`` specifies that traffic on the socket should interpret 3319 telnet escape sequences. 3320 3321 ``websocket=on|off`` specifies that the socket uses WebSocket protocol for 3322 communication. 3323 3324 ``reconnect`` sets the timeout for reconnecting on non-server 3325 sockets when the remote end goes away. qemu will delay this many 3326 seconds and then attempt to reconnect. Zero disables reconnecting, 3327 and is the default. 3328 3329 ``tls-creds`` requests enablement of the TLS protocol for 3330 encryption, and specifies the id of the TLS credentials to use for 3331 the handshake. The credentials must be previously created with the 3332 ``-object tls-creds`` argument. 3333 3334 ``tls-auth`` provides the ID of the QAuthZ authorization object 3335 against which the client's x509 distinguished name will be 3336 validated. This object is only resolved at time of use, so can be 3337 deleted and recreated on the fly while the chardev server is active. 3338 If missing, it will default to denying access. 3339 3340 TCP and unix socket options are given below: 3341 3342 ``TCP options: port=port[,host=host][,to=to][,ipv4=on|off][,ipv6=on|off][,nodelay=on|off]`` 3343 ``host`` for a listening socket specifies the local address to 3344 be bound. For a connecting socket species the remote host to 3345 connect to. ``host`` is optional for listening sockets. If not 3346 specified it defaults to ``0.0.0.0``. 3347 3348 ``port`` for a listening socket specifies the local port to be 3349 bound. For a connecting socket specifies the port on the remote 3350 host to connect to. ``port`` can be given as either a port 3351 number or a service name. ``port`` is required. 3352 3353 ``to`` is only relevant to listening sockets. If it is 3354 specified, and ``port`` cannot be bound, QEMU will attempt to 3355 bind to subsequent ports up to and including ``to`` until it 3356 succeeds. ``to`` must be specified as a port number. 3357 3358 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4 3359 or IPv6 must be used. If neither is specified the socket may 3360 use either protocol. 3361 3362 ``nodelay=on|off`` disables the Nagle algorithm. 3363 3364 ``unix options: path=path[,abstract=on|off][,tight=on|off]`` 3365 ``path`` specifies the local path of the unix socket. ``path`` 3366 is required. 3367 ``abstract=on|off`` specifies the use of the abstract socket namespace, 3368 rather than the filesystem. Optional, defaults to false. 3369 ``tight=on|off`` sets the socket length of abstract sockets to their minimum, 3370 rather than the full sun_path length. Optional, defaults to true. 3371 3372``-chardev udp,id=id[,host=host],port=port[,localaddr=localaddr][,localport=localport][,ipv4=on|off][,ipv6=on|off]`` 3373 Sends all traffic from the guest to a remote host over UDP. 3374 3375 ``host`` specifies the remote host to connect to. If not specified 3376 it defaults to ``localhost``. 3377 3378 ``port`` specifies the port on the remote host to connect to. 3379 ``port`` is required. 3380 3381 ``localaddr`` specifies the local address to bind to. If not 3382 specified it defaults to ``0.0.0.0``. 3383 3384 ``localport`` specifies the local port to bind to. If not specified 3385 any available local port will be used. 3386 3387 ``ipv4=on|off`` and ``ipv6=on|off`` specify that either IPv4 or IPv6 must be used. 3388 If neither is specified the device may use either protocol. 3389 3390``-chardev msmouse,id=id`` 3391 Forward QEMU's emulated msmouse events to the guest. ``msmouse`` 3392 does not take any options. 3393 3394``-chardev vc,id=id[[,width=width][,height=height]][[,cols=cols][,rows=rows]]`` 3395 Connect to a QEMU text console. ``vc`` may optionally be given a 3396 specific size. 3397 3398 ``width`` and ``height`` specify the width and height respectively 3399 of the console, in pixels. 3400 3401 ``cols`` and ``rows`` specify that the console be sized to fit a 3402 text console with the given dimensions. 3403 3404``-chardev ringbuf,id=id[,size=size]`` 3405 Create a ring buffer with fixed size ``size``. size must be a power 3406 of two and defaults to ``64K``. 3407 3408``-chardev file,id=id,path=path`` 3409 Log all traffic received from the guest to a file. 3410 3411 ``path`` specifies the path of the file to be opened. This file will 3412 be created if it does not already exist, and overwritten if it does. 3413 ``path`` is required. 3414 3415``-chardev pipe,id=id,path=path`` 3416 Create a two-way connection to the guest. The behaviour differs 3417 slightly between Windows hosts and other hosts: 3418 3419 On Windows, a single duplex pipe will be created at 3420 ``\\.pipe\path``. 3421 3422 On other hosts, 2 pipes will be created called ``path.in`` and 3423 ``path.out``. Data written to ``path.in`` will be received by the 3424 guest. Data written by the guest can be read from ``path.out``. QEMU 3425 will not create these fifos, and requires them to be present. 3426 3427 ``path`` forms part of the pipe path as described above. ``path`` is 3428 required. 3429 3430``-chardev console,id=id`` 3431 Send traffic from the guest to QEMU's standard output. ``console`` 3432 does not take any options. 3433 3434 ``console`` is only available on Windows hosts. 3435 3436``-chardev serial,id=id,path=path`` 3437 Send traffic from the guest to a serial device on the host. 3438 3439 On Unix hosts serial will actually accept any tty device, not only 3440 serial lines. 3441 3442 ``path`` specifies the name of the serial device to open. 3443 3444``-chardev pty,id=id`` 3445 Create a new pseudo-terminal on the host and connect to it. ``pty`` 3446 does not take any options. 3447 3448 ``pty`` is not available on Windows hosts. 3449 3450``-chardev stdio,id=id[,signal=on|off]`` 3451 Connect to standard input and standard output of the QEMU process. 3452 3453 ``signal`` controls if signals are enabled on the terminal, that 3454 includes exiting QEMU with the key sequence Control-c. This option 3455 is enabled by default, use ``signal=off`` to disable it. 3456 3457``-chardev braille,id=id`` 3458 Connect to a local BrlAPI server. ``braille`` does not take any 3459 options. 3460 3461``-chardev tty,id=id,path=path`` 3462 ``tty`` is only available on Linux, Sun, FreeBSD, NetBSD, OpenBSD 3463 and DragonFlyBSD hosts. It is an alias for ``serial``. 3464 3465 ``path`` specifies the path to the tty. ``path`` is required. 3466 3467``-chardev parallel,id=id,path=path`` 3468 \ 3469``-chardev parport,id=id,path=path`` 3470 ``parallel`` is only available on Linux, FreeBSD and DragonFlyBSD 3471 hosts. 3472 3473 Connect to a local parallel port. 3474 3475 ``path`` specifies the path to the parallel port device. ``path`` is 3476 required. 3477 3478``-chardev spicevmc,id=id,debug=debug,name=name`` 3479 ``spicevmc`` is only available when spice support is built in. 3480 3481 ``debug`` debug level for spicevmc 3482 3483 ``name`` name of spice channel to connect to 3484 3485 Connect to a spice virtual machine channel, such as vdiport. 3486 3487``-chardev spiceport,id=id,debug=debug,name=name`` 3488 ``spiceport`` is only available when spice support is built in. 3489 3490 ``debug`` debug level for spicevmc 3491 3492 ``name`` name of spice port to connect to 3493 3494 Connect to a spice port, allowing a Spice client to handle the 3495 traffic identified by a name (preferably a fqdn). 3496ERST 3497 3498DEFHEADING() 3499 3500#ifdef CONFIG_TPM 3501DEFHEADING(TPM device options:) 3502 3503DEF("tpmdev", HAS_ARG, QEMU_OPTION_tpmdev, \ 3504 "-tpmdev passthrough,id=id[,path=path][,cancel-path=path]\n" 3505 " use path to provide path to a character device; default is /dev/tpm0\n" 3506 " use cancel-path to provide path to TPM's cancel sysfs entry; if\n" 3507 " not provided it will be searched for in /sys/class/misc/tpm?/device\n" 3508 "-tpmdev emulator,id=id,chardev=dev\n" 3509 " configure the TPM device using chardev backend\n", 3510 QEMU_ARCH_ALL) 3511SRST 3512The general form of a TPM device option is: 3513 3514``-tpmdev backend,id=id[,options]`` 3515 The specific backend type will determine the applicable options. The 3516 ``-tpmdev`` option creates the TPM backend and requires a 3517 ``-device`` option that specifies the TPM frontend interface model. 3518 3519 Use ``-tpmdev help`` to print all available TPM backend types. 3520 3521The available backends are: 3522 3523``-tpmdev passthrough,id=id,path=path,cancel-path=cancel-path`` 3524 (Linux-host only) Enable access to the host's TPM using the 3525 passthrough driver. 3526 3527 ``path`` specifies the path to the host's TPM device, i.e., on a 3528 Linux host this would be ``/dev/tpm0``. ``path`` is optional and by 3529 default ``/dev/tpm0`` is used. 3530 3531 ``cancel-path`` specifies the path to the host TPM device's sysfs 3532 entry allowing for cancellation of an ongoing TPM command. 3533 ``cancel-path`` is optional and by default QEMU will search for the 3534 sysfs entry to use. 3535 3536 Some notes about using the host's TPM with the passthrough driver: 3537 3538 The TPM device accessed by the passthrough driver must not be used 3539 by any other application on the host. 3540 3541 Since the host's firmware (BIOS/UEFI) has already initialized the 3542 TPM, the VM's firmware (BIOS/UEFI) will not be able to initialize 3543 the TPM again and may therefore not show a TPM-specific menu that 3544 would otherwise allow the user to configure the TPM, e.g., allow the 3545 user to enable/disable or activate/deactivate the TPM. Further, if 3546 TPM ownership is released from within a VM then the host's TPM will 3547 get disabled and deactivated. To enable and activate the TPM again 3548 afterwards, the host has to be rebooted and the user is required to 3549 enter the firmware's menu to enable and activate the TPM. If the TPM 3550 is left disabled and/or deactivated most TPM commands will fail. 3551 3552 To create a passthrough TPM use the following two options: 3553 3554 :: 3555 3556 -tpmdev passthrough,id=tpm0 -device tpm-tis,tpmdev=tpm0 3557 3558 Note that the ``-tpmdev`` id is ``tpm0`` and is referenced by 3559 ``tpmdev=tpm0`` in the device option. 3560 3561``-tpmdev emulator,id=id,chardev=dev`` 3562 (Linux-host only) Enable access to a TPM emulator using Unix domain 3563 socket based chardev backend. 3564 3565 ``chardev`` specifies the unique ID of a character device backend 3566 that provides connection to the software TPM server. 3567 3568 To create a TPM emulator backend device with chardev socket backend: 3569 3570 :: 3571 3572 -chardev socket,id=chrtpm,path=/tmp/swtpm-sock -tpmdev emulator,id=tpm0,chardev=chrtpm -device tpm-tis,tpmdev=tpm0 3573ERST 3574 3575DEFHEADING() 3576 3577#endif 3578 3579DEFHEADING(Linux/Multiboot boot specific:) 3580SRST 3581When using these options, you can use a given Linux or Multiboot kernel 3582without installing it in the disk image. It can be useful for easier 3583testing of various kernels. 