You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

qemu-doc.texi 93KB

  1. \input texinfo @c -*- texinfo -*-
  2. @c %**start of header
  3. @setfilename
  4. @include version.texi
  5. @documentlanguage en
  6. @documentencoding UTF-8
  7. @settitle QEMU version @value{VERSION} User Documentation
  8. @exampleindent 0
  9. @paragraphindent 0
  10. @c %**end of header
  11. @ifinfo
  12. @direntry
  13. * QEMU: (qemu-doc). The QEMU Emulator User Documentation.
  14. @end direntry
  15. @end ifinfo
  16. @iftex
  17. @titlepage
  18. @sp 7
  19. @center @titlefont{QEMU version @value{VERSION}}
  20. @sp 1
  21. @center @titlefont{User Documentation}
  22. @sp 3
  23. @end titlepage
  24. @end iftex
  25. @ifnottex
  26. @node Top
  27. @top
  28. @menu
  29. * Introduction::
  30. * QEMU PC System emulator::
  31. * QEMU System emulator for non PC targets::
  32. * QEMU Guest Agent::
  33. * QEMU User space emulator::
  34. * System requirements::
  35. * Security::
  36. * Implementation notes::
  37. * Deprecated features::
  38. * Supported build platforms::
  39. * License::
  40. * Index::
  41. @end menu
  42. @end ifnottex
  43. @contents
  44. @node Introduction
  45. @chapter Introduction
  46. @menu
  47. * intro_features:: Features
  48. @end menu
  49. @node intro_features
  50. @section Features
  51. QEMU is a FAST! processor emulator using dynamic translation to
  52. achieve good emulation speed.
  53. @cindex operating modes
  54. QEMU has two operating modes:
  55. @itemize
  56. @cindex system emulation
  57. @item Full system emulation. In this mode, QEMU emulates a full system (for
  58. example a PC), including one or several processors and various
  59. peripherals. It can be used to launch different Operating Systems
  60. without rebooting the PC or to debug system code.
  61. @cindex user mode emulation
  62. @item User mode emulation. In this mode, QEMU can launch
  63. processes compiled for one CPU on another CPU. It can be used to
  64. launch the Wine Windows API emulator (@url{}) or
  65. to ease cross-compilation and cross-debugging.
  66. @end itemize
  67. QEMU has the following features:
  68. @itemize
  69. @item QEMU can run without a host kernel driver and yet gives acceptable
  70. performance. It uses dynamic translation to native code for reasonable speed,
  71. with support for self-modifying code and precise exceptions.
  72. @item It is portable to several operating systems (GNU/Linux, *BSD, Mac OS X,
  73. Windows) and architectures.
  74. @item It performs accurate software emulation of the FPU.
  75. @end itemize
  76. QEMU user mode emulation has the following features:
  77. @itemize
  78. @item Generic Linux system call converter, including most ioctls.
  79. @item clone() emulation using native CPU clone() to use Linux scheduler for threads.
  80. @item Accurate signal handling by remapping host signals to target signals.
  81. @end itemize
  82. QEMU full system emulation has the following features:
  83. @itemize
  84. @item
  85. QEMU uses a full software MMU for maximum portability.
  86. @item
  87. QEMU can optionally use an in-kernel accelerator, like kvm. The accelerators
  88. execute most of the guest code natively, while
  89. continuing to emulate the rest of the machine.
  90. @item
  91. Various hardware devices can be emulated and in some cases, host
  92. devices (e.g. serial and parallel ports, USB, drives) can be used
  93. transparently by the guest Operating System. Host device passthrough
  94. can be used for talking to external physical peripherals (e.g. a
  95. webcam, modem or tape drive).
  96. @item
  97. Symmetric multiprocessing (SMP) support. Currently, an in-kernel
  98. accelerator is required to use more than one host CPU for emulation.
  99. @end itemize
  100. @node QEMU PC System emulator
  101. @chapter QEMU PC System emulator
  102. @cindex system emulation (PC)
  103. @menu
  104. * pcsys_introduction:: Introduction
  105. * pcsys_quickstart:: Quick Start
  106. * sec_invocation:: Invocation
  107. * pcsys_keys:: Keys in the graphical frontends
  108. * mux_keys:: Keys in the character backend multiplexer
  109. * pcsys_monitor:: QEMU Monitor
  110. * cpu_models:: CPU models
  111. * disk_images:: Disk Images
  112. * pcsys_network:: Network emulation
  113. * pcsys_other_devs:: Other Devices
  114. * direct_linux_boot:: Direct Linux Boot
  115. * pcsys_usb:: USB emulation
  116. * vnc_security:: VNC security
  117. * network_tls:: TLS setup for network services
  118. * gdb_usage:: GDB usage
  119. * pcsys_os_specific:: Target OS specific information
  120. @end menu
  121. @node pcsys_introduction
  122. @section Introduction
  123. @c man begin DESCRIPTION
  124. The QEMU PC System emulator simulates the
  125. following peripherals:
  126. @itemize @minus
  127. @item
  128. i440FX host PCI bridge and PIIX3 PCI to ISA bridge
  129. @item
  130. Cirrus CLGD 5446 PCI VGA card or dummy VGA card with Bochs VESA
  131. extensions (hardware level, including all non standard modes).
  132. @item
  133. PS/2 mouse and keyboard
  134. @item
  135. 2 PCI IDE interfaces with hard disk and CD-ROM support
  136. @item
  137. Floppy disk
  138. @item
  139. PCI and ISA network adapters
  140. @item
  141. Serial ports
  142. @item
  143. IPMI BMC, either and internal or external one
  144. @item
  145. Creative SoundBlaster 16 sound card
  146. @item
  147. ENSONIQ AudioPCI ES1370 sound card
  148. @item
  149. Intel 82801AA AC97 Audio compatible sound card
  150. @item
  151. Intel HD Audio Controller and HDA codec
  152. @item
  153. Adlib (OPL2) - Yamaha YM3812 compatible chip
  154. @item
  155. Gravis Ultrasound GF1 sound card
  156. @item
  157. CS4231A compatible sound card
  158. @item
  159. PCI UHCI, OHCI, EHCI or XHCI USB controller and a virtual USB-1.1 hub.
  160. @end itemize
  161. SMP is supported with up to 255 CPUs.
  162. QEMU uses the PC BIOS from the Seabios project and the Plex86/Bochs LGPL
  163. VGA BIOS.
  164. QEMU uses YM3812 emulation by Tatsuyuki Satoh.
  165. QEMU uses GUS emulation (GUSEMU32 @url{})
  166. by Tibor "TS" Schütz.
  167. Note that, by default, GUS shares IRQ(7) with parallel ports and so
  168. QEMU must be told to not have parallel ports to have working GUS.
  169. @example
  170. qemu-system-i386 dos.img -soundhw gus -parallel none
  171. @end example
  172. Alternatively:
  173. @example
  174. qemu-system-i386 dos.img -device gus,irq=5
  175. @end example
  176. Or some other unclaimed IRQ.
  177. CS4231A is the chip used in Windows Sound System and GUSMAX products
  178. @c man end
  179. @node pcsys_quickstart
  180. @section Quick Start
  181. @cindex quick start
  182. Download and uncompress the linux image (@file{linux.img}) and type:
  183. @example
  184. qemu-system-i386 linux.img
  185. @end example
  186. Linux should boot and give you a prompt.
  187. @node sec_invocation
  188. @section Invocation
  189. @example
  190. @c man begin SYNOPSIS
  191. @command{qemu-system-i386} [@var{options}] [@var{disk_image}]
  192. @c man end
  193. @end example
  194. @c man begin OPTIONS
  195. @var{disk_image} is a raw hard disk image for IDE hard disk 0. Some
  196. targets do not need a disk image.
  197. @include qemu-options.texi
  198. @c man end
  199. @subsection Device URL Syntax
  200. @c TODO merge this with section Disk Images
  201. @c man begin NOTES
  202. In addition to using normal file images for the emulated storage devices,
  203. QEMU can also use networked resources such as iSCSI devices. These are
  204. specified using a special URL syntax.
  205. @table @option
  206. @item iSCSI
  207. iSCSI support allows QEMU to access iSCSI resources directly and use as
  208. images for the guest storage. Both disk and cdrom images are supported.
  209. Syntax for specifying iSCSI LUNs is
  210. ``iscsi://<target-ip>[:<port>]/<target-iqn>/<lun>''
  211. By default qemu will use the iSCSI initiator-name
  212. '[:<name>]' but this can also be set from the command
  213. line or a configuration file.
  214. Since version Qemu 2.4 it is possible to specify a iSCSI request timeout to detect
  215. stalled requests and force a reestablishment of the session. The timeout
  216. is specified in seconds. The default is 0 which means no timeout. Libiscsi
  217. 1.15.0 or greater is required for this feature.