3584 3585 3586ERST 3587 3588DEF("kernel", HAS_ARG, QEMU_OPTION_kernel, \ 3589 "-kernel bzImage use 'bzImage' as kernel image\n", QEMU_ARCH_ALL) 3590SRST 3591``-kernel bzImage`` 3592 Use bzImage as kernel image. The kernel can be either a Linux kernel 3593 or in multiboot format. 3594ERST 3595 3596DEF("append", HAS_ARG, QEMU_OPTION_append, \ 3597 "-append cmdline use 'cmdline' as kernel command line\n", QEMU_ARCH_ALL) 3598SRST 3599``-append cmdline`` 3600 Use cmdline as kernel command line 3601ERST 3602 3603DEF("initrd", HAS_ARG, QEMU_OPTION_initrd, \ 3604 "-initrd file use 'file' as initial ram disk\n", QEMU_ARCH_ALL) 3605SRST 3606``-initrd file`` 3607 Use file as initial ram disk. 3608 3609``-initrd "file1 arg=foo,file2"`` 3610 This syntax is only available with multiboot. 3611 3612 Use file1 and file2 as modules and pass arg=foo as parameter to the 3613 first module. 3614ERST 3615 3616DEF("dtb", HAS_ARG, QEMU_OPTION_dtb, \ 3617 "-dtb file use 'file' as device tree image\n", QEMU_ARCH_ALL) 3618SRST 3619``-dtb file`` 3620 Use file as a device tree binary (dtb) image and pass it to the 3621 kernel on boot. 3622ERST 3623 3624DEFHEADING() 3625 3626DEFHEADING(Debug/Expert options:) 3627 3628DEF("compat", HAS_ARG, QEMU_OPTION_compat, 3629 "-compat [deprecated-input=accept|reject|crash][,deprecated-output=accept|hide]\n" 3630 " Policy for handling deprecated management interfaces\n", 3631 QEMU_ARCH_ALL) 3632SRST 3633``-compat [deprecated-input=@var{input-policy}][,deprecated-output=@var{output-policy}]`` 3634 Set policy for handling deprecated management interfaces (experimental): 3635 3636 ``deprecated-input=accept`` (default) 3637 Accept deprecated commands and arguments 3638 ``deprecated-input=reject`` 3639 Reject deprecated commands and arguments 3640 ``deprecated-input=crash`` 3641 Crash on deprecated commands and arguments 3642 ``deprecated-output=accept`` (default) 3643 Emit deprecated command results and events 3644 ``deprecated-output=hide`` 3645 Suppress deprecated command results and events 3646 3647 Limitation: covers only syntactic aspects of QMP. 3648ERST 3649 3650DEF("fw_cfg", HAS_ARG, QEMU_OPTION_fwcfg, 3651 "-fw_cfg [name=]<name>,file=<file>\n" 3652 " add named fw_cfg entry with contents from file\n" 3653 "-fw_cfg [name=]<name>,string=<str>\n" 3654 " add named fw_cfg entry with contents from string\n", 3655 QEMU_ARCH_ALL) 3656SRST 3657``-fw_cfg [name=]name,file=file`` 3658 Add named fw\_cfg entry with contents from file file. 3659 3660``-fw_cfg [name=]name,string=str`` 3661 Add named fw\_cfg entry with contents from string str. 3662 3663 The terminating NUL character of the contents of str will not be 3664 included as part of the fw\_cfg item data. To insert contents with 3665 embedded NUL characters, you have to use the file parameter. 3666 3667 The fw\_cfg entries are passed by QEMU through to the guest. 3668 3669 Example: 3670 3671 :: 3672 3673 -fw_cfg name=opt/com.mycompany/blob,file=./my_blob.bin 3674 3675 creates an fw\_cfg entry named opt/com.mycompany/blob with contents 3676 from ./my\_blob.bin. 3677ERST 3678 3679DEF("serial", HAS_ARG, QEMU_OPTION_serial, \ 3680 "-serial dev redirect the serial port to char device 'dev'\n", 3681 QEMU_ARCH_ALL) 3682SRST 3683``-serial dev`` 3684 Redirect the virtual serial port to host character device dev. The 3685 default device is ``vc`` in graphical mode and ``stdio`` in non 3686 graphical mode. 3687 3688 This option can be used several times to simulate up to 4 serial 3689 ports. 3690 3691 Use ``-serial none`` to disable all serial ports. 3692 3693 Available character devices are: 3694 3695 ``vc[:WxH]`` 3696 Virtual console. Optionally, a width and height can be given in 3697 pixel with 3698 3699 :: 3700 3701 vc:800x600 3702 3703 It is also possible to specify width or height in characters: 3704 3705 :: 3706 3707 vc:80Cx24C 3708 3709 ``pty`` 3710 [Linux only] Pseudo TTY (a new PTY is automatically allocated) 3711 3712 ``none`` 3713 No device is allocated. 3714 3715 ``null`` 3716 void device 3717 3718 ``chardev:id`` 3719 Use a named character device defined with the ``-chardev`` 3720 option. 3721 3722 ``/dev/XXX`` 3723 [Linux only] Use host tty, e.g. ``/dev/ttyS0``. The host serial 3724 port parameters are set according to the emulated ones. 3725 3726 ``/dev/parportN`` 3727 [Linux only, parallel port only] Use host parallel port N. 3728 Currently SPP and EPP parallel port features can be used. 3729 3730 ``file:filename`` 3731 Write output to filename. No character can be read. 3732 3733 ``stdio`` 3734 [Unix only] standard input/output 3735 3736 ``pipe:filename`` 3737 name pipe filename 3738 3739 ``COMn`` 3740 [Windows only] Use host serial port n 3741 3742 ``udp:[remote_host]:remote_port[@[src_ip]:src_port]`` 3743 This implements UDP Net Console. When remote\_host or src\_ip 3744 are not specified they default to ``0.0.0.0``. When not using a 3745 specified src\_port a random port is automatically chosen. 3746 3747 If you just want a simple readonly console you can use 3748 ``netcat`` or ``nc``, by starting QEMU with: 3749 ``-serial udp::4555`` and nc as: ``nc -u -l -p 4555``. Any time 3750 QEMU writes something to that port it will appear in the 3751 netconsole session. 3752 3753 If you plan to send characters back via netconsole or you want 3754 to stop and start QEMU a lot of times, you should have QEMU use 3755 the same source port each time by using something like ``-serial 3756 udp::4555@:4556`` to QEMU. Another approach is to use a patched 3757 version of netcat which can listen to a TCP port and send and 3758 receive characters via udp. If you have a patched version of 3759 netcat which activates telnet remote echo and single char 3760 transfer, then you can use the following options to set up a 3761 netcat redirector to allow telnet on port 5555 to access the 3762 QEMU port. 3763 3764 ``QEMU Options:`` 3765 -serial udp::4555@:4556 3766 3767 ``netcat options:`` 3768 -u -P 4555 -L 0.0.0.0:4556 -t -p 5555 -I -T 3769 3770 ``telnet options:`` 3771 localhost 5555 3772 3773 ``tcp:[host]:port[,server=on|off][,wait=on|off][,nodelay=on|off][,reconnect=seconds]`` 3774 The TCP Net Console has two modes of operation. It can send the 3775 serial I/O to a location or wait for a connection from a 3776 location. By default the TCP Net Console is sent to host at the 3777 port. If you use the ``server=on`` option QEMU will wait for a client 3778 socket application to connect to the port before continuing, 3779 unless the ``wait=on|off`` option was specified. The ``nodelay=on|off`` 3780 option disables the Nagle buffering algorithm. The ``reconnect=on`` 3781 option only applies if ``server=no`` is set, if the connection goes 3782 down it will attempt to reconnect at the given interval. If host 3783 is omitted, 0.0.0.0 is assumed. Only one TCP connection at a 3784 time is accepted. You can use ``telnet=on`` to connect to the 3785 corresponding character device. 3786 3787 ``Example to send tcp console to 192.168.0.2 port 4444`` 3788 -serial tcp:192.168.0.2:4444 3789 3790 ``Example to listen and wait on port 4444 for connection`` 3791 -serial tcp::4444,server=on 3792 3793 ``Example to not wait and listen on ip 192.168.0.100 port 4444`` 3794 -serial tcp:192.168.0.100:4444,server=on,wait=off 3795 3796 ``telnet:host:port[,server=on|off][,wait=on|off][,nodelay=on|off]`` 3797 The telnet protocol is used instead of raw tcp sockets. The 3798 options work the same as if you had specified ``-serial tcp``. 3799 The difference is that the port acts like a telnet server or 3800 client using telnet option negotiation. This will also allow you 3801 to send the MAGIC\_SYSRQ sequence if you use a telnet that 3802 supports sending the break sequence. Typically in unix telnet 3803 you do it with Control-] and then type "send break" followed by 3804 pressing the enter key. 3805 3806 ``websocket:host:port,server=on[,wait=on|off][,nodelay=on|off]`` 3807 The WebSocket protocol is used instead of raw tcp socket. The 3808 port acts as a WebSocket server. Client mode is not supported. 3809 3810 ``unix:path[,server=on|off][,wait=on|off][,reconnect=seconds]`` 3811 A unix domain socket is used instead of a tcp socket. The option 3812 works the same as if you had specified ``-serial tcp`` except 3813 the unix domain socket path is used for connections. 3814 3815 ``mon:dev_string`` 3816 This is a special option to allow the monitor to be multiplexed 3817 onto another serial port. The monitor is accessed with key 3818 sequence of Control-a and then pressing c. dev\_string should be 3819 any one of the serial devices specified above. An example to 3820 multiplex the monitor onto a telnet server listening on port 3821 4444 would be: 3822 3823 ``-serial mon:telnet::4444,server=on,wait=off`` 3824 3825 When the monitor is multiplexed to stdio in this way, Ctrl+C 3826 will not terminate QEMU any more but will be passed to the guest 3827 instead. 3828 3829 ``braille`` 3830 Braille device. This will use BrlAPI to display the braille 3831 output on a real or fake device. 3832 3833 ``msmouse`` 3834 Three button serial mouse. Configure the guest to use Microsoft 3835 protocol. 3836ERST 3837 3838DEF("parallel", HAS_ARG, QEMU_OPTION_parallel, \ 3839 "-parallel dev redirect the parallel port to char device 'dev'\n", 3840 QEMU_ARCH_ALL) 3841SRST 3842``-parallel dev`` 3843 Redirect the virtual parallel port to host device dev (same devices 3844 as the serial port). On Linux hosts, ``/dev/parportN`` can be used 3845 to use hardware devices connected on the corresponding host parallel 3846 port. 3847 3848 This option can be used several times to simulate up to 3 parallel 3849 ports. 3850 3851 Use ``-parallel none`` to disable all parallel ports. 3852ERST 3853 3854DEF("monitor", HAS_ARG, QEMU_OPTION_monitor, \ 3855 "-monitor dev redirect the monitor to char device 'dev'\n", 3856 QEMU_ARCH_ALL) 3857SRST 3858``-monitor dev`` 3859 Redirect the monitor to host device dev (same devices as the serial 3860 port). The default device is ``vc`` in graphical mode and ``stdio`` 3861 in non graphical mode. Use ``-monitor none`` to disable the default 3862 monitor. 3863ERST 3864DEF("qmp", HAS_ARG, QEMU_OPTION_qmp, \ 3865 "-qmp dev like -monitor but opens in 'control' mode\n", 3866 QEMU_ARCH_ALL) 3867SRST 3868``-qmp dev`` 3869 Like -monitor but opens in 'control' mode. 3870ERST 3871DEF("qmp-pretty", HAS_ARG, QEMU_OPTION_qmp_pretty, \ 3872 "-qmp-pretty dev like -qmp but uses pretty JSON formatting\n", 3873 QEMU_ARCH_ALL) 3874SRST 3875``-qmp-pretty dev`` 3876 Like -qmp but uses pretty JSON formatting. 