  218. Example (without authentication):
  219. @example
  220. qemu-system-i386 -iscsi \
  221. -cdrom iscsi:// \
  222. -drive file=iscsi://
  223. @end example
  224. Example (CHAP username/password via URL):
  225. @example
  226. qemu-system-i386 -drive file=iscsi://user%password@@
  227. @end example
  228. Example (CHAP username/password via environment variables):
  229. @example
  231. LIBISCSI_CHAP_PASSWORD="password" \
  232. qemu-system-i386 -drive file=iscsi://
  233. @end example
  234. @item NBD
  235. QEMU supports NBD (Network Block Devices) both using TCP protocol as well
  236. as Unix Domain Sockets.
  237. Syntax for specifying a NBD device using TCP
  238. ``nbd:<server-ip>:<port>[:exportname=<export>]''
  239. Syntax for specifying a NBD device using Unix Domain Sockets
  240. ``nbd:unix:<domain-socket>[:exportname=<export>]''
  241. Example for TCP
  242. @example
  243. qemu-system-i386 --drive file=nbd:
  244. @end example
  245. Example for Unix Domain Sockets
  246. @example
  247. qemu-system-i386 --drive file=nbd:unix:/tmp/nbd-socket
  248. @end example
  249. @item SSH
  250. QEMU supports SSH (Secure Shell) access to remote disks.
  251. Examples:
  252. @example
  253. qemu-system-i386 -drive file=ssh://user@@host/path/to/disk.img
  254. qemu-system-i386 -drive file.driver=ssh,file.user=user,,file.port=22,file.path=/path/to/disk.img
  255. @end example
  256. Currently authentication must be done using ssh-agent. Other
  257. authentication methods may be supported in future.
  258. @item Sheepdog
  259. Sheepdog is a distributed storage system for QEMU.
  260. QEMU supports using either local sheepdog devices or remote networked
  261. devices.
  262. Syntax for specifying a sheepdog device
  263. @example
  264. sheepdog[+tcp|+unix]://[host:port]/vdiname[?socket=path][#snapid|#tag]
  265. @end example
  266. Example
  267. @example
  268. qemu-system-i386 --drive file=sheepdog://
  269. @end example
  270. See also @url{}.
  271. @item GlusterFS
  272. GlusterFS is a user space distributed file system.
  273. QEMU supports the use of GlusterFS volumes for hosting VM disk images using
  274. TCP, Unix Domain Sockets and RDMA transport protocols.
  275. Syntax for specifying a VM disk image on GlusterFS volume is
  276. @example
  277. URI:
  278. gluster[+type]://[host[:port]]/volume/path[?socket=...][,debug=N][,logfile=...]
  279. JSON:
  280. 'json:@{"driver":"qcow2","file":@{"driver":"gluster","volume":"testvol","path":"a.img","debug":N,"logfile":"...",
  281. @ "server":[@{"type":"tcp","host":"...","port":"..."@},
  282. @ @{"type":"unix","socket":"..."@}]@}@}'
  283. @end example
  284. Example
  285. @example
  286. URI:
  287. qemu-system-x86_64 --drive file=gluster://,
  288. @ file.debug=9,file.logfile=/var/log/qemu-gluster.log
  289. JSON:
  290. qemu-system-x86_64 'json:@{"driver":"qcow2",
  291. @ "file":@{"driver":"gluster",
  292. @ "volume":"testvol","path":"a.img",
  293. @ "debug":9,"logfile":"/var/log/qemu-gluster.log",
  294. @ "server":[@{"type":"tcp","host":"","port":24007@},
  295. @ @{"type":"unix","socket":"/var/run/glusterd.socket"@}]@}@}'
  296. qemu-system-x86_64 -drive driver=qcow2,file.driver=gluster,file.volume=testvol,file.path=/path/a.img,
  297. @ file.debug=9,file.logfile=/var/log/qemu-gluster.log,
  298. @ file.server.0.type=tcp,,file.server.0.port=24007,
  299. @ file.server.1.type=unix,file.server.1.socket=/var/run/glusterd.socket
  300. @end example
  301. See also @url{}.
  302. @item HTTP/HTTPS/FTP/FTPS
  303. QEMU supports read-only access to files accessed over http(s) and ftp(s).
  304. Syntax using a single filename:
  305. @example
  306. <protocol>://[<username>[:<password>]@@]<host>/<path>
  307. @end example
  308. where:
  309. @table @option
  310. @item protocol
  311. 'http', 'https', 'ftp', or 'ftps'.
  312. @item username
  313. Optional username for authentication to the remote server.
  314. @item password
  315. Optional password for authentication to the remote server.
  316. @item host
  317. Address of the remote server.
  318. @item path
  319. Path on the remote server, including any query string.
  320. @end table
  321. The following options are also supported:
  322. @table @option
  323. @item url
  324. The full URL when passing options to the driver explicitly.
  325. @item readahead
  326. The amount of data to read ahead with each range request to the remote server.
  327. This value may optionally have the suffix 'T', 'G', 'M', 'K', 'k' or 'b'. If it
  328. does not have a suffix, it will be assumed to be in bytes. The value must be a
  329. multiple of 512 bytes. It defaults to 256k.
  330. @item sslverify
  331. Whether to verify the remote server's certificate when connecting over SSL. It
  332. can have the value 'on' or 'off'. It defaults to 'on'.
  333. @item cookie
  334. Send this cookie (it can also be a list of cookies separated by ';') with
  335. each outgoing request. Only supported when using protocols such as HTTP
  336. which support cookies, otherwise ignored.
  337. @item timeout
  338. Set the timeout in seconds of the CURL connection. This timeout is the time
  339. that CURL waits for a response from the remote server to get the size of the
  340. image to be downloaded. If not set, the default timeout of 5 seconds is used.
  341. @end table
  342. Note that when passing options to qemu explicitly, @option{driver} is the value
  343. of <protocol>.
  344. Example: boot from a remote Fedora 20 live ISO image
  345. @example
  346. qemu-system-x86_64 --drive media=cdrom,file=,readonly
  347. qemu-system-x86_64 --drive media=cdrom,file.driver=http,file.url=,readonly
  348. @end example
  349. Example: boot from a remote Fedora 20 cloud image using a local overlay for
  350. writes, copy-on-read, and a readahead of 64k
  351. @example
  352. qemu-img create -f qcow2 -o backing_file='json:@{"file.driver":"http",, "file.url":"",, "file.readahead":"64k"@}' /tmp/Fedora-x86_64-20-20131211.1-sda.qcow2
  353. qemu-system-x86_64 -drive file=/tmp/Fedora-x86_64-20-20131211.1-sda.qcow2,copy-on-read=on
  354. @end example
  355. Example: boot from an image stored on a VMware vSphere server with a self-signed
  356. certificate using a local overlay for writes, a readahead of 64k and a timeout
  357. of 10 seconds.
  358. @example
  359. qemu-img create -f qcow2 -o backing_file='json:@{"file.driver":"https",, "file.url":"",, "file.sslverify":"off",, "file.readahead":"64k",, "file.timeout":10@}' /tmp/test.qcow2
  360. qemu-system-x86_64 -drive file=/tmp/test.qcow2
  361. @end example
  362. @end table
  363. @c man end
  364. @node pcsys_keys
  365. @section Keys in the graphical frontends
  366. @c man begin OPTIONS
  367. During the graphical emulation, you can use special key combinations to change
  368. modes. The default key mappings are shown below, but if you use @code{-alt-grab}
  369. then the modifier is Ctrl-Alt-Shift (instead of Ctrl-Alt) and if you use
  370. @code{-ctrl-grab} then the modifier is the right Ctrl key (instead of Ctrl-Alt):
  371. @table @key
  372. @item Ctrl-Alt-f
  373. @kindex Ctrl-Alt-f
  374. Toggle full screen
  375. @item Ctrl-Alt-+
  376. @kindex Ctrl-Alt-+
  377. Enlarge the screen
  378. @item Ctrl-Alt--
  379. @kindex Ctrl-Alt--
  380. Shrink the screen
  381. @item Ctrl-Alt-u
  382. @kindex Ctrl-Alt-u
  383. Restore the screen's un-scaled dimensions
  384. @item Ctrl-Alt-n
  385. @kindex Ctrl-Alt-n
  386. Switch to virtual console 'n'. Standard console mappings are:
  387. @table @emph
  388. @item 1
  389. Target system display
  390. @item 2
  391. Monitor
  392. @item 3
  393. Serial port
  394. @end table
  395. @item Ctrl-Alt
  396. @kindex Ctrl-Alt
  397. Toggle mouse and keyboard grab.
  398. @end table
  399. @kindex Ctrl-Up
  400. @kindex Ctrl-Down
  401. @kindex Ctrl-PageUp
  402. @kindex Ctrl-PageDown
  403. In the virtual consoles, you can use @key{Ctrl-Up}, @key{Ctrl-Down},
  404. @key{Ctrl-PageUp} and @key{Ctrl-PageDown} to move in the back log.