3877ERST 3878 3879DEF("mon", HAS_ARG, QEMU_OPTION_mon, \ 3880 "-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]\n", QEMU_ARCH_ALL) 3881SRST 3882``-mon [chardev=]name[,mode=readline|control][,pretty[=on|off]]`` 3883 Setup monitor on chardev name. ``mode=control`` configures 3884 a QMP monitor (a JSON RPC-style protocol) and it is not the 3885 same as HMP, the human monitor that has a "(qemu)" prompt. 3886 ``pretty`` is only valid when ``mode=control``, 3887 turning on JSON pretty printing to ease 3888 human reading and debugging. 3889ERST 3890 3891DEF("debugcon", HAS_ARG, QEMU_OPTION_debugcon, \ 3892 "-debugcon dev redirect the debug console to char device 'dev'\n", 3893 QEMU_ARCH_ALL) 3894SRST 3895``-debugcon dev`` 3896 Redirect the debug console to host device dev (same devices as the 3897 serial port). The debug console is an I/O port which is typically 3898 port 0xe9; writing to that I/O port sends output to this device. The 3899 default device is ``vc`` in graphical mode and ``stdio`` in non 3900 graphical mode. 3901ERST 3902 3903DEF("pidfile", HAS_ARG, QEMU_OPTION_pidfile, \ 3904 "-pidfile file write PID to 'file'\n", QEMU_ARCH_ALL) 3905SRST 3906``-pidfile file`` 3907 Store the QEMU process PID in file. It is useful if you launch QEMU 3908 from a script. 3909ERST 3910 3911DEF("singlestep", 0, QEMU_OPTION_singlestep, \ 3912 "-singlestep always run in singlestep mode\n", QEMU_ARCH_ALL) 3913SRST 3914``-singlestep`` 3915 Run the emulation in single step mode. 3916ERST 3917 3918DEF("preconfig", 0, QEMU_OPTION_preconfig, \ 3919 "--preconfig pause QEMU before machine is initialized (experimental)\n", 3920 QEMU_ARCH_ALL) 3921SRST 3922``--preconfig`` 3923 Pause QEMU for interactive configuration before the machine is 3924 created, which allows querying and configuring properties that will 3925 affect machine initialization. Use QMP command 'x-exit-preconfig' to 3926 exit the preconfig state and move to the next state (i.e. run guest 3927 if -S isn't used or pause the second time if -S is used). This 3928 option is experimental. 3929ERST 3930 3931DEF("S", 0, QEMU_OPTION_S, \ 3932 "-S freeze CPU at startup (use 'c' to start execution)\n", 3933 QEMU_ARCH_ALL) 3934SRST 3935``-S`` 3936 Do not start CPU at startup (you must type 'c' in the monitor). 3937ERST 3938 3939DEF("overcommit", HAS_ARG, QEMU_OPTION_overcommit, 3940 "-overcommit [mem-lock=on|off][cpu-pm=on|off]\n" 3941 " run qemu with overcommit hints\n" 3942 " mem-lock=on|off controls memory lock support (default: off)\n" 3943 " cpu-pm=on|off controls cpu power management (default: off)\n", 3944 QEMU_ARCH_ALL) 3945SRST 3946``-overcommit mem-lock=on|off`` 3947 \ 3948``-overcommit cpu-pm=on|off`` 3949 Run qemu with hints about host resource overcommit. The default is 3950 to assume that host overcommits all resources. 3951 3952 Locking qemu and guest memory can be enabled via ``mem-lock=on`` 3953 (disabled by default). This works when host memory is not 3954 overcommitted and reduces the worst-case latency for guest. 3955 3956 Guest ability to manage power state of host cpus (increasing latency 3957 for other processes on the same host cpu, but decreasing latency for 3958 guest) can be enabled via ``cpu-pm=on`` (disabled by default). This 3959 works best when host CPU is not overcommitted. When used, host 3960 estimates of CPU cycle and power utilization will be incorrect, not 3961 taking into account guest idle time. 3962ERST 3963 3964DEF("gdb", HAS_ARG, QEMU_OPTION_gdb, \ 3965 "-gdb dev accept gdb connection on 'dev'. (QEMU defaults to starting\n" 3966 " the guest without waiting for gdb to connect; use -S too\n" 3967 " if you want it to not start execution.)\n", 3968 QEMU_ARCH_ALL) 3969SRST 3970``-gdb dev`` 3971 Accept a gdb connection on device dev (see the :ref:`GDB usage` chapter 3972 in the System Emulation Users Guide). Note that this option does not pause QEMU 3973 execution -- if you want QEMU to not start the guest until you 3974 connect with gdb and issue a ``continue`` command, you will need to 3975 also pass the ``-S`` option to QEMU. 3976 3977 The most usual configuration is to listen on a local TCP socket:: 3978 3979 -gdb tcp::3117 3980 3981 but you can specify other backends; UDP, pseudo TTY, or even stdio 3982 are all reasonable use cases. For example, a stdio connection 3983 allows you to start QEMU from within gdb and establish the 3984 connection via a pipe: 3985 3986 .. parsed-literal:: 3987 3988 (gdb) target remote | exec |qemu_system| -gdb stdio ... 3989ERST 3990 3991DEF("s", 0, QEMU_OPTION_s, \ 3992 "-s shorthand for -gdb tcp::" DEFAULT_GDBSTUB_PORT "\n", 3993 QEMU_ARCH_ALL) 3994SRST 3995``-s`` 3996 Shorthand for -gdb tcp::1234, i.e. open a gdbserver on TCP port 1234 3997 (see the :ref:`GDB usage` chapter in the System Emulation Users Guide). 3998ERST 3999 4000DEF("d", HAS_ARG, QEMU_OPTION_d, \ 4001 "-d item1,... enable logging of specified items (use '-d help' for a list of log items)\n", 4002 QEMU_ARCH_ALL) 4003SRST 4004``-d item1[,...]`` 4005 Enable logging of specified items. Use '-d help' for a list of log 4006 items. 4007ERST 4008 4009DEF("D", HAS_ARG, QEMU_OPTION_D, \ 4010 "-D logfile output log to logfile (default stderr)\n", 4011 QEMU_ARCH_ALL) 4012SRST 4013``-D logfile`` 4014 Output log in logfile instead of to stderr 4015ERST 4016 4017DEF("dfilter", HAS_ARG, QEMU_OPTION_DFILTER, \ 4018 "-dfilter range,.. filter debug output to range of addresses (useful for -d cpu,exec,etc..)\n", 4019 QEMU_ARCH_ALL) 4020SRST 4021``-dfilter range1[,...]`` 4022 Filter debug output to that relevant to a range of target addresses. 4023 The filter spec can be either start+size, start-size or start..end 4024 where start end and size are the addresses and sizes required. For 4025 example: 4026 4027 :: 4028 4029 -dfilter 0x8000..0x8fff,0xffffffc000080000+0x200,0xffffffc000060000-0x1000 4030 4031 Will dump output for any code in the 0x1000 sized block starting at 4032 0x8000 and the 0x200 sized block starting at 0xffffffc000080000 and 4033 another 0x1000 sized block starting at 0xffffffc00005f000. 4034ERST 4035 4036DEF("seed", HAS_ARG, QEMU_OPTION_seed, \ 4037 "-seed number seed the pseudo-random number generator\n", 4038 QEMU_ARCH_ALL) 4039SRST 4040``-seed number`` 4041 Force the guest to use a deterministic pseudo-random number 4042 generator, seeded with number. This does not affect crypto routines 4043 within the host. 4044ERST 4045 4046DEF("L", HAS_ARG, QEMU_OPTION_L, \ 4047 "-L path set the directory for the BIOS, VGA BIOS and keymaps\n", 4048 QEMU_ARCH_ALL) 4049SRST 4050``-L path`` 4051 Set the directory for the BIOS, VGA BIOS and keymaps. 4052 4053 To list all the data directories, use ``-L help``. 4054ERST 4055 4056DEF("bios", HAS_ARG, QEMU_OPTION_bios, \ 4057 "-bios file set the filename for the BIOS\n", QEMU_ARCH_ALL) 4058SRST 4059``-bios file`` 4060 Set the filename for the BIOS. 4061ERST 4062 4063DEF("enable-kvm", 0, QEMU_OPTION_enable_kvm, \ 4064 "-enable-kvm enable KVM full virtualization support\n", QEMU_ARCH_ALL) 4065SRST 4066``-enable-kvm`` 4067 Enable KVM full virtualization support. This option is only 4068 available if KVM support is enabled when compiling. 4069ERST 4070 4071DEF("xen-domid", HAS_ARG, QEMU_OPTION_xen_domid, 4072 "-xen-domid id specify xen guest domain id\n", QEMU_ARCH_ALL) 4073DEF("xen-attach", 0, QEMU_OPTION_xen_attach, 4074 "-xen-attach attach to existing xen domain\n" 4075 " libxl will use this when starting QEMU\n", 4076 QEMU_ARCH_ALL) 4077DEF("xen-domid-restrict", 0, QEMU_OPTION_xen_domid_restrict, 4078 "-xen-domid-restrict restrict set of available xen operations\n" 4079 " to specified domain id. (Does not affect\n" 4080 " xenpv machine type).\n", 4081 QEMU_ARCH_ALL) 4082SRST 4083``-xen-domid id`` 4084 Specify xen guest domain id (XEN only). 4085 4086``-xen-attach`` 4087 Attach to existing xen domain. libxl will use this when starting 4088 QEMU (XEN only). Restrict set of available xen operations to 4089 specified domain id (XEN only). 4090ERST 4091 4092DEF("no-reboot", 0, QEMU_OPTION_no_reboot, \ 4093 "-no-reboot exit instead of rebooting\n", QEMU_ARCH_ALL) 4094SRST 4095``-no-reboot`` 4096 Exit instead of rebooting. 4097ERST 4098 4099DEF("no-shutdown", 0, QEMU_OPTION_no_shutdown, \ 4100 "-no-shutdown stop before shutdown\n", QEMU_ARCH_ALL) 4101SRST 4102``-no-shutdown`` 4103 Don't exit QEMU on guest shutdown, but instead only stop the 4104 emulation. This allows for instance switching to monitor to commit 4105 changes to the disk image. 4106ERST 4107 4108DEF("action", HAS_ARG, QEMU_OPTION_action, 4109 "-action reboot=reset|shutdown\n" 4110 " action when guest reboots [default=reset]\n" 4111 "-action shutdown=poweroff|pause\n" 4112 " action when guest shuts down [default=poweroff]\n" 4113 "-action panic=pause|shutdown|none\n" 4114 " action when guest panics [default=shutdown]\n" 4115 "-action watchdog=reset|shutdown|poweroff|inject-nmi|pause|debug|none\n" 4116 " action when watchdog fires [default=reset]\n", 4117 QEMU_ARCH_ALL) 4118SRST 4119``-action event=action`` 4120 The action parameter serves to modify QEMU's default behavior when 4121 certain guest events occur. It provides a generic method for specifying the 4122 same behaviors that are modified by the ``-no-reboot`` and ``-no-shutdown`` 4123 parameters. 4124 4125 Examples: 4126 4127 ``-action panic=none`` 4128 ``-action reboot=shutdown,shutdown=pause`` 4129 ``-watchdog i6300esb -action watchdog=pause`` 4130 4131ERST 4132 4133DEF("loadvm", HAS_ARG, QEMU_OPTION_loadvm, \ 4134 "-loadvm [tag|id]\n" \ 4135 " start right away with a saved state (loadvm in monitor)\n", 4136 QEMU_ARCH_ALL) 4137SRST 4138``-loadvm file`` 4139 Start right away with a saved state (``loadvm`` in monitor) 4140ERST 4141 4142#ifndef _WIN32 4143DEF("daemonize", 0, QEMU_OPTION_daemonize, \ 4144 "-daemonize daemonize QEMU after initializing\n", QEMU_ARCH_ALL) 4145#endif 4146SRST 4147``-daemonize`` 4148 Daemonize the QEMU process after initialization. QEMU will not 4149 detach from standard IO until it is ready to receive connections on 4150 any of its devices. This option is a useful way for external 4151 programs to launch QEMU without having to cope with initialization 4152 race conditions. 4153ERST 4154 4155DEF("option-rom", HAS_ARG, QEMU_OPTION_option_rom, \ 4156 "-option-rom rom load a file, rom, into the option ROM space\n", 4157 QEMU_ARCH_ALL) 4158SRST 4159``-option-rom file`` 4160 Load the contents of file as an option ROM. This option is useful to 4161 load things like EtherBoot. 