  405. @c man end
  406. @node mux_keys
  407. @section Keys in the character backend multiplexer
  408. @c man begin OPTIONS
  409. During emulation, if you are using a character backend multiplexer
  410. (which is the default if you are using @option{-nographic}) then
  411. several commands are available via an escape sequence. These
  412. key sequences all start with an escape character, which is @key{Ctrl-a}
  413. by default, but can be changed with @option{-echr}. The list below assumes
  414. you're using the default.
  415. @table @key
  416. @item Ctrl-a h
  417. @kindex Ctrl-a h
  418. Print this help
  419. @item Ctrl-a x
  420. @kindex Ctrl-a x
  421. Exit emulator
  422. @item Ctrl-a s
  423. @kindex Ctrl-a s
  424. Save disk data back to file (if -snapshot)
  425. @item Ctrl-a t
  426. @kindex Ctrl-a t
  427. Toggle console timestamps
  428. @item Ctrl-a b
  429. @kindex Ctrl-a b
  430. Send break (magic sysrq in Linux)
  431. @item Ctrl-a c
  432. @kindex Ctrl-a c
  433. Rotate between the frontends connected to the multiplexer (usually
  434. this switches between the monitor and the console)
  435. @item Ctrl-a Ctrl-a
  436. @kindex Ctrl-a Ctrl-a
  437. Send the escape character to the frontend
  438. @end table
  439. @c man end
  440. @ignore
  441. @c man begin SEEALSO
  442. The HTML documentation of QEMU for more precise information and Linux
  443. user mode emulator invocation.
  444. @c man end
  445. @c man begin AUTHOR
  446. Fabrice Bellard
  447. @c man end
  448. @end ignore
  449. @node pcsys_monitor
  450. @section QEMU Monitor
  451. @cindex QEMU monitor
  452. The QEMU monitor is used to give complex commands to the QEMU
  453. emulator. You can use it to:
  454. @itemize @minus
  455. @item
  456. Remove or insert removable media images
  457. (such as CD-ROM or floppies).
  458. @item
  459. Freeze/unfreeze the Virtual Machine (VM) and save or restore its state
  460. from a disk file.
  461. @item Inspect the VM state without an external debugger.
  462. @end itemize
  463. @subsection Commands
  464. The following commands are available:
  465. @include qemu-monitor.texi
  466. @include qemu-monitor-info.texi
  467. @subsection Integer expressions
  468. The monitor understands integers expressions for every integer
  469. argument. You can use register names to get the value of specifics
  470. CPU registers by prefixing them with @emph{$}.
  471. @node cpu_models
  472. @section CPU models
  473. @include docs/qemu-cpu-models.texi
  474. @node disk_images
  475. @section Disk Images
  476. QEMU supports many disk image formats, including growable disk images
  477. (their size increase as non empty sectors are written), compressed and
  478. encrypted disk images.
  479. @menu
  480. * disk_images_quickstart:: Quick start for disk image creation
  481. * disk_images_snapshot_mode:: Snapshot mode
  482. * vm_snapshots:: VM snapshots
  483. * qemu_img_invocation:: qemu-img Invocation
  484. * qemu_nbd_invocation:: qemu-nbd Invocation
  485. * disk_images_formats:: Disk image file formats
  486. * host_drives:: Using host drives
  487. * disk_images_fat_images:: Virtual FAT disk images
  488. * disk_images_nbd:: NBD access
  489. * disk_images_sheepdog:: Sheepdog disk images
  490. * disk_images_iscsi:: iSCSI LUNs
  491. * disk_images_gluster:: GlusterFS disk images
  492. * disk_images_ssh:: Secure Shell (ssh) disk images
  493. * disk_images_nvme:: NVMe userspace driver
  494. * disk_image_locking:: Disk image file locking
  495. @end menu
  496. @node disk_images_quickstart
  497. @subsection Quick start for disk image creation
  498. You can create a disk image with the command:
  499. @example
  500. qemu-img create myimage.img mysize
  501. @end example
  502. where @var{myimage.img} is the disk image filename and @var{mysize} is its
  503. size in kilobytes. You can add an @code{M} suffix to give the size in
  504. megabytes and a @code{G} suffix for gigabytes.
  505. See @ref{qemu_img_invocation} for more information.
  506. @node disk_images_snapshot_mode
  507. @subsection Snapshot mode
  508. If you use the option @option{-snapshot}, all disk images are
  509. considered as read only. When sectors in written, they are written in
  510. a temporary file created in @file{/tmp}. You can however force the
  511. write back to the raw disk images by using the @code{commit} monitor
  512. command (or @key{C-a s} in the serial console).
  513. @node vm_snapshots
  514. @subsection VM snapshots
  515. VM snapshots are snapshots of the complete virtual machine including
  516. CPU state, RAM, device state and the content of all the writable
  517. disks. In order to use VM snapshots, you must have at least one non
  518. removable and writable block device using the @code{qcow2} disk image
  519. format. Normally this device is the first virtual hard drive.
  520. Use the monitor command @code{savevm} to create a new VM snapshot or
  521. replace an existing one. A human readable name can be assigned to each
  522. snapshot in addition to its numerical ID.
  523. Use @code{loadvm} to restore a VM snapshot and @code{delvm} to remove
  524. a VM snapshot. @code{info snapshots} lists the available snapshots
  525. with their associated information:
  526. @example
  527. (qemu) info snapshots
  528. Snapshot devices: hda
  529. Snapshot list (from hda):
  531. 1 start 41M 2006-08-06 12:38:02 00:00:14.954
  532. 2 40M 2006-08-06 12:43:29 00:00:18.633
  533. 3 msys 40M 2006-08-06 12:44:04 00:00:23.514
  534. @end example
  535. A VM snapshot is made of a VM state info (its size is shown in
  536. @code{info snapshots}) and a snapshot of every writable disk image.
  537. The VM state info is stored in the first @code{qcow2} non removable
  538. and writable block device. The disk image snapshots are stored in
  539. every disk image. The size of a snapshot in a disk image is difficult
  540. to evaluate and is not shown by @code{info snapshots} because the
  541. associated disk sectors are shared among all the snapshots to save
  542. disk space (otherwise each snapshot would need a full copy of all the
  543. disk images).
  544. When using the (unrelated) @code{-snapshot} option
  545. (@ref{disk_images_snapshot_mode}), you can always make VM snapshots,
  546. but they are deleted as soon as you exit QEMU.
  547. VM snapshots currently have the following known limitations:
  548. @itemize
  549. @item
  550. They cannot cope with removable devices if they are removed or
  551. inserted after a snapshot is done.
  552. @item
  553. A few device drivers still have incomplete snapshot support so their
  554. state is not saved or restored properly (in particular USB).
  555. @end itemize
  556. @node qemu_img_invocation
  557. @subsection @code{qemu-img} Invocation
  558. @include qemu-img.texi
  559. @node qemu_nbd_invocation
  560. @subsection @code{qemu-nbd} Invocation
  561. @include qemu-nbd.texi
  562. @include docs/qemu-block-drivers.texi
  563. @node pcsys_network
  564. @section Network emulation
  565. QEMU can simulate several network cards (e.g. PCI or ISA cards on the PC
  566. target) and can connect them to a network backend on the host or an emulated
  567. hub. The various host network backends can either be used to connect the NIC of
  568. the guest to a real network (e.g. by using a TAP devices or the non-privileged
  569. user mode network stack), or to other guest instances running in another QEMU
  570. process (e.g. by using the socket host network backend).
  571. @subsection Using TAP network interfaces
  572. This is the standard way to connect QEMU to a real network. QEMU adds
  573. a virtual network device on your host (called @code{tapN}), and you
  574. can then configure it as if it was a real ethernet card.
  575. @subsubsection Linux host
  576. As an example, you can download the @file{linux-test-xxx.tar.gz}
  577. archive and copy the script @file{qemu-ifup} in @file{/etc} and
  578. configure properly @code{sudo} so that the command @code{ifconfig}
  579. contained in @file{qemu-ifup} can be executed as root. You must verify
  580. that your host kernel supports the TAP network interfaces: the
  581. device @file{/dev/net/tun} must be present.
  582. See @ref{sec_invocation} to have examples of command lines using the
  583. TAP network interfaces.
  584. @subsubsection Windows host
  585. There is a virtual ethernet driver for Windows 2000/XP systems, called
  586. TAP-Win32. But it is not included in standard QEMU for Windows,
  587. so you will need to get it separately. It is part of OpenVPN package,
  588. so download OpenVPN from : @url{}.
  589. @subsection Using the user mode network stack
  590. By using the option @option{-net user} (default configuration if no
  591. @option{-net} option is specified), QEMU uses a completely user mode
  592. network stack (you don't need root privilege to use the virtual
  593. network). The virtual network configuration is the following:
  594. @example
  595. guest ( <------> Firewall/DHCP server <-----> Internet
  596. | (
  597. |
  598. ----> DNS server (
  599. |
  600. ----> SMB server (
  601. @end example
  602. The QEMU VM behaves as if it was behind a firewall which blocks all
  603. incoming connections. You can use a DHCP client to automatically
  604. configure the network in the QEMU VM. The DHCP server assign addresses
  605. to the hosts starting from
  606. In order to check that the user mode network is working, you can ping
  607. the address and verify that you got an address in the range
  608. 10.0.2.x from the QEMU virtual DHCP server.