4162ERST 4163 4164DEF("rtc", HAS_ARG, QEMU_OPTION_rtc, \ 4165 "-rtc [base=utc|localtime|<datetime>][,clock=host|rt|vm][,driftfix=none|slew]\n" \ 4166 " set the RTC base and clock, enable drift fix for clock ticks (x86 only)\n", 4167 QEMU_ARCH_ALL) 4168 4169SRST 4170``-rtc [base=utc|localtime|datetime][,clock=host|rt|vm][,driftfix=none|slew]`` 4171 Specify ``base`` as ``utc`` or ``localtime`` to let the RTC start at 4172 the current UTC or local time, respectively. ``localtime`` is 4173 required for correct date in MS-DOS or Windows. To start at a 4174 specific point in time, provide datetime in the format 4175 ``2006-06-17T16:01:21`` or ``2006-06-17``. The default base is UTC. 4176 4177 By default the RTC is driven by the host system time. This allows 4178 using of the RTC as accurate reference clock inside the guest, 4179 specifically if the host time is smoothly following an accurate 4180 external reference clock, e.g. via NTP. If you want to isolate the 4181 guest time from the host, you can set ``clock`` to ``rt`` instead, 4182 which provides a host monotonic clock if host support it. To even 4183 prevent the RTC from progressing during suspension, you can set 4184 ``clock`` to ``vm`` (virtual clock). '\ ``clock=vm``\ ' is 4185 recommended especially in icount mode in order to preserve 4186 determinism; however, note that in icount mode the speed of the 4187 virtual clock is variable and can in general differ from the host 4188 clock. 4189 4190 Enable ``driftfix`` (i386 targets only) if you experience time drift 4191 problems, specifically with Windows' ACPI HAL. This option will try 4192 to figure out how many timer interrupts were not processed by the 4193 Windows guest and will re-inject them. 4194ERST 4195 4196DEF("icount", HAS_ARG, QEMU_OPTION_icount, \ 4197 "-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=<filename>[,rrsnapshot=<snapshot>]]\n" \ 4198 " enable virtual instruction counter with 2^N clock ticks per\n" \ 4199 " instruction, enable aligning the host and virtual clocks\n" \ 4200 " or disable real time cpu sleeping, and optionally enable\n" \ 4201 " record-and-replay mode\n", QEMU_ARCH_ALL) 4202SRST 4203``-icount [shift=N|auto][,align=on|off][,sleep=on|off][,rr=record|replay,rrfile=filename[,rrsnapshot=snapshot]]`` 4204 Enable virtual instruction counter. The virtual cpu will execute one 4205 instruction every 2^N ns of virtual time. If ``auto`` is specified 4206 then the virtual cpu speed will be automatically adjusted to keep 4207 virtual time within a few seconds of real time. 4208 4209 Note that while this option can give deterministic behavior, it does 4210 not provide cycle accurate emulation. Modern CPUs contain 4211 superscalar out of order cores with complex cache hierarchies. The 4212 number of instructions executed often has little or no correlation 4213 with actual performance. 4214 4215 When the virtual cpu is sleeping, the virtual time will advance at 4216 default speed unless ``sleep=on`` is specified. With 4217 ``sleep=on``, the virtual time will jump to the next timer 4218 deadline instantly whenever the virtual cpu goes to sleep mode and 4219 will not advance if no timer is enabled. This behavior gives 4220 deterministic execution times from the guest point of view. 4221 The default if icount is enabled is ``sleep=off``. 4222 ``sleep=on`` cannot be used together with either ``shift=auto`` 4223 or ``align=on``. 4224 4225 ``align=on`` will activate the delay algorithm which will try to 4226 synchronise the host clock and the virtual clock. The goal is to 4227 have a guest running at the real frequency imposed by the shift 4228 option. Whenever the guest clock is behind the host clock and if 4229 ``align=on`` is specified then we print a message to the user to 4230 inform about the delay. Currently this option does not work when 4231 ``shift`` is ``auto``. Note: The sync algorithm will work for those 4232 shift values for which the guest clock runs ahead of the host clock. 4233 Typically this happens when the shift value is high (how high 4234 depends on the host machine). The default if icount is enabled 4235 is ``align=off``. 4236 4237 When the ``rr`` option is specified deterministic record/replay is 4238 enabled. The ``rrfile=`` option must also be provided to 4239 specify the path to the replay log. In record mode data is written 4240 to this file, and in replay mode it is read back. 4241 If the ``rrsnapshot`` option is given then it specifies a VM snapshot 4242 name. In record mode, a new VM snapshot with the given name is created 4243 at the start of execution recording. In replay mode this option 4244 specifies the snapshot name used to load the initial VM state. 4245ERST 4246 4247DEF("watchdog", HAS_ARG, QEMU_OPTION_watchdog, \ 4248 "-watchdog model\n" \ 4249 " enable virtual hardware watchdog [default=none]\n", 4250 QEMU_ARCH_ALL) 4251SRST 4252``-watchdog model`` 4253 Create a virtual hardware watchdog device. Once enabled (by a guest 4254 action), the watchdog must be periodically polled by an agent inside 4255 the guest or else the guest will be restarted. Choose a model for 4256 which your guest has drivers. 4257 4258 The model is the model of hardware watchdog to emulate. Use 4259 ``-watchdog help`` to list available hardware models. Only one 4260 watchdog can be enabled for a guest. 4261 4262 The following models may be available: 4263 4264 ``ib700`` 4265 iBASE 700 is a very simple ISA watchdog with a single timer. 4266 4267 ``i6300esb`` 4268 Intel 6300ESB I/O controller hub is a much more featureful 4269 PCI-based dual-timer watchdog. 4270 4271 ``diag288`` 4272 A virtual watchdog for s390x backed by the diagnose 288 4273 hypercall (currently KVM only). 4274ERST 4275 4276DEF("watchdog-action", HAS_ARG, QEMU_OPTION_watchdog_action, \ 4277 "-watchdog-action reset|shutdown|poweroff|inject-nmi|pause|debug|none\n" \ 4278 " action when watchdog fires [default=reset]\n", 4279 QEMU_ARCH_ALL) 4280SRST 4281``-watchdog-action action`` 4282 The action controls what QEMU will do when the watchdog timer 4283 expires. The default is ``reset`` (forcefully reset the guest). 4284 Other possible actions are: ``shutdown`` (attempt to gracefully 4285 shutdown the guest), ``poweroff`` (forcefully poweroff the guest), 4286 ``inject-nmi`` (inject a NMI into the guest), ``pause`` (pause the 4287 guest), ``debug`` (print a debug message and continue), or ``none`` 4288 (do nothing). 4289 4290 Note that the ``shutdown`` action requires that the guest responds 4291 to ACPI signals, which it may not be able to do in the sort of 4292 situations where the watchdog would have expired, and thus 4293 ``-watchdog-action shutdown`` is not recommended for production use. 4294 4295 Examples: 4296 4297 ``-watchdog i6300esb -watchdog-action pause``; \ ``-watchdog ib700`` 4298 4299ERST 4300 4301DEF("echr", HAS_ARG, QEMU_OPTION_echr, \ 4302 "-echr chr set terminal escape character instead of ctrl-a\n", 4303 QEMU_ARCH_ALL) 4304SRST 4305``-echr numeric_ascii_value`` 4306 Change the escape character used for switching to the monitor when 4307 using monitor and serial sharing. The default is ``0x01`` when using 4308 the ``-nographic`` option. ``0x01`` is equal to pressing 4309 ``Control-a``. You can select a different character from the ascii 4310 control keys where 1 through 26 map to Control-a through Control-z. 4311 For instance you could use the either of the following to change the 4312 escape character to Control-t. 4313 4314 ``-echr 0x14``; \ ``-echr 20`` 4315 4316ERST 4317 4318DEF("incoming", HAS_ARG, QEMU_OPTION_incoming, \ 4319 "-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]\n" \ 4320 "-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]\n" \ 4321 "-incoming unix:socketpath\n" \ 4322 " prepare for incoming migration, listen on\n" \ 4323 " specified protocol and socket address\n" \ 4324 "-incoming fd:fd\n" \ 4325 "-incoming exec:cmdline\n" \ 4326 " accept incoming migration on given file descriptor\n" \ 4327 " or from given external command\n" \ 4328 "-incoming defer\n" \ 4329 " wait for the URI to be specified via migrate_incoming\n", 4330 QEMU_ARCH_ALL) 4331SRST 4332``-incoming tcp:[host]:port[,to=maxport][,ipv4=on|off][,ipv6=on|off]`` 4333 \ 4334``-incoming rdma:host:port[,ipv4=on|off][,ipv6=on|off]`` 4335 Prepare for incoming migration, listen on a given tcp port. 4336 4337``-incoming unix:socketpath`` 4338 Prepare for incoming migration, listen on a given unix socket. 4339 4340``-incoming fd:fd`` 4341 Accept incoming migration from a given filedescriptor. 4342 4343``-incoming exec:cmdline`` 4344 Accept incoming migration as an output from specified external 4345 command. 4346 4347``-incoming defer`` 4348 Wait for the URI to be specified via migrate\_incoming. The monitor 4349 can be used to change settings (such as migration parameters) prior 4350 to issuing the migrate\_incoming to allow the migration to begin. 4351ERST 4352 4353DEF("only-migratable", 0, QEMU_OPTION_only_migratable, \ 4354 "-only-migratable allow only migratable devices\n", QEMU_ARCH_ALL) 4355SRST 4356``-only-migratable`` 4357 Only allow migratable devices. Devices will not be allowed to enter 4358 an unmigratable state. 4359ERST 4360 4361DEF("nodefaults", 0, QEMU_OPTION_nodefaults, \ 4362 "-nodefaults don't create default devices\n", QEMU_ARCH_ALL) 4363SRST 4364``-nodefaults`` 4365 Don't create default devices. Normally, QEMU sets the default 4366 devices like serial port, parallel port, virtual console, monitor 4367 device, VGA adapter, floppy and CD-ROM drive and others. The 4368 ``-nodefaults`` option will disable all those default devices. 4369ERST 4370 4371#ifndef _WIN32 4372DEF("chroot", HAS_ARG, QEMU_OPTION_chroot, \ 4373 "-chroot dir chroot to dir just before starting the VM\n", 4374 QEMU_ARCH_ALL) 4375#endif 4376SRST 4377``-chroot dir`` 4378 Immediately before starting guest execution, chroot to the specified 4379 directory. Especially useful in combination with -runas. 4380ERST 4381 4382#ifndef _WIN32 4383DEF("runas", HAS_ARG, QEMU_OPTION_runas, \ 4384 "-runas user change to user id user just before starting the VM\n" \ 4385 " user can be numeric uid:gid instead\n", 4386 QEMU_ARCH_ALL) 4387#endif 4388SRST 4389``-runas user`` 4390 Immediately before starting guest execution, drop root privileges, 4391 switching to the specified user. 4392ERST 4393 4394DEF("prom-env", HAS_ARG, QEMU_OPTION_prom_env, 4395 "-prom-env variable=value\n" 4396 " set OpenBIOS nvram variables\n", 4397 QEMU_ARCH_PPC | QEMU_ARCH_SPARC) 4398SRST 4399``-prom-env variable=value`` 4400 Set OpenBIOS nvram variable to given value (PPC, SPARC only). 4401 4402 :: 4403 4404 qemu-system-sparc -prom-env 'auto-boot?=false' \ 4405 -prom-env 'boot-device=sd(0,2,0):d' -prom-env 'boot-args=linux single' 4406 4407 :: 4408 4409 qemu-system-ppc -prom-env 'auto-boot?=false' \ 4410 -prom-env 'boot-device=hd:2,\yaboot' \ 4411 -prom-env 'boot-args=conf=hd:2,\yaboot.conf' 4412ERST 4413DEF("semihosting", 0, QEMU_OPTION_semihosting, 4414 "-semihosting semihosting mode\n", 4415 QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA | 4416 QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV) 4417SRST 4418``-semihosting`` 4419 Enable semihosting mode (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V only). 4420 4421 Note that this allows guest direct access to the host filesystem, so 4422 should only be used with a trusted guest OS. 4423 4424 See the -semihosting-config option documentation for further 4425 information about the facilities this enables. 4426ERST 4427DEF("semihosting-config", HAS_ARG, QEMU_OPTION_semihosting_config, 4428 "-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,arg=str[,...]]\n" \ 4429 " semihosting configuration\n", 4430QEMU_ARCH_ARM | QEMU_ARCH_M68K | QEMU_ARCH_XTENSA | 4431QEMU_ARCH_MIPS | QEMU_ARCH_NIOS2 | QEMU_ARCH_RISCV) 4432SRST 4433``-semihosting-config [enable=on|off][,target=native|gdb|auto][,chardev=id][,arg=str[,...]]`` 4434 Enable and configure semihosting (ARM, M68K, Xtensa, MIPS, Nios II, RISC-V 4435 only). 4436 4437 Note that this allows guest direct access to the host filesystem, so 4438 should only be used with a trusted guest OS. 4439 4440 On Arm this implements the standard semihosting API, version 2.0. 4441 4442 On M68K this implements the "ColdFire GDB" interface used by 4443 libgloss. 4444 4445 Xtensa semihosting provides basic file IO calls, such as 4446 open/read/write/seek/select. Tensilica baremetal libc for ISS and 4447 linux platform "sim" use this interface. 4448 4449 On RISC-V this implements the standard semihosting API, version 0.2. 4450 4451 ``target=native|gdb|auto`` 4452 Defines where the semihosting calls will be addressed, to QEMU 4453 (``native``) or to GDB (``gdb``). The default is ``auto``, which 4454 means ``gdb`` during debug sessions and ``native`` otherwise. 4455 4456 ``chardev=str1`` 4457 Send the output to a chardev backend output for native or auto 4458 output when not in gdb 4459 4460 ``arg=str1,arg=str2,...`` 4461 Allows the user to pass input arguments, and can be used 4462 multiple times to build up a list. The old-style 4463 ``-kernel``/``-append`` method of passing a command line is 4464 still supported for backward compatibility. If both the 4465 ``--semihosting-config arg`` and the ``-kernel``/``-append`` are 4466 specified, the former is passed to semihosting as it always 4467 takes precedence. 4468ERST 4469DEF("old-param", 0, QEMU_OPTION_old_param, 4470 "-old-param old param mode\n", QEMU_ARCH_ARM) 4471SRST 4472``-old-param`` 4473 Old param mode (ARM only). 4474ERST 4475 4476DEF("sandbox", HAS_ARG, QEMU_OPTION_sandbox, \ 4477 "-sandbox on[,obsolete=allow|deny][,elevateprivileges=allow|deny|children]\n" \ 4478 " [,spawn=allow|deny][,resourcecontrol=allow|deny]\n" \ 4479 " Enable seccomp mode 2 system call filter (default 'off').\n" \ 4480 " use 'obsolete' to allow obsolete system calls that are provided\n" \ 4481 " by the kernel, but typically no longer used by modern\n" \ 4482 " C library implementations.\n" \ 4483 " use 'elevateprivileges' to allow or deny the QEMU process ability\n" \ 4484 " to elevate privileges using set*uid|gid system calls.\n" \ 4485 " The value 'children' will deny set*uid|gid system calls for\n" \ 4486 " main QEMU process but will allow forks and execves to run unprivileged\n" \ 4487 " use 'spawn' to avoid QEMU to spawn new threads or processes by\n" \ 4488 " blocking *fork and execve\n" \ 4489 " use 'resourcecontrol' to disable process affinity and schedular priority\n", 4490 QEMU_ARCH_ALL) 4491SRST 4492``-sandbox arg[,obsolete=string][,elevateprivileges=string][,spawn=string][,resourcecontrol=string]`` 4493 Enable Seccomp mode 2 system call filter. 'on' will enable syscall 4494 filtering and 'off' will disable it. The default is 'off'. 4495 4496 ``obsolete=string`` 4497 Enable Obsolete system calls 4498 4499 ``elevateprivileges=string`` 4500 Disable set\*uid\|gid system calls 4501 4502 ``spawn=string`` 4503 Disable \*fork and execve 4504 4505 ``resourcecontrol=string`` 4506 Disable process affinity and schedular priority 4507ERST 4508 4509DEF("readconfig", HAS_ARG, QEMU_OPTION_readconfig, 4510 "-readconfig <file>\n", QEMU_ARCH_ALL) 4511SRST 4512``-readconfig file`` 4513 Read device configuration from file. This approach is useful when 4514 you want to spawn QEMU process with many command line options but 4515 you don't want to exceed the command line character limit. 4516ERST 4517DEF("writeconfig", HAS_ARG, QEMU_OPTION_writeconfig, 4518 "-writeconfig <file>\n" 4519 " read/write config file (deprecated)\n", QEMU_ARCH_ALL) 4520SRST 4521ERST 4522 4523DEF("no-user-config", 0, QEMU_OPTION_nouserconfig, 4524 "-no-user-config\n" 4525 " do not load default user-provided config files at startup\n", 4526 QEMU_ARCH_ALL) 4527SRST 4528``-no-user-config`` 4529 The ``-no-user-config`` option makes QEMU not load any of the 4530 user-provided config files on sysconfdir. 4531ERST 4532 4533DEF("trace", HAS_ARG, QEMU_OPTION_trace, 4534 "-trace [[enable=]<pattern>][,events=<file>][,file=<file>]\n" 4535 " specify tracing options\n", 4536 QEMU_ARCH_ALL) 4537SRST 4538``-trace [[enable=]pattern][,events=file][,file=file]`` 4539 .. include:: ../qemu-option-trace.rst.inc 4540 4541ERST 4542DEF("plugin", HAS_ARG, QEMU_OPTION_plugin, 4543 "-plugin [file=]<file>[,<argname>=<argvalue>]\n" 4544 " load a plugin\n", 4545 QEMU_ARCH_ALL) 4546SRST 4547``-plugin file=file[,argname=argvalue]`` 4548 Load a plugin. 4549 4550 ``file=file`` 4551 Load the given plugin from a shared library file. 4552 4553 ``argname=argvalue`` 4554 Argument passed to the plugin. (Can be given multiple times.) 4555ERST 4556 4557HXCOMM Internal use 4558DEF("qtest", HAS_ARG, QEMU_OPTION_qtest, "", QEMU_ARCH_ALL) 4559DEF("qtest-log", HAS_ARG, QEMU_OPTION_qtest_log, "", QEMU_ARCH_ALL) 4560 4561#ifdef __linux__ 4562DEF("enable-fips", 0, QEMU_OPTION_enablefips, 4563 "-enable-fips enable FIPS 140-2 compliance\n", 4564 QEMU_ARCH_ALL) 4565#endif 4566SRST 4567``-enable-fips`` 4568 Enable FIPS 140-2 compliance mode. 4569ERST 4570 4571DEF("msg", HAS_ARG, QEMU_OPTION_msg, 4572 "-msg [timestamp[=on|off]][,guest-name=[on|off]]\n" 4573 " control error message format\n" 4574 " timestamp=on enables timestamps (default: off)\n" 4575 " guest-name=on enables guest name prefix but only if\n" 4576 " -name guest option is set (default: off)\n", 4577 QEMU_ARCH_ALL) 4578SRST 4579``-msg [timestamp[=on|off]][,guest-name[=on|off]]`` 4580 Control error message format. 4581 4582 ``timestamp=on|off`` 4583 Prefix messages with a timestamp. Default is off. 4584 4585 ``guest-name=on|off`` 4586 Prefix messages with guest name but only if -name guest option is set 4587 otherwise the option is ignored. Default is off. 4588ERST 4589 4590DEF("dump-vmstate", HAS_ARG, QEMU_OPTION_dump_vmstate, 4591 "-dump-vmstate <file>\n" 4592 " Output vmstate information in JSON format to file.\n" 4593 " Use the scripts/vmstate-static-checker.py file to\n" 4594 " check for possible regressions in migration code\n" 4595 " by comparing two such vmstate dumps.\n", 4596 QEMU_ARCH_ALL) 4597SRST 4598``-dump-vmstate file`` 4599 Dump json-encoded vmstate information for current machine type to 4600 file in file 4601ERST 4602 4603DEF("enable-sync-profile", 0, QEMU_OPTION_enable_sync_profile, 4604 "-enable-sync-profile\n" 4605 " enable synchronization profiling\n", 4606 QEMU_ARCH_ALL) 4607SRST 4608``-enable-sync-profile`` 4609 Enable synchronization profiling. 4610ERST 4611 4612DEFHEADING() 4613 4614DEFHEADING(Generic object creation:) 4615 4616DEF("object", HAS_ARG, QEMU_OPTION_object, 4617 "-object TYPENAME[,PROP1=VALUE1,...]\n" 4618 " create a new object of type TYPENAME setting properties\n" 4619 " in the order they are specified. Note that the 'id'\n" 4620 " property must be set. These objects are placed in the\n" 4621 " '/objects' path.\n", 4622 QEMU_ARCH_ALL) 4623SRST 4624``-object typename[,prop1=value1,...]`` 4625 Create a new object of type typename setting properties in the order 4626 they are specified. Note that the 'id' property must be set. These 4627 objects are placed in the '/objects' path. 4628 4629 ``-object memory-backend-file,id=id,size=size,mem-path=dir,share=on|off,discard-data=on|off,merge=on|off,dump=on|off,prealloc=on|off,host-nodes=host-nodes,policy=default|preferred|bind|interleave,align=align,readonly=on|off`` 4630 Creates a memory file backend object, which can be used to back 4631 the guest RAM with huge pages. 4632 4633 The ``id`` parameter is a unique ID that will be used to 4634 reference this memory region in other parameters, e.g. ``-numa``, 4635 ``-device nvdimm``, etc. 4636 4637 The ``size`` option provides the size of the memory region, and 4638 accepts common suffixes, e.g. ``500M``. 4639 4640 The ``mem-path`` provides the path to either a shared memory or 4641 huge page filesystem mount. 4642 4643 The ``share`` boolean option determines whether the memory 4644 region is marked as private to QEMU, or shared. The latter 4645 allows a co-operating external process to access the QEMU memory 4646 region. 4647 4648 The ``share`` is also required for pvrdma devices due to 4649 limitations in the RDMA API provided by Linux. 4650 4651 Setting share=on might affect the ability to configure NUMA 4652 bindings for the memory backend under some circumstances, see 4653 Documentation/vm/numa\_memory\_policy.txt on the Linux kernel 4654 source tree for additional details. 4655 4656 Setting the ``discard-data`` boolean option to on indicates that 4657 file contents can be destroyed when QEMU exits, to avoid 4658 unnecessarily flushing data to the backing file. Note that 4659 ``discard-data`` is only an optimization, and QEMU might not 4660 discard file contents if it aborts unexpectedly or is terminated 4661 using SIGKILL. 4662 4663 The ``merge`` boolean option enables memory merge, also known as 4664 MADV\_MERGEABLE, so that Kernel Samepage Merging will consider 4665 the pages for memory deduplication. 4666 4667 Setting the ``dump`` boolean option to off excludes the memory 4668 from core dumps. This feature is also known as MADV\_DONTDUMP. 4669 4670 The ``prealloc`` boolean option enables memory preallocation. 