  609. Note that ICMP traffic in general does not work with user mode networking.
  610. @code{ping}, aka. ICMP echo, to the local router ( shall work,
  611. however. If you're using QEMU on Linux >= 3.0, it can use unprivileged ICMP
  612. ping sockets to allow @code{ping} to the Internet. The host admin has to set
  613. the ping_group_range in order to grant access to those sockets. To allow ping
  614. for GID 100 (usually users group):
  615. @example
  616. echo 100 100 > /proc/sys/net/ipv4/ping_group_range
  617. @end example
  618. When using the built-in TFTP server, the router is also the TFTP
  619. server.
  620. When using the @option{'-netdev user,hostfwd=...'} option, TCP or UDP
  621. connections can be redirected from the host to the guest. It allows for
  622. example to redirect X11, telnet or SSH connections.
  623. @subsection Hubs
  624. QEMU can simulate several hubs. A hub can be thought of as a virtual connection
  625. between several network devices. These devices can be for example QEMU virtual
  626. ethernet cards or virtual Host ethernet devices (TAP devices). You can connect
  627. guest NICs or host network backends to such a hub using the @option{-netdev
  628. hubport} or @option{-nic hubport} options. The legacy @option{-net} option
  629. also connects the given device to the emulated hub with ID 0 (i.e. the default
  630. hub) unless you specify a netdev with @option{-net nic,netdev=xxx} here.
  631. @subsection Connecting emulated networks between QEMU instances
  632. Using the @option{-netdev socket} (or @option{-nic socket} or
  633. @option{-net socket}) option, it is possible to create emulated
  634. networks that span several QEMU instances.
  635. See the description of the @option{-netdev socket} option in the
  636. @ref{sec_invocation,,Invocation chapter} to have a basic example.
  637. @node pcsys_other_devs
  638. @section Other Devices
  639. @subsection Inter-VM Shared Memory device
  640. On Linux hosts, a shared memory device is available. The basic syntax
  641. is:
  642. @example
  643. qemu-system-x86_64 -device ivshmem-plain,memdev=@var{hostmem}
  644. @end example
  645. where @var{hostmem} names a host memory backend. For a POSIX shared
  646. memory backend, use something like
  647. @example
  648. -object memory-backend-file,size=1M,share,mem-path=/dev/shm/ivshmem,id=@var{hostmem}
  649. @end example
  650. If desired, interrupts can be sent between guest VMs accessing the same shared
  651. memory region. Interrupt support requires using a shared memory server and
  652. using a chardev socket to connect to it. The code for the shared memory server
  653. is qemu.git/contrib/ivshmem-server. An example syntax when using the shared
  654. memory server is:
  655. @example
  656. # First start the ivshmem server once and for all
  657. ivshmem-server -p @var{pidfile} -S @var{path} -m @var{shm-name} -l @var{shm-size} -n @var{vectors}
  658. # Then start your qemu instances with matching arguments
  659. qemu-system-x86_64 -device ivshmem-doorbell,vectors=@var{vectors},chardev=@var{id}
  660. -chardev socket,path=@var{path},id=@var{id}
  661. @end example
  662. When using the server, the guest will be assigned a VM ID (>=0) that allows guests
  663. using the same server to communicate via interrupts. Guests can read their
  664. VM ID from a device register (see ivshmem-spec.txt).
  665. @subsubsection Migration with ivshmem
  666. With device property @option{master=on}, the guest will copy the shared
  667. memory on migration to the destination host. With @option{master=off},
  668. the guest will not be able to migrate with the device attached. In the
  669. latter case, the device should be detached and then reattached after
  670. migration using the PCI hotplug support.
  671. At most one of the devices sharing the same memory can be master. The
  672. master must complete migration before you plug back the other devices.
  673. @subsubsection ivshmem and hugepages
  674. Instead of specifying the <shm size> using POSIX shm, you may specify
  675. a memory backend that has hugepage support:
  676. @example
  677. qemu-system-x86_64 -object memory-backend-file,size=1G,mem-path=/dev/hugepages/my-shmem-file,share,id=mb1
  678. -device ivshmem-plain,memdev=mb1
  679. @end example
  680. ivshmem-server also supports hugepages mount points with the
  681. @option{-m} memory path argument.
  682. @node direct_linux_boot
  683. @section Direct Linux Boot
  684. This section explains how to launch a Linux kernel inside QEMU without
  685. having to make a full bootable image. It is very useful for fast Linux
  686. kernel testing.
  687. The syntax is:
  688. @example
  689. qemu-system-i386 -kernel arch/i386/boot/bzImage -hda root-2.4.20.img -append "root=/dev/hda"
  690. @end example
  691. Use @option{-kernel} to provide the Linux kernel image and
  692. @option{-append} to give the kernel command line arguments. The
  693. @option{-initrd} option can be used to provide an INITRD image.
  694. When using the direct Linux boot, a disk image for the first hard disk
  695. @file{hda} is required because its boot sector is used to launch the
  696. Linux kernel.
  697. If you do not need graphical output, you can disable it and redirect
  698. the virtual serial port and the QEMU monitor to the console with the
  699. @option{-nographic} option. The typical command line is:
  700. @example
  701. qemu-system-i386 -kernel arch/i386/boot/bzImage -hda root-2.4.20.img \
  702. -append "root=/dev/hda console=ttyS0" -nographic
  703. @end example
  704. Use @key{Ctrl-a c} to switch between the serial console and the
  705. monitor (@pxref{pcsys_keys}).
  706. @node pcsys_usb
  707. @section USB emulation
  708. QEMU can emulate a PCI UHCI, OHCI, EHCI or XHCI USB controller. You can
  709. plug virtual USB devices or real host USB devices (only works with certain
  710. host operating systems). QEMU will automatically create and connect virtual
  711. USB hubs as necessary to connect multiple USB devices.
  712. @menu
  713. * usb_devices::
  714. * host_usb_devices::
  715. @end menu
  716. @node usb_devices
  717. @subsection Connecting USB devices
  718. USB devices can be connected with the @option{-device usb-...} command line
  719. option or the @code{device_add} monitor command. Available devices are:
  720. @table @code
  721. @item usb-mouse
  722. Virtual Mouse. This will override the PS/2 mouse emulation when activated.
  723. @item usb-tablet
  724. Pointer device that uses absolute coordinates (like a touchscreen).
  725. This means QEMU is able to report the mouse position without having
  726. to grab the mouse. Also overrides the PS/2 mouse emulation when activated.
  727. @item usb-storage,drive=@var{drive_id}
  728. Mass storage device backed by @var{drive_id} (@pxref{disk_images})
  729. @item usb-uas
  730. USB attached SCSI device, see
  731. @url{;a=blob_plain;f=docs/usb-storage.txt,usb-storage.txt}
  732. for details
  733. @item usb-bot
  734. Bulk-only transport storage device, see
  735. @url{;a=blob_plain;f=docs/usb-storage.txt,usb-storage.txt}
  736. for details here, too
  737. @item usb-mtp,rootdir=@var{dir}
  738. Media transfer protocol device, using @var{dir} as root of the file tree
  739. that is presented to the guest.
  740. @item usb-host,hostbus=@var{bus},hostaddr=@var{addr}
  741. Pass through the host device identified by @var{bus} and @var{addr}
  742. @item usb-host,vendorid=@var{vendor},productid=@var{product}
  743. Pass through the host device identified by @var{vendor} and @var{product} ID
  744. @item usb-wacom-tablet
  745. Virtual Wacom PenPartner tablet. This device is similar to the @code{tablet}
  746. above but it can be used with the tslib library because in addition to touch
  747. coordinates it reports touch pressure.
  748. @item usb-kbd
  749. Standard USB keyboard. Will override the PS/2 keyboard (if present).
  750. @item usb-serial,chardev=@var{id}
  751. Serial converter. This emulates an FTDI FT232BM chip connected to host character
  752. device @var{id}.
  753. @item usb-braille,chardev=@var{id}
  754. Braille device. This will use BrlAPI to display the braille output on a real
  755. or fake device referenced by @var{id}.
  756. @item usb-net[,netdev=@var{id}]
  757. Network adapter that supports CDC ethernet and RNDIS protocols. @var{id}
  758. specifies a netdev defined with @code{-netdev @dots{},id=@var{id}}.
  759. For instance, user-mode networking can be used with
  760. @example
  761. qemu-system-i386 [...] -netdev user,id=net0 -device usb-net,netdev=net0
  762. @end example
  763. @item usb-ccid
  764. Smartcard reader device
  765. @item usb-audio
  766. USB audio device
  767. @item usb-bt-dongle
  768. Bluetooth dongle for the transport layer of HCI. It is connected to HCI
  769. scatternet 0 by default (corresponds to @code{-bt hci,vlan=0}).
  770. Note that the syntax for the @code{-device usb-bt-dongle} option is not as
  771. useful yet as it was with the legacy @code{-usbdevice} option. So to
  772. configure an USB bluetooth device, you might need to use
  773. "@code{-usbdevice bt}[:@var{hci-type}]" instead. This configures a
  774. bluetooth dongle whose type is specified in the same format as with
  775. the @option{-bt hci} option, @pxref{bt-hcis,,allowed HCI types}. If
  776. no type is given, the HCI logic corresponds to @code{-bt hci,vlan=0}.
  777. This USB device implements the USB Transport Layer of HCI. Example
  778. usage:
  779. @example
  780. @command{qemu-system-i386} [...@var{OPTIONS}...] @option{-usbdevice} bt:hci,vlan=3 @option{-bt} device:keyboard,vlan=3
  781. @end example
  782. @end table
  783. @node host_usb_devices
  784. @subsection Using host USB devices on a Linux host
  785. WARNING: this is an experimental feature. QEMU will slow down when
  786. using it. USB devices requiring real time streaming (i.e. USB Video
  787. Cameras) are not supported yet.
  788. @enumerate
  789. @item If you use an early Linux 2.4 kernel, verify that no Linux driver
  790. is actually using the USB device. A simple way to do that is simply to
  791. disable the corresponding kernel module by renaming it from @file{mydriver.o}
  792. to @file{mydriver.o.disabled}.
  793. @item Verify that @file{/proc/bus/usb} is working (most Linux distributions should enable it by default). You should see something like that:
  794. @example
  795. ls /proc/bus/usb
  796. 001 devices drivers
  797. @end example
  798. @item Since only root can access to the USB devices directly, you can either launch QEMU as root or change the permissions of the USB devices you want to use. For testing, the following suffices:
  799. @example
  800. chown -R myuid /proc/bus/usb
  801. @end example
  802. @item Launch QEMU and do in the monitor:
  803. @example
  804. info usbhost
  805. Device 1.2, speed 480 Mb/s
  806. Class 00: USB device 1234:5678, USB DISK
  807. @end example
  808. You should see the list of the devices you can use (Never try to use
  809. hubs, it won't work).
  810. @item Add the device in QEMU by using:
  811. @example
  812. device_add usb-host,vendorid=0x1234,productid=0x5678
  813. @end example
  814. Normally the guest OS should report that a new USB device is plugged.
  815. You can use the option @option{-device usb-host,...} to do the same.
  816. @item Now you can try to use the host USB device in QEMU.
  817. @end enumerate
  818. When relaunching QEMU, you may have to unplug and plug again the USB
  819. device to make it work again (this is a bug).
  820. @node vnc_security
  821. @section VNC security
  822. The VNC server capability provides access to the graphical console
  823. of the guest VM across the network. This has a number of security
  824. considerations depending on the deployment scenarios.
  825. @menu
  826. * vnc_sec_none::
  827. * vnc_sec_password::
  828. * vnc_sec_certificate::
  829. * vnc_sec_certificate_verify::
  830. * vnc_sec_certificate_pw::
  831. * vnc_sec_sasl::
  832. * vnc_sec_certificate_sasl::
  833. * vnc_setup_sasl::
  834. @end menu
  835. @node vnc_sec_none
  836. @subsection Without passwords
  837. The simplest VNC server setup does not include any form of authentication.
  838. For this setup it is recommended to restrict it to listen on a UNIX domain
  839. socket only. For example
  840. @example
  841. qemu-system-i386 [...OPTIONS...] -vnc unix:/home/joebloggs/.qemu-myvm-vnc
  842. @end example
  843. This ensures that only users on local box with read/write access to that
  844. path can access the VNC server. To securely access the VNC server from a
  845. remote machine, a combination of netcat+ssh can be used to provide a secure
  846. tunnel.
  847. @node vnc_sec_password
  848. @subsection With passwords
  849. The VNC protocol has limited support for password based authentication. Since
  850. the protocol limits passwords to 8 characters it should not be considered
  851. to provide high security. The password can be fairly easily brute-forced by
  852. a client making repeat connections. For this reason, a VNC server using password
  853. authentication should be restricted to only listen on the loopback interface
  854. or UNIX domain sockets. Password authentication is not supported when operating
  855. in FIPS 140-2 compliance mode as it requires the use of the DES cipher. Password
  856. authentication is requested with the @code{password} option, and then once QEMU
  857. is running the password is set with the monitor. Until the monitor is used to
  858. set the password all clients will be rejected.
  859. @example
  860. qemu-system-i386 [...OPTIONS...] -vnc :1,password -monitor stdio
  861. (qemu) change vnc password
  862. Password: ********
  863. (qemu)
  864. @end example
  865. @node vnc_sec_certificate
  866. @subsection With x509 certificates
  867. The QEMU VNC server also implements the VeNCrypt extension allowing use of
  868. TLS for encryption of the session, and x509 certificates for authentication.
  869. The use of x509 certificates is strongly recommended, because TLS on its
  870. own is susceptible to man-in-the-middle attacks. Basic x509 certificate
  871. support provides a secure session, but no authentication. This allows any
  872. client to connect, and provides an encrypted session.
  873. @example
  874. qemu-system-i386 [...OPTIONS...] \
  875. -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=server,verify-peer=no \
  876. -vnc :1,tls-creds=tls0 -monitor stdio
  877. @end example
  878. In the above example @code{/etc/pki/qemu} should contain at least three files,
  879. @code{ca-cert.pem}, @code{server-cert.pem} and @code{server-key.pem}. Unprivileged
  880. users will want to use a private directory, for example @code{$HOME/.pki/qemu}.
  881. NB the @code{server-key.pem} file should be protected with file mode 0600 to
  882. only be readable by the user owning it.
  883. @node vnc_sec_certificate_verify
  884. @subsection With x509 certificates and client verification
  885. Certificates can also provide a means to authenticate the client connecting.
  886. The server will request that the client provide a certificate, which it will
  887. then validate against the CA certificate. This is a good choice if deploying
  888. in an environment with a private internal certificate authority. It uses the
  889. same syntax as previously, but with @code{verify-peer} set to @code{yes}
  890. instead.
  891. @example
  892. qemu-system-i386 [...OPTIONS...] \
  893. -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=server,verify-peer=yes \
  894. -vnc :1,tls-creds=tls0 -monitor stdio
  895. @end example
  896. @node vnc_sec_certificate_pw
  897. @subsection With x509 certificates, client verification and passwords
  898. Finally, the previous method can be combined with VNC password authentication
  899. to provide two layers of authentication for clients.
  900. @example
  901. qemu-system-i386 [...OPTIONS...] \
  902. -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=server,verify-peer=yes \
  903. -vnc :1,tls-creds=tls0,password -monitor stdio
  904. (qemu) change vnc password
  905. Password: ********
  906. (qemu)
  907. @end example
  908. @node vnc_sec_sasl
  909. @subsection With SASL authentication
  910. The SASL authentication method is a VNC extension, that provides an
  911. easily extendable, pluggable authentication method. This allows for
  912. integration with a wide range of authentication mechanisms, such as
  913. PAM, GSSAPI/Kerberos, LDAP, SQL databases, one-time keys and more.
  914. The strength of the authentication depends on the exact mechanism
  915. configured. If the chosen mechanism also provides a SSF layer, then
  916. it will encrypt the datastream as well.
  917. Refer to the later docs on how to choose the exact SASL mechanism
  918. used for authentication, but assuming use of one supporting SSF,
  919. then QEMU can be launched with:
  920. @example
  921. qemu-system-i386 [...OPTIONS...] -vnc :1,sasl -monitor stdio
  922. @end example
  923. @node vnc_sec_certificate_sasl
  924. @subsection With x509 certificates and SASL authentication
  925. If the desired SASL authentication mechanism does not supported
  926. SSF layers, then it is strongly advised to run it in combination
  927. with TLS and x509 certificates. This provides securely encrypted
  928. data stream, avoiding risk of compromising of the security
  929. credentials. This can be enabled, by combining the 'sasl' option
  930. with the aforementioned TLS + x509 options:
  931. @example
  932. qemu-system-i386 [...OPTIONS...] \
  933. -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=server,verify-peer=yes \
  934. -vnc :1,tls-creds=tls0,sasl -monitor stdio
  935. @end example
  936. @node vnc_setup_sasl
  937. @subsection Configuring SASL mechanisms
  938. The following documentation assumes use of the Cyrus SASL implementation on a
  939. Linux host, but the principles should apply to any other SASL implementation
  940. or host. When SASL is enabled, the mechanism configuration will be loaded from
  941. system default SASL service config /etc/sasl2/qemu.conf. If running QEMU as an
  942. unprivileged user, an environment variable SASL_CONF_PATH can be used to make
  943. it search alternate locations for the service config file.