4671 4672 The ``host-nodes`` option binds the memory range to a list of 4673 NUMA host nodes. 4674 4675 The ``policy`` option sets the NUMA policy to one of the 4676 following values: 4677 4678 ``default`` 4679 default host policy 4680 4681 ``preferred`` 4682 prefer the given host node list for allocation 4683 4684 ``bind`` 4685 restrict memory allocation to the given host node list 4686 4687 ``interleave`` 4688 interleave memory allocations across the given host node 4689 list 4690 4691 The ``align`` option specifies the base address alignment when 4692 QEMU mmap(2) ``mem-path``, and accepts common suffixes, eg 4693 ``2M``. Some backend store specified by ``mem-path`` requires an 4694 alignment different than the default one used by QEMU, eg the 4695 device DAX /dev/dax0.0 requires 2M alignment rather than 4K. In 4696 such cases, users can specify the required alignment via this 4697 option. 4698 4699 The ``pmem`` option specifies whether the backing file specified 4700 by ``mem-path`` is in host persistent memory that can be 4701 accessed using the SNIA NVM programming model (e.g. Intel 4702 NVDIMM). If ``pmem`` is set to 'on', QEMU will take necessary 4703 operations to guarantee the persistence of its own writes to 4704 ``mem-path`` (e.g. in vNVDIMM label emulation and live 4705 migration). Also, we will map the backend-file with MAP\_SYNC 4706 flag, which ensures the file metadata is in sync for 4707 ``mem-path`` in case of host crash or a power failure. MAP\_SYNC 4708 requires support from both the host kernel (since Linux kernel 4709 4.15) and the filesystem of ``mem-path`` mounted with DAX 4710 option. 4711 4712 The ``readonly`` option specifies whether the backing file is opened 4713 read-only or read-write (default). 4714 4715 ``-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`` 4716 Creates a memory backend object, which can be used to back the 4717 guest RAM. Memory backend objects offer more control than the 4718 ``-m`` option that is traditionally used to define guest RAM. 4719 Please refer to ``memory-backend-file`` for a description of the 4720 options. 4721 4722 ``-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`` 4723 Creates an anonymous memory file backend object, which allows 4724 QEMU to share the memory with an external process (e.g. when 4725 using vhost-user). The memory is allocated with memfd and 4726 optional sealing. (Linux only) 4727 4728 The ``seal`` option creates a sealed-file, that will block 4729 further resizing the memory ('on' by default). 4730 4731 The ``hugetlb`` option specify the file to be created resides in 4732 the hugetlbfs filesystem (since Linux 4.14). Used in conjunction 4733 with the ``hugetlb`` option, the ``hugetlbsize`` option specify 4734 the hugetlb page size on systems that support multiple hugetlb 4735 page sizes (it must be a power of 2 value supported by the 4736 system). 4737 4738 In some versions of Linux, the ``hugetlb`` option is 4739 incompatible with the ``seal`` option (requires at least Linux 4740 4.16). 4741 4742 Please refer to ``memory-backend-file`` for a description of the 4743 other options. 4744 4745 The ``share`` boolean option is on by default with memfd. 4746 4747 ``-object rng-builtin,id=id`` 4748 Creates a random number generator backend which obtains entropy 4749 from QEMU builtin functions. The ``id`` parameter is a unique ID 4750 that will be used to reference this entropy backend from the 4751 ``virtio-rng`` device. By default, the ``virtio-rng`` device 4752 uses this RNG backend. 4753 4754 ``-object rng-random,id=id,filename=/dev/random`` 4755 Creates a random number generator backend which obtains entropy 4756 from a device on the host. The ``id`` parameter is a unique ID 4757 that will be used to reference this entropy backend from the 4758 ``virtio-rng`` device. The ``filename`` parameter specifies 4759 which file to obtain entropy from and if omitted defaults to 4760 ``/dev/urandom``. 4761 4762 ``-object rng-egd,id=id,chardev=chardevid`` 4763 Creates a random number generator backend which obtains entropy 4764 from an external daemon running on the host. The ``id`` 4765 parameter is a unique ID that will be used to reference this 4766 entropy backend from the ``virtio-rng`` device. The ``chardev`` 4767 parameter is the unique ID of a character device backend that 4768 provides the connection to the RNG daemon. 4769 4770 ``-object tls-creds-anon,id=id,endpoint=endpoint,dir=/path/to/cred/dir,verify-peer=on|off`` 4771 Creates a TLS anonymous credentials object, which can be used to 4772 provide TLS support on network backends. The ``id`` parameter is 4773 a unique ID which network backends will use to access the 4774 credentials. The ``endpoint`` is either ``server`` or ``client`` 4775 depending on whether the QEMU network backend that uses the 4776 credentials will be acting as a client or as a server. If 4777 ``verify-peer`` is enabled (the default) then once the handshake 4778 is completed, the peer credentials will be verified, though this 4779 is a no-op for anonymous credentials. 4780 4781 The dir parameter tells QEMU where to find the credential files. 4782 For server endpoints, this directory may contain a file 4783 dh-params.pem providing diffie-hellman parameters to use for the 4784 TLS server. If the file is missing, QEMU will generate a set of 4785 DH parameters at startup. This is a computationally expensive 4786 operation that consumes random pool entropy, so it is 4787 recommended that a persistent set of parameters be generated 4788 upfront and saved. 4789 4790 ``-object tls-creds-psk,id=id,endpoint=endpoint,dir=/path/to/keys/dir[,username=username]`` 4791 Creates a TLS Pre-Shared Keys (PSK) credentials object, which 4792 can be used to provide TLS support on network backends. The 4793 ``id`` parameter is a unique ID which network backends will use 4794 to access the credentials. The ``endpoint`` is either ``server`` 4795 or ``client`` depending on whether the QEMU network backend that 4796 uses the credentials will be acting as a client or as a server. 4797 For clients only, ``username`` is the username which will be 4798 sent to the server. If omitted it defaults to "qemu". 4799 4800 The dir parameter tells QEMU where to find the keys file. It is 4801 called "dir/keys.psk" and contains "username:key" pairs. This 4802 file can most easily be created using the GnuTLS ``psktool`` 4803 program. 4804 4805 For server endpoints, dir may also contain a file dh-params.pem 4806 providing diffie-hellman parameters to use for the TLS server. 4807 If the file is missing, QEMU will generate a set of DH 4808 parameters at startup. This is a computationally expensive 4809 operation that consumes random pool entropy, so it is 4810 recommended that a persistent set of parameters be generated up 4811 front and saved. 4812 4813 ``-object tls-creds-x509,id=id,endpoint=endpoint,dir=/path/to/cred/dir,priority=priority,verify-peer=on|off,passwordid=id`` 4814 Creates a TLS anonymous credentials object, which can be used to 4815 provide TLS support on network backends. The ``id`` parameter is 4816 a unique ID which network backends will use to access the 4817 credentials. The ``endpoint`` is either ``server`` or ``client`` 4818 depending on whether the QEMU network backend that uses the 4819 credentials will be acting as a client or as a server. If 4820 ``verify-peer`` is enabled (the default) then once the handshake 4821 is completed, the peer credentials will be verified. With x509 4822 certificates, this implies that the clients must be provided 4823 with valid client certificates too. 4824 4825 The dir parameter tells QEMU where to find the credential files. 4826 For server endpoints, this directory may contain a file 4827 dh-params.pem providing diffie-hellman parameters to use for the 4828 TLS server. If the file is missing, QEMU will generate a set of 4829 DH parameters at startup. This is a computationally expensive 4830 operation that consumes random pool entropy, so it is 4831 recommended that a persistent set of parameters be generated 4832 upfront and saved. 4833 4834 For x509 certificate credentials the directory will contain 4835 further files providing the x509 certificates. The certificates 4836 must be stored in PEM format, in filenames ca-cert.pem, 4837 ca-crl.pem (optional), server-cert.pem (only servers), 4838 server-key.pem (only servers), client-cert.pem (only clients), 4839 and client-key.pem (only clients). 4840 4841 For the server-key.pem and client-key.pem files which contain 4842 sensitive private keys, it is possible to use an encrypted 4843 version by providing the passwordid parameter. This provides the 4844 ID of a previously created ``secret`` object containing the 4845 password for decryption. 4846 4847 The priority parameter allows to override the global default 4848 priority used by gnutls. This can be useful if the system 4849 administrator needs to use a weaker set of crypto priorities for 4850 QEMU without potentially forcing the weakness onto all 4851 applications. Or conversely if one wants wants a stronger 4852 default for QEMU than for all other applications, they can do 4853 this through this parameter. Its format is a gnutls priority 4854 string as described at 4855 https://gnutls.org/manual/html_node/Priority-Strings.html. 4856 4857 ``-object tls-cipher-suites,id=id,priority=priority`` 4858 Creates a TLS cipher suites object, which can be used to control 4859 the TLS cipher/protocol algorithms that applications are permitted 4860 to use. 4861 4862 The ``id`` parameter is a unique ID which frontends will use to 4863 access the ordered list of permitted TLS cipher suites from the 4864 host. 4865 4866 The ``priority`` parameter allows to override the global default 4867 priority used by gnutls. This can be useful if the system 4868 administrator needs to use a weaker set of crypto priorities for 4869 QEMU without potentially forcing the weakness onto all 4870 applications. Or conversely if one wants wants a stronger 4871 default for QEMU than for all other applications, they can do 4872 this through this parameter. Its format is a gnutls priority 4873 string as described at 4874 https://gnutls.org/manual/html_node/Priority-Strings.html. 4875 4876 An example of use of this object is to control UEFI HTTPS Boot. 4877 The tls-cipher-suites object exposes the ordered list of permitted 4878 TLS cipher suites from the host side to the guest firmware, via 4879 fw_cfg. The list is represented as an array of IANA_TLS_CIPHER 4880 objects. The firmware uses the IANA_TLS_CIPHER array for configuring 4881 guest-side TLS. 4882 4883 In the following example, the priority at which the host-side policy 4884 is retrieved is given by the ``priority`` property. 4885 Given that QEMU uses GNUTLS, ``priority=@SYSTEM`` may be used to 4886 refer to /etc/crypto-policies/back-ends/gnutls.config. 4887 4888 .. parsed-literal:: 4889 4890 # |qemu_system| \\ 4891 -object tls-cipher-suites,id=mysuite0,priority=@SYSTEM \\ 4892 -fw_cfg name=etc/edk2/https/ciphers,gen_id=mysuite0 4893 4894 ``-object filter-buffer,id=id,netdev=netdevid,interval=t[,queue=all|rx|tx][,status=on|off][,position=head|tail|id=<id>][,insert=behind|before]`` 4895 Interval t can't be 0, this filter batches the packet delivery: 4896 all packets arriving in a given interval on netdev netdevid are 4897 delayed until the end of the interval. Interval is in 4898 microseconds. ``status`` is optional that indicate whether the 4899 netfilter is on (enabled) or off (disabled), the default status 4900 for netfilter will be 'on'. 4901 4902 queue all\|rx\|tx is an option that can be applied to any 4903 netfilter. 4904 4905 ``all``: the filter is attached both to the receive and the 4906 transmit queue of the netdev (default). 4907 4908 ``rx``: the filter is attached to the receive queue of the 4909 netdev, where it will receive packets sent to the netdev. 4910 4911 ``tx``: the filter is attached to the transmit queue of the 4912 netdev, where it will receive packets sent by the netdev. 4913 4914 position head\|tail\|id=<id> is an option to specify where the 4915 filter should be inserted in the filter list. It can be applied 4916 to any netfilter. 4917 4918 ``head``: the filter is inserted at the head of the filter list, 4919 before any existing filters. 4920 4921 ``tail``: the filter is inserted at the tail of the filter list, 4922 behind any existing filters (default). 4923 4924 ``id=<id>``: the filter is inserted before or behind the filter 4925 specified by <id>, see the insert option below. 4926 4927 insert behind\|before is an option to specify where to insert 4928 the new filter relative to the one specified with 4929 position=id=<id>. It can be applied to any netfilter. 4930 4931 ``before``: insert before the specified filter. 4932 4933 ``behind``: insert behind the specified filter (default). 4934 4935 ``-object filter-mirror,id=id,netdev=netdevid,outdev=chardevid,queue=all|rx|tx[,vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]`` 4936 filter-mirror on netdev netdevid,mirror net packet to 4937 chardevchardevid, if it has the vnet\_hdr\_support flag, 4938 filter-mirror will mirror packet with vnet\_hdr\_len. 4939 4940 ``-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]`` 4941 filter-redirector on netdev netdevid,redirect filter's net 4942 packet to chardev chardevid,and redirect indev's packet to 4943 filter.if it has the vnet\_hdr\_support flag, filter-redirector 4944 will redirect packet with vnet\_hdr\_len. Create a 4945 filter-redirector we need to differ outdev id from indev id, id 4946 can not be the same. we can just use indev or outdev, but at 4947 least one of indev or outdev need to be specified. 4948 4949 ``-object filter-rewriter,id=id,netdev=netdevid,queue=all|rx|tx,[vnet_hdr_support][,position=head|tail|id=<id>][,insert=behind|before]`` 4950 Filter-rewriter is a part of COLO project.It will rewrite tcp 4951 packet to secondary from primary to keep secondary tcp 4952 connection,and rewrite tcp packet to primary from secondary make 4953 tcp packet can be handled by client.if it has the 4954 vnet\_hdr\_support flag, we can parse packet with vnet header. 4955 4956 usage: colo secondary: -object 4957 filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 -object 4958 filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 -object 4959 filter-rewriter,id=rew0,netdev=hn0,queue=all 4960 4961 ``-object filter-dump,id=id,netdev=dev[,file=filename][,maxlen=len][,position=head|tail|id=<id>][,insert=behind|before]`` 4962 Dump the network traffic on netdev dev to the file specified by 4963 filename. At most len bytes (64k by default) per packet are 4964 stored. The file format is libpcap, so it can be analyzed with 4965 tools such as tcpdump or Wireshark. 4966 4967 ``-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}]`` 4968 Colo-compare gets packet from primary\_in chardevid and 4969 secondary\_in, then compare whether the payload of primary packet 4970 and secondary packet are the same. If same, it will output 4971 primary packet to out\_dev, else it will notify COLO-framework to do 4972 checkpoint and send primary packet to out\_dev. In order to 4973 improve efficiency, we need to put the task of comparison in 4974 another iothread. If it has the vnet\_hdr\_support flag, 4975 colo compare will send/recv packet with vnet\_hdr\_len. 4976 The compare\_timeout=@var{ms} determines the maximum time of the 4977 colo-compare hold the packet. The expired\_scan\_cycle=@var{ms} 4978 is to set the period of scanning expired primary node network packets. 4979 The max\_queue\_size=@var{size} is to set the max compare queue 4980 size depend on user environment. 4981 If user want to use Xen COLO, need to add the notify\_dev to 4982 notify Xen colo-frame to do checkpoint. 4983 4984 COLO-compare must be used with the help of filter-mirror, 4985 filter-redirector and filter-rewriter. 4986 4987 :: 4988 4989 KVM COLO 4990 4991 primary: 4992 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown 4993 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66 4994 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off 4995 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off 4996 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off 4997 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001 4998 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off 4999 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005 5000 -object iothread,id=iothread1 5001 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0 5002 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out 5003 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0 5004 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,iothread=iothread1 5005 5006 secondary: 5007 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown 5008 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66 5009 -chardev socket,id=red0,host=3.3.3.3,port=9003 5010 -chardev socket,id=red1,host=3.3.3.3,port=9004 5011 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 5012 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 5013 5014 5015 Xen COLO 5016 5017 primary: 5018 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,downscript=/etc/qemu-ifdown 5019 -device e1000,id=e0,netdev=hn0,mac=52:a4:00:12:78:66 5020 -chardev socket,id=mirror0,host=3.3.3.3,port=9003,server=on,wait=off 5021 -chardev socket,id=compare1,host=3.3.3.3,port=9004,server=on,wait=off 5022 -chardev socket,id=compare0,host=3.3.3.3,port=9001,server=on,wait=off 5023 -chardev socket,id=compare0-0,host=3.3.3.3,port=9001 5024 -chardev socket,id=compare_out,host=3.3.3.3,port=9005,server=on,wait=off 5025 -chardev socket,id=compare_out0,host=3.3.3.3,port=9005 5026 -chardev socket,id=notify_way,host=3.3.3.3,port=9009,server=on,wait=off 5027 -object filter-mirror,id=m0,netdev=hn0,queue=tx,outdev=mirror0 5028 -object filter-redirector,netdev=hn0,id=redire0,queue=rx,indev=compare_out 5029 -object filter-redirector,netdev=hn0,id=redire1,queue=rx,outdev=compare0 5030 -object iothread,id=iothread1 5031 -object colo-compare,id=comp0,primary_in=compare0-0,secondary_in=compare1,outdev=compare_out0,notify_dev=nofity_way,iothread=iothread1 5032 5033 secondary: 5034 -netdev tap,id=hn0,vhost=off,script=/etc/qemu-ifup,down script=/etc/qemu-ifdown 5035 -device e1000,netdev=hn0,mac=52:a4:00:12:78:66 5036 -chardev socket,id=red0,host=3.3.3.3,port=9003 5037 -chardev socket,id=red1,host=3.3.3.3,port=9004 5038 -object filter-redirector,id=f1,netdev=hn0,queue=tx,indev=red0 5039 -object filter-redirector,id=f2,netdev=hn0,queue=rx,outdev=red1 5040 5041 If you want to know the detail of above command line, you can 5042 read the colo-compare git log. 5043 5044 ``-object cryptodev-backend-builtin,id=id[,queues=queues]`` 5045 Creates a cryptodev backend which executes crypto opreation from 5046 the QEMU cipher APIS. The id parameter is a unique ID that will 5047 be used to reference this cryptodev backend from the 5048 ``virtio-crypto`` device. The queues parameter is optional, 5049 which specify the queue number of cryptodev backend, the default 5050 of queues is 1. 5051 5052 .. parsed-literal:: 5053 5054 # |qemu_system| \\ 5055 [...] \\ 5056 -object cryptodev-backend-builtin,id=cryptodev0 \\ 5057 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\ 5058 [...] 5059 5060 ``-object cryptodev-vhost-user,id=id,chardev=chardevid[,queues=queues]`` 5061 Creates a vhost-user cryptodev backend, backed by a chardev 5062 chardevid. The id parameter is a unique ID that will be used to 5063 reference this cryptodev backend from the ``virtio-crypto`` 5064 device. The chardev should be a unix domain socket backed one. 5065 The vhost-user uses a specifically defined protocol to pass 5066 vhost ioctl replacement messages to an application on the other 5067 end of the socket. The queues parameter is optional, which 5068 specify the queue number of cryptodev backend for multiqueue 5069 vhost-user, the default of queues is 1. 5070 5071 .. parsed-literal:: 5072 5073 # |qemu_system| \\ 5074 [...] \\ 5075 -chardev socket,id=chardev0,path=/path/to/socket \\ 5076 -object cryptodev-vhost-user,id=cryptodev0,chardev=chardev0 \\ 5077 -device virtio-crypto-pci,id=crypto0,cryptodev=cryptodev0 \\ 5078 [...] 5079 5080 ``-object secret,id=id,data=string,format=raw|base64[,keyid=secretid,iv=string]`` 5081 \ 5082 ``-object secret,id=id,file=filename,format=raw|base64[,keyid=secretid,iv=string]`` 5083 Defines a secret to store a password, encryption key, or some 5084 other sensitive data. The sensitive data can either be passed 5085 directly via the data parameter, or indirectly via the file 5086 parameter. Using the data parameter is insecure unless the 5087 sensitive data is encrypted. 5088 5089 The sensitive data can be provided in raw format (the default), 5090 or base64. When encoded as JSON, the raw format only supports 5091 valid UTF-8 characters, so base64 is recommended for sending 5092 binary data. QEMU will convert from which ever format is 5093 provided to the format it needs internally. eg, an RBD password 5094 can be provided in raw format, even though it will be base64 5095 encoded when passed onto the RBD sever. 