  944. If the TLS option is enabled for VNC, then it will provide session encryption,
  945. otherwise the SASL mechanism will have to provide encryption. In the latter
  946. case the list of possible plugins that can be used is drastically reduced. In
  947. fact only the GSSAPI SASL mechanism provides an acceptable level of security
  948. by modern standards. Previous versions of QEMU referred to the DIGEST-MD5
  949. mechanism, however, it has multiple serious flaws described in detail in
  950. RFC 6331 and thus should never be used any more. The SCRAM-SHA-1 mechanism
  951. provides a simple username/password auth facility similar to DIGEST-MD5, but
  952. does not support session encryption, so can only be used in combination with
  953. TLS.
  954. When not using TLS the recommended configuration is
  955. @example
  956. mech_list: gssapi
  957. keytab: /etc/qemu/
  958. @end example
  959. This says to use the 'GSSAPI' mechanism with the Kerberos v5 protocol, with
  960. the server principal stored in /etc/qemu/ For this to work the
  961. administrator of your KDC must generate a Kerberos principal for the server,
  962. with a name of 'qemu/' replacing
  963. '' with the fully qualified host name of the machine
  964. running QEMU, and 'EXAMPLE.COM' with the Kerberos Realm.
  965. When using TLS, if username+password authentication is desired, then a
  966. reasonable configuration is
  967. @example
  968. mech_list: scram-sha-1
  969. sasldb_path: /etc/qemu/passwd.db
  970. @end example
  971. The @code{saslpasswd2} program can be used to populate the @code{passwd.db}
  972. file with accounts.
  973. Other SASL configurations will be left as an exercise for the reader. Note that
  974. all mechanisms, except GSSAPI, should be combined with use of TLS to ensure a
  975. secure data channel.
  976. @node network_tls
  977. @section TLS setup for network services
  978. Almost all network services in QEMU have the ability to use TLS for
  979. session data encryption, along with x509 certificates for simple
  980. client authentication. What follows is a description of how to
  981. generate certificates suitable for usage with QEMU, and applies to
  982. the VNC server, character devices with the TCP backend, NBD server
  983. and client, and migration server and client.
  984. At a high level, QEMU requires certificates and private keys to be
  985. provided in PEM format. Aside from the core fields, the certificates
  986. should include various extension data sets, including v3 basic
  987. constraints data, key purpose, key usage and subject alt name.
  988. The GnuTLS package includes a command called @code{certtool} which can
  989. be used to easily generate certificates and keys in the required format
  990. with expected data present. Alternatively a certificate management
  991. service may be used.
  992. At a minimum it is necessary to setup a certificate authority, and
  993. issue certificates to each server. If using x509 certificates for
  994. authentication, then each client will also need to be issued a
  995. certificate.
  996. Assuming that the QEMU network services will only ever be exposed to
  997. clients on a private intranet, there is no need to use a commercial
  998. certificate authority to create certificates. A self-signed CA is
  999. sufficient, and in fact likely to be more secure since it removes
  1000. the ability of malicious 3rd parties to trick the CA into mis-issuing
  1001. certs for impersonating your services. The only likely exception
  1002. where a commercial CA might be desirable is if enabling the VNC
  1003. websockets server and exposing it directly to remote browser clients.
  1004. In such a case it might be useful to use a commercial CA to avoid
  1005. needing to install custom CA certs in the web browsers.
  1006. The recommendation is for the server to keep its certificates in either
  1007. @code{/etc/pki/qemu} or for unprivileged users in @code{$HOME/.pki/qemu}.
  1008. @menu
  1009. * tls_generate_ca::
  1010. * tls_generate_server::
  1011. * tls_generate_client::
  1012. * tls_creds_setup::
  1013. * tls_psk::
  1014. @end menu
  1015. @node tls_generate_ca
  1016. @subsection Setup the Certificate Authority
  1017. This step only needs to be performed once per organization / organizational
  1018. unit. First the CA needs a private key. This key must be kept VERY secret
  1019. and secure. If this key is compromised the entire trust chain of the certificates
  1020. issued with it is lost.
  1021. @example
  1022. # certtool --generate-privkey > ca-key.pem
  1023. @end example
  1024. To generate a self-signed certificate requires one core piece of information,
  1025. the name of the organization. A template file @code{} should be
  1026. populated with the desired data to avoid having to deal with interactive
  1027. prompts from certtool:
  1028. @example
  1029. # cat > <<EOF
  1030. cn = Name of your organization
  1031. ca
  1032. cert_signing_key
  1033. EOF
  1034. # certtool --generate-self-signed \
  1035. --load-privkey ca-key.pem
  1036. --template \
  1037. --outfile ca-cert.pem
  1038. @end example
  1039. The @code{ca} keyword in the template sets the v3 basic constraints extension
  1040. to indicate this certificate is for a CA, while @code{cert_signing_key} sets
  1041. the key usage extension to indicate this will be used for signing other keys.
  1042. The generated @code{ca-cert.pem} file should be copied to all servers and
  1043. clients wishing to utilize TLS support in the VNC server. The @code{ca-key.pem}
  1044. must not be disclosed/copied anywhere except the host responsible for issuing
  1045. certificates.
  1046. @node tls_generate_server
  1047. @subsection Issuing server certificates
  1048. Each server (or host) needs to be issued with a key and certificate. When connecting
  1049. the certificate is sent to the client which validates it against the CA certificate.
  1050. The core pieces of information for a server certificate are the hostnames and/or IP
  1051. addresses that will be used by clients when connecting. The hostname / IP address
  1052. that the client specifies when connecting will be validated against the hostname(s)
  1053. and IP address(es) recorded in the server certificate, and if no match is found
  1054. the client will close the connection.
  1055. Thus it is recommended that the server certificate include both the fully qualified
  1056. and unqualified hostnames. If the server will have permanently assigned IP address(es),
  1057. and clients are likely to use them when connecting, they may also be included in the
  1058. certificate. Both IPv4 and IPv6 addresses are supported. Historically certificates
  1059. only included 1 hostname in the @code{CN} field, however, usage of this field for
  1060. validation is now deprecated. Instead modern TLS clients will validate against the
  1061. Subject Alt Name extension data, which allows for multiple entries. In the future
  1062. usage of the @code{CN} field may be discontinued entirely, so providing SAN
  1063. extension data is strongly recommended.
  1064. On the host holding the CA, create template files containing the information
  1065. for each server, and use it to issue server certificates.
  1066. @example
  1067. # cat > <<EOF
  1068. organization = Name of your organization
  1069. cn =
  1070. dns_name = hostNNN
  1071. dns_name =
  1072. ip_address =
  1073. ip_address =
  1074. ip_address = 2620:0:cafe::87
  1075. ip_address = 2001:24::92
  1076. tls_www_server
  1077. encryption_key
  1078. signing_key
  1079. EOF
  1080. # certtool --generate-privkey > server-hostNNN-key.pem
  1081. # certtool --generate-certificate \
  1082. --load-ca-certificate ca-cert.pem \
  1083. --load-ca-privkey ca-key.pem \
  1084. --load-privkey server-hostNNN-key.pem \
  1085. --template \
  1086. --outfile server-hostNNN-cert.pem
  1087. @end example
  1088. The @code{dns_name} and @code{ip_address} fields in the template are setting
  1089. the subject alt name extension data. The @code{tls_www_server} keyword is the
  1090. key purpose extension to indicate this certificate is intended for usage in
  1091. a web server. Although QEMU network services are not in fact HTTP servers
  1092. (except for VNC websockets), setting this key purpose is still recommended.
  1093. The @code{encryption_key} and @code{signing_key} keyword is the key usage
  1094. extension to indicate this certificate is intended for usage in the data
  1095. session.
  1096. The @code{server-hostNNN-key.pem} and @code{server-hostNNN-cert.pem} files
  1097. should now be securely copied to the server for which they were generated,
  1098. and renamed to @code{server-key.pem} and @code{server-cert.pem} when added
  1099. to the @code{/etc/pki/qemu} directory on the target host. The @code{server-key.pem}
  1100. file is security sensitive and should be kept protected with file mode 0600
  1101. to prevent disclosure.
  1102. @node tls_generate_client
  1103. @subsection Issuing client certificates
  1104. The QEMU x509 TLS credential setup defaults to enabling client verification
  1105. using certificates, providing a simple authentication mechanism. If this
  1106. default is used, each client also needs to be issued a certificate. The client
  1107. certificate contains enough metadata to uniquely identify the client with the
  1108. scope of the certificate authority. The client certificate would typically
  1109. include fields for organization, state, city, building, etc.
  1110. Once again on the host holding the CA, create template files containing the
  1111. information for each client, and use it to issue client certificates.
  1112. @example
  1113. # cat > <<EOF
  1114. country = GB
  1115. state = London
  1116. locality = City Of London
  1117. organization = Name of your organization
  1118. cn =
  1119. tls_www_client
  1120. encryption_key
  1121. signing_key
  1122. EOF
  1123. # certtool --generate-privkey > client-hostNNN-key.pem
  1124. # certtool --generate-certificate \
  1125. --load-ca-certificate ca-cert.pem \
  1126. --load-ca-privkey ca-key.pem \
  1127. --load-privkey client-hostNNN-key.pem \
  1128. --template \
  1129. --outfile client-hostNNN-cert.pem
  1130. @end example
  1131. The subject alt name extension data is not required for clients, so the
  1132. the @code{dns_name} and @code{ip_address} fields are not included.