5096 5097 For added protection, it is possible to encrypt the data 5098 associated with a secret using the AES-256-CBC cipher. Use of 5099 encryption is indicated by providing the keyid and iv 5100 parameters. The keyid parameter provides the ID of a previously 5101 defined secret that contains the AES-256 decryption key. This 5102 key should be 32-bytes long and be base64 encoded. The iv 5103 parameter provides the random initialization vector used for 5104 encryption of this particular secret and should be a base64 5105 encrypted string of the 16-byte IV. 5106 5107 The simplest (insecure) usage is to provide the secret inline 5108 5109 .. parsed-literal:: 5110 5111 # |qemu_system| -object secret,id=sec0,data=letmein,format=raw 5112 5113 The simplest secure usage is to provide the secret via a file 5114 5115 # printf "letmein" > mypasswd.txt # QEMU\_SYSTEM\_MACRO -object 5116 secret,id=sec0,file=mypasswd.txt,format=raw 5117 5118 For greater security, AES-256-CBC should be used. To illustrate 5119 usage, consider the openssl command line tool which can encrypt 5120 the data. Note that when encrypting, the plaintext must be 5121 padded to the cipher block size (32 bytes) using the standard 5122 PKCS#5/6 compatible padding algorithm. 5123 5124 First a master key needs to be created in base64 encoding: 5125 5126 :: 5127 5128 # openssl rand -base64 32 > key.b64 5129 # KEY=$(base64 -d key.b64 | hexdump -v -e '/1 "%02X"') 5130 5131 Each secret to be encrypted needs to have a random 5132 initialization vector generated. These do not need to be kept 5133 secret 5134 5135 :: 5136 5137 # openssl rand -base64 16 > iv.b64 5138 # IV=$(base64 -d iv.b64 | hexdump -v -e '/1 "%02X"') 5139 5140 The secret to be defined can now be encrypted, in this case 5141 we're telling openssl to base64 encode the result, but it could 5142 be left as raw bytes if desired. 5143 5144 :: 5145 5146 # SECRET=$(printf "letmein" | 5147 openssl enc -aes-256-cbc -a -K $KEY -iv $IV) 5148 5149 When launching QEMU, create a master secret pointing to 5150 ``key.b64`` and specify that to be used to decrypt the user 5151 password. Pass the contents of ``iv.b64`` to the second secret 5152 5153 .. parsed-literal:: 5154 5155 # |qemu_system| \\ 5156 -object secret,id=secmaster0,format=base64,file=key.b64 \\ 5157 -object secret,id=sec0,keyid=secmaster0,format=base64,\\ 5158 data=$SECRET,iv=$(<iv.b64) 5159 5160 ``-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]`` 5161 Create a Secure Encrypted Virtualization (SEV) guest object, 5162 which can be used to provide the guest memory encryption support 5163 on AMD processors. 5164 5165 When memory encryption is enabled, one of the physical address 5166 bit (aka the C-bit) is utilized to mark if a memory page is 5167 protected. The ``cbitpos`` is used to provide the C-bit 5168 position. The C-bit position is Host family dependent hence user 5169 must provide this value. On EPYC, the value should be 47. 5170 5171 When memory encryption is enabled, we loose certain bits in 5172 physical address space. The ``reduced-phys-bits`` is used to 5173 provide the number of bits we loose in physical address space. 5174 Similar to C-bit, the value is Host family dependent. On EPYC, 5175 the value should be 5. 5176 5177 The ``sev-device`` provides the device file to use for 5178 communicating with the SEV firmware running inside AMD Secure 5179 Processor. The default device is '/dev/sev'. If hardware 5180 supports memory encryption then /dev/sev devices are created by 5181 CCP driver. 5182 5183 The ``policy`` provides the guest policy to be enforced by the 5184 SEV firmware and restrict what configuration and operational 5185 commands can be performed on this guest by the hypervisor. The 5186 policy should be provided by the guest owner and is bound to the 5187 guest and cannot be changed throughout the lifetime of the 5188 guest. The default is 0. 5189 5190 If guest ``policy`` allows sharing the key with another SEV 5191 guest then ``handle`` can be use to provide handle of the guest 5192 from which to share the key. 5193 5194 The ``dh-cert-file`` and ``session-file`` provides the guest 5195 owner's Public Diffie-Hillman key defined in SEV spec. The PDH 5196 and session parameters are used for establishing a cryptographic 5197 session with the guest owner to negotiate keys used for 5198 attestation. The file must be encoded in base64. 5199 5200 e.g to launch a SEV guest 5201 5202 .. parsed-literal:: 5203 5204 # |qemu_system_x86| \\ 5205 ...... \\ 5206 -object sev-guest,id=sev0,cbitpos=47,reduced-phys-bits=5 \\ 5207 -machine ...,memory-encryption=sev0 \\ 5208 ..... 5209 5210 ``-object authz-simple,id=id,identity=string`` 5211 Create an authorization object that will control access to 5212 network services. 5213 5214 The ``identity`` parameter is identifies the user and its format 5215 depends on the network service that authorization object is 5216 associated with. For authorizing based on TLS x509 certificates, 5217 the identity must be the x509 distinguished name. Note that care 5218 must be taken to escape any commas in the distinguished name. 5219 5220 An example authorization object to validate a x509 distinguished 5221 name would look like: 5222 5223 .. parsed-literal:: 5224 5225 # |qemu_system| \\ 5226 ... \\ 5227 -object 'authz-simple,id=auth0,identity=CN=laptop.example.com,,O=Example Org,,L=London,,ST=London,,C=GB' \\ 5228 ... 5229 5230 Note the use of quotes due to the x509 distinguished name 5231 containing whitespace, and escaping of ','. 5232 5233 ``-object authz-listfile,id=id,filename=path,refresh=on|off`` 5234 Create an authorization object that will control access to 5235 network services. 5236 5237 The ``filename`` parameter is the fully qualified path to a file 5238 containing the access control list rules in JSON format. 5239 5240 An example set of rules that match against SASL usernames might 5241 look like: 5242 5243 :: 5244 5245 { 5246 "rules": [ 5247 { "match": "fred", "policy": "allow", "format": "exact" }, 5248 { "match": "bob", "policy": "allow", "format": "exact" }, 5249 { "match": "danb", "policy": "deny", "format": "glob" }, 5250 { "match": "dan*", "policy": "allow", "format": "exact" }, 5251 ], 5252 "policy": "deny" 5253 } 5254 5255 When checking access the object will iterate over all the rules 5256 and the first rule to match will have its ``policy`` value 5257 returned as the result. If no rules match, then the default 5258 ``policy`` value is returned. 5259 5260 The rules can either be an exact string match, or they can use 5261 the simple UNIX glob pattern matching to allow wildcards to be 5262 used. 5263 5264 If ``refresh`` is set to true the file will be monitored and 5265 automatically reloaded whenever its content changes. 5266 5267 As with the ``authz-simple`` object, the format of the identity 5268 strings being matched depends on the network service, but is 5269 usually a TLS x509 distinguished name, or a SASL username. 5270 5271 An example authorization object to validate a SASL username 5272 would look like: 5273 5274 .. parsed-literal:: 5275 5276 # |qemu_system| \\ 5277 ... \\ 5278 -object authz-simple,id=auth0,filename=/etc/qemu/vnc-sasl.acl,refresh=on \\ 5279 ... 5280 5281 ``-object authz-pam,id=id,service=string`` 5282 Create an authorization object that will control access to 5283 network services. 5284 5285 The ``service`` parameter provides the name of a PAM service to 5286 use for authorization. It requires that a file 5287 ``/etc/pam.d/service`` exist to provide the configuration for 5288 the ``account`` subsystem. 5289 5290 An example authorization object to validate a TLS x509 5291 distinguished name would look like: 5292 5293 .. parsed-literal:: 5294 5295 # |qemu_system| \\ 5296 ... \\ 5297 -object authz-pam,id=auth0,service=qemu-vnc \\ 5298 ... 5299 5300 There would then be a corresponding config file for PAM at 5301 ``/etc/pam.d/qemu-vnc`` that contains: 5302 5303 :: 5304 5305 account requisite pam_listfile.so item=user sense=allow \ 5306 file=/etc/qemu/vnc.allow 5307 5308 Finally the ``/etc/qemu/vnc.allow`` file would contain the list 5309 of x509 distingished names that are permitted access 5310 5311 :: 5312 5313 CN=laptop.example.com,O=Example Home,L=London,ST=London,C=GB 5314 5315 ``-object iothread,id=id,poll-max-ns=poll-max-ns,poll-grow=poll-grow,poll-shrink=poll-shrink,aio-max-batch=aio-max-batch`` 5316 Creates a dedicated event loop thread that devices can be 5317 assigned to. This is known as an IOThread. By default device 5318 emulation happens in vCPU threads or the main event loop thread. 5319 This can become a scalability bottleneck. IOThreads allow device 5320 emulation and I/O to run on other host CPUs. 5321 5322 The ``id`` parameter is a unique ID that will be used to 5323 reference this IOThread from ``-device ...,iothread=id``. 5324 Multiple devices can be assigned to an IOThread. Note that not 5325 all devices support an ``iothread`` parameter. 5326 5327 The ``query-iothreads`` QMP command lists IOThreads and reports 5328 their thread IDs so that the user can configure host CPU 5329 pinning/affinity. 5330 5331 IOThreads use an adaptive polling algorithm to reduce event loop 5332 latency. Instead of entering a blocking system call to monitor 5333 file descriptors and then pay the cost of being woken up when an 5334 event occurs, the polling algorithm spins waiting for events for 5335 a short time. The algorithm's default parameters are suitable 5336 for many cases but can be adjusted based on knowledge of the 5337 workload and/or host device latency. 5338 5339 The ``poll-max-ns`` parameter is the maximum number of 5340 nanoseconds to busy wait for events. Polling can be disabled by 5341 setting this value to 0. 5342 5343 The ``poll-grow`` parameter is the multiplier used to increase 5344 the polling time when the algorithm detects it is missing events 5345 due to not polling long enough. 5346 5347 The ``poll-shrink`` parameter is the divisor used to decrease 5348 the polling time when the algorithm detects it is spending too 5349 long polling without encountering events. 5350 5351 The ``aio-max-batch`` parameter is the maximum number of requests 5352 in a batch for the AIO engine, 0 means that the engine will use 5353 its default. 5354 5355 The IOThread parameters can be modified at run-time using the 5356 ``qom-set`` command (where ``iothread1`` is the IOThread's 5357 ``id``): 5358 5359 :: 5360 5361 (qemu) qom-set /objects/iothread1 poll-max-ns 100000 5362ERST 5363 5364 5365HXCOMM This is the last statement. Insert new options before this line! 5366 5367#undef DEF 5368#undef DEFHEADING 5369#undef ARCHHEADING 5370