  1133. The @code{tls_www_client} keyword is the key purpose extension to indicate
  1134. this certificate is intended for usage in a web client. Although QEMU
  1135. network clients are not in fact HTTP clients, setting this key purpose is
  1136. still recommended. The @code{encryption_key} and @code{signing_key} keyword
  1137. is the key usage extension to indicate this certificate is intended for
  1138. usage in the data session.
  1139. The @code{client-hostNNN-key.pem} and @code{client-hostNNN-cert.pem} files
  1140. should now be securely copied to the client for which they were generated,
  1141. and renamed to @code{client-key.pem} and @code{client-cert.pem} when added
  1142. to the @code{/etc/pki/qemu} directory on the target host. The @code{client-key.pem}
  1143. file is security sensitive and should be kept protected with file mode 0600
  1144. to prevent disclosure.
  1145. If a single host is going to be using TLS in both a client and server
  1146. role, it is possible to create a single certificate to cover both roles.
  1147. This would be quite common for the migration and NBD services, where a
  1148. QEMU process will be started by accepting a TLS protected incoming migration,
  1149. and later itself be migrated out to another host. To generate a single
  1150. certificate, simply include the template data from both the client and server
  1151. instructions in one.
  1152. @example
  1153. # cat > <<EOF
  1154. country = GB
  1155. state = London
  1156. locality = City Of London
  1157. organization = Name of your organization
  1158. cn =
  1159. dns_name = hostNNN
  1160. dns_name =
  1161. ip_address =
  1162. ip_address =
  1163. ip_address = 2620:0:cafe::87
  1164. ip_address = 2001:24::92
  1165. tls_www_server
  1166. tls_www_client
  1167. encryption_key
  1168. signing_key
  1169. EOF
  1170. # certtool --generate-privkey > both-hostNNN-key.pem
  1171. # certtool --generate-certificate \
  1172. --load-ca-certificate ca-cert.pem \
  1173. --load-ca-privkey ca-key.pem \
  1174. --load-privkey both-hostNNN-key.pem \
  1175. --template \
  1176. --outfile both-hostNNN-cert.pem
  1177. @end example
  1178. When copying the PEM files to the target host, save them twice,
  1179. once as @code{server-cert.pem} and @code{server-key.pem}, and
  1180. again as @code{client-cert.pem} and @code{client-key.pem}.
  1181. @node tls_creds_setup
  1182. @subsection TLS x509 credential configuration
  1183. QEMU has a standard mechanism for loading x509 credentials that will be
  1184. used for network services and clients. It requires specifying the
  1185. @code{tls-creds-x509} class name to the @code{--object} command line
  1186. argument for the system emulators. Each set of credentials loaded should
  1187. be given a unique string identifier via the @code{id} parameter. A single
  1188. set of TLS credentials can be used for multiple network backends, so VNC,
  1189. migration, NBD, character devices can all share the same credentials. Note,
  1190. however, that credentials for use in a client endpoint must be loaded
  1191. separately from those used in a server endpoint.
  1192. When specifying the object, the @code{dir} parameters specifies which
  1193. directory contains the credential files. This directory is expected to
  1194. contain files with the names mentioned previously, @code{ca-cert.pem},
  1195. @code{server-key.pem}, @code{server-cert.pem}, @code{client-key.pem}
  1196. and @code{client-cert.pem} as appropriate. It is also possible to
  1197. include a set of pre-generated Diffie-Hellman (DH) parameters in a file
  1198. @code{dh-params.pem}, which can be created using the
  1199. @code{certtool --generate-dh-params} command. If omitted, QEMU will
  1200. dynamically generate DH parameters when loading the credentials.
  1201. The @code{endpoint} parameter indicates whether the credentials will
  1202. be used for a network client or server, and determines which PEM
  1203. files are loaded.
  1204. The @code{verify} parameter determines whether x509 certificate
  1205. validation should be performed. This defaults to enabled, meaning
  1206. clients will always validate the server hostname against the
  1207. certificate subject alt name fields and/or CN field. It also
  1208. means that servers will request that clients provide a certificate
  1209. and validate them. Verification should never be turned off for
  1210. client endpoints, however, it may be turned off for server endpoints
  1211. if an alternative mechanism is used to authenticate clients. For
  1212. example, the VNC server can use SASL to authenticate clients
  1213. instead.
  1214. To load server credentials with client certificate validation
  1215. enabled
  1216. @example
  1217. $QEMU -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=server
  1218. @end example
  1219. while to load client credentials use
  1220. @example
  1221. $QEMU -object tls-creds-x509,id=tls0,dir=/etc/pki/qemu,endpoint=client
  1222. @end example
  1223. Network services which support TLS will all have a @code{tls-creds}
  1224. parameter which expects the ID of the TLS credentials object. For
  1225. example with VNC:
  1226. @example
  1227. $QEMU -vnc,tls-creds=tls0
  1228. @end example
  1229. @node tls_psk
  1230. @subsection TLS Pre-Shared Keys (PSK)
  1231. Instead of using certificates, you may also use TLS Pre-Shared Keys
  1232. (TLS-PSK). This can be simpler to set up than certificates but is
  1233. less scalable.
  1234. Use the GnuTLS @code{psktool} program to generate a @code{keys.psk}
  1235. file containing one or more usernames and random keys:
  1236. @example
  1237. mkdir -m 0700 /tmp/keys
  1238. psktool -u rich -p /tmp/keys/keys.psk
  1239. @end example
  1240. TLS-enabled servers such as qemu-nbd can use this directory like so:
  1241. @example
  1242. qemu-nbd \
  1243. -t -x / \
  1244. --object tls-creds-psk,id=tls0,endpoint=server,dir=/tmp/keys \
  1245. --tls-creds tls0 \
  1246. image.qcow2
  1247. @end example
  1248. When connecting from a qemu-based client you must specify the
  1249. directory containing @code{keys.psk} and an optional @var{username}
  1250. (defaults to ``qemu''):
  1251. @example
  1252. qemu-img info \
  1253. --object tls-creds-psk,id=tls0,dir=/tmp/keys,username=rich,endpoint=client \
  1254. --image-opts \
  1255. file.driver=nbd,,file.port=10809,file.tls-creds=tls0,file.export=/
  1256. @end example
  1257. @node gdb_usage
  1258. @section GDB usage
  1259. QEMU has a primitive support to work with gdb, so that you can do
  1260. 'Ctrl-C' while the virtual machine is running and inspect its state.
  1261. In order to use gdb, launch QEMU with the '-s' option. It will wait for a
  1262. gdb connection:
  1263. @example
  1264. qemu-system-i386 -s -kernel arch/i386/boot/bzImage -hda root-2.4.20.img \
  1265. -append "root=/dev/hda"
  1266. Connected to host network interface: tun0
  1267. Waiting gdb connection on port 1234
  1268. @end example
  1269. Then launch gdb on the 'vmlinux' executable:
  1270. @example
  1271. > gdb vmlinux
  1272. @end example
  1273. In gdb, connect to QEMU:
  1274. @example
  1275. (gdb) target remote localhost:1234
  1276. @end example
  1277. Then you can use gdb normally. For example, type 'c' to launch the kernel:
  1278. @example
  1279. (gdb) c
  1280. @end example
  1281. Here are some useful tips in order to use gdb on system code:
  1282. @enumerate
  1283. @item
  1284. Use @code{info reg} to display all the CPU registers.
  1285. @item
  1286. Use @code{x/10i $eip} to display the code at the PC position.
  1287. @item
  1288. Use @code{set architecture i8086} to dump 16 bit code. Then use
  1289. @code{x/10i $cs*16+$eip} to dump the code at the PC position.
  1290. @end enumerate
  1291. Advanced debugging options:
  1292. The default single stepping behavior is step with the IRQs and timer service routines off. It is set this way because when gdb executes a single step it expects to advance beyond the current instruction. With the IRQs and timer service routines on, a single step might jump into the one of the interrupt or exception vectors instead of executing the current instruction. This means you may hit the same breakpoint a number of times before executing the instruction gdb wants to have executed. Because there are rare circumstances where you want to single step into an interrupt vector the behavior can be controlled from GDB. There are three commands you can query and set the single step behavior:
  1293. @table @code
  1294. @item maintenance packet qqemu.sstepbits
  1295. This will display the MASK bits used to control the single stepping IE:
  1296. @example
  1297. (gdb) maintenance packet qqemu.sstepbits
  1298. sending: "qqemu.sstepbits"
  1299. received: "ENABLE=1,NOIRQ=2,NOTIMER=4"
  1300. @end example
  1301. @item maintenance packet qqemu.sstep
  1302. This will display the current value of the mask used when single stepping IE:
  1303. @example
  1304. (gdb) maintenance packet qqemu.sstep
  1305. sending: "qqemu.sstep"
  1306. received: "0x7"
  1307. @end example
  1308. @item maintenance packet Qqemu.sstep=HEX_VALUE
  1309. This will change the single step mask, so if wanted to enable IRQs on the single step, but not timers, you would use:
  1310. @example
  1311. (gdb) maintenance packet Qqemu.sstep=0x5
  1312. sending: "qemu.sstep=0x5"
  1313. received: "OK"
  1314. @end example
  1315. @end table
  1316. @node pcsys_os_specific
  1317. @section Target OS specific information
  1318. @subsection Linux
  1319. To have access to SVGA graphic modes under X11, use the @code{vesa} or
  1320. the @code{cirrus} X11 driver. For optimal performances, use 16 bit
  1321. color depth in the guest and the host OS.
  1322. When using a 2.6 guest Linux kernel, you should add the option
  1323. @code{clock=pit} on the kernel command line because the 2.6 Linux
  1324. kernels make very strict real time clock checks by default that QEMU
  1325. cannot simulate exactly.
  1326. When using a 2.6 guest Linux kernel, verify that the 4G/4G patch is
  1327. not activated because QEMU is slower with this patch. The QEMU
  1328. Accelerator Module is also much slower in this case. Earlier Fedora
  1329. Core 3 Linux kernel (< 2.6.9-1.724_FC3) were known to incorporate this
  1330. patch by default. Newer kernels don't have it.
  1331. @subsection Windows
  1332. If you have a slow host, using Windows 95 is better as it gives the
  1333. best speed. Windows 2000 is also a good choice.
  1334. @subsubsection SVGA graphic modes support
  1335. QEMU emulates a Cirrus Logic GD5446 Video
  1336. card. All Windows versions starting from Windows 95 should recognize
  1337. and use this graphic card. For optimal performances, use 16 bit color
  1338. depth in the guest and the host OS.
  1339. If you are using Windows XP as guest OS and if you want to use high
  1340. resolution modes which the Cirrus Logic BIOS does not support (i.e. >=
  1341. 1280x1024x16), then you should use the VESA VBE virtual graphic card
  1342. (option @option{-std-vga}).
  1343. @subsubsection CPU usage reduction
  1344. Windows 9x does not correctly use the CPU HLT
  1345. instruction. The result is that it takes host CPU cycles even when
  1346. idle. You can install the utility from
  1347. @url{}
  1348. to solve this problem. Note that no such tool is needed for NT, 2000 or XP.
  1349. @subsubsection Windows 2000 disk full problem
  1350. Windows 2000 has a bug which gives a disk full problem during its
  1351. installation. When installing it, use the @option{-win2k-hack} QEMU
  1352. option to enable a specific workaround. After Windows 2000 is
  1353. installed, you no longer need this option (this option slows down the
  1354. IDE transfers).
  1355. @subsubsection Windows 2000 shutdown
  1356. Windows 2000 cannot automatically shutdown in QEMU although Windows 98
  1357. can. It comes from the fact that Windows 2000 does not automatically
  1358. use the APM driver provided by the BIOS.
  1359. In order to correct that, do the following (thanks to Struan
  1360. Bartlett): go to the Control Panel => Add/Remove Hardware & Next =>
  1361. Add/Troubleshoot a device => Add a new device & Next => No, select the
  1362. hardware from a list & Next => NT Apm/Legacy Support & Next => Next
  1363. (again) a few times. Now the driver is installed and Windows 2000 now
  1364. correctly instructs QEMU to shutdown at the appropriate moment.
  1365. @subsubsection Share a directory between Unix and Windows
  1366. See @ref{sec_invocation} about the help of the option
  1367. @option{'-netdev user,smb=...'}.
  1368. @subsubsection Windows XP security problem
  1369. Some releases of Windows XP install correctly but give a security
  1370. error when booting:
  1371. @example
  1372. A problem is preventing Windows from accurately checking the
  1373. license for this computer. Error code: 0x800703e6.
  1374. @end example
  1375. The workaround is to install a service pack for XP after a boot in safe
  1376. mode. Then reboot, and the problem should go away. Since there is no
  1377. network while in safe mode, its recommended to download the full
  1378. installation of SP1 or SP2 and transfer that via an ISO or using the
  1379. vvfat block device ("-hdb fat:directory_which_holds_the_SP").
  1380. @subsection MS-DOS and FreeDOS
  1381. @subsubsection CPU usage reduction
  1382. DOS does not correctly use the CPU HLT instruction. The result is that
  1383. it takes host CPU cycles even when idle. You can install the utility from
  1384. @url{}
  1385. to solve this problem.
  1386. @node QEMU System emulator for non PC targets
  1387. @chapter QEMU System emulator for non PC targets
  1388. QEMU is a generic emulator and it emulates many non PC
  1389. machines. Most of the options are similar to the PC emulator. The
  1390. differences are mentioned in the following sections.
  1391. @menu
  1392. * PowerPC System emulator::
  1393. * Sparc32 System emulator::
  1394. * Sparc64 System emulator::
  1395. * MIPS System emulator::
  1396. * ARM System emulator::
  1397. * ColdFire System emulator::
  1398. * Cris System emulator::
  1399. * Microblaze System emulator::
  1400. * SH4 System emulator::
  1401. * Xtensa System emulator::
  1402. @end menu
  1403. @node PowerPC System emulator
  1404. @section PowerPC System emulator
  1405. @cindex system emulation (PowerPC)
  1406. Use the executable @file{qemu-system-ppc} to simulate a complete PREP
  1407. or PowerMac PowerPC system.
  1408. QEMU emulates the following PowerMac peripherals:
  1409. @itemize @minus
  1410. @item
  1411. UniNorth or Grackle PCI Bridge
  1412. @item
  1413. PCI VGA compatible card with VESA Bochs Extensions
  1414. @item
  1415. 2 PMAC IDE interfaces with hard disk and CD-ROM support
  1416. @item
  1417. NE2000 PCI adapters
  1418. @item
  1419. Non Volatile RAM
  1420. @item
  1421. VIA-CUDA with ADB keyboard and mouse.
  1422. @end itemize
  1423. QEMU emulates the following PREP peripherals:
  1424. @itemize @minus
  1425. @item
  1426. PCI Bridge
  1427. @item
  1428. PCI VGA compatible card with VESA Bochs Extensions
  1429. @item
  1430. 2 IDE interfaces with hard disk and CD-ROM support
  1431. @item
  1432. Floppy disk
  1433. @item
  1434. NE2000 network adapters
  1435. @item
  1436. Serial port
  1437. @item
  1438. PREP Non Volatile RAM
  1439. @item
  1440. PC compatible keyboard and mouse.
  1441. @end itemize
  1442. QEMU uses the Open Hack'Ware Open Firmware Compatible BIOS available at
  1443. @url{}.
  1444. Since version 0.9.1, QEMU uses OpenBIOS @url{}
  1445. for the g3beige and mac99 PowerMac machines. OpenBIOS is a free (GPL
  1446. v2) portable firmware implementation. The goal is to implement a 100%
  1447. IEEE 1275-1994 (referred to as Open Firmware) compliant firmware.
  1448. @c man begin OPTIONS
  1449. The following options are specific to the PowerPC emulation:
  1450. @table @option
  1451. @item -g @var{W}x@var{H}[x@var{DEPTH}]
  1452. Set the initial VGA graphic mode. The default is 800x600x32.
  1453. @item -prom-env @var{string}
  1454. Set OpenBIOS variables in NVRAM, for example:
  1455. @example
  1456. qemu-system-ppc -prom-env 'auto-boot?=false' \
  1457. -prom-env 'boot-device=hd:2,\yaboot' \
  1458. -prom-env 'boot-args=conf=hd:2,\yaboot.conf'
  1459. @end example
  1460. These variables are not used by Open Hack'Ware.
  1461. @end table
  1462. @c man end
  1463. More information is available at
  1464. @url{}.
  1465. @node Sparc32 System emulator
  1466. @section Sparc32 System emulator
  1467. @cindex system emulation (Sparc32)
  1468. Use the executable @file{qemu-system-sparc} to simulate the following
  1469. Sun4m architecture machines:
  1470. @itemize @minus
  1471. @item
  1472. SPARCstation 4
  1473. @item
  1474. SPARCstation 5
  1475. @item
  1476. SPARCstation 10
  1477. @item
  1478. SPARCstation 20
  1479. @item
  1480. SPARCserver 600MP
  1481. @item
  1482. SPARCstation LX
  1483. @item
  1484. SPARCstation Voyager
  1485. @item
  1486. SPARCclassic
  1487. @item
  1488. SPARCbook
  1489. @end itemize
  1490. The emulation is somewhat complete. SMP up to 16 CPUs is supported,
  1491. but Linux limits the number of usable CPUs to 4.
  1492. QEMU emulates the following sun4m peripherals:
  1493. @itemize @minus
  1494. @item
  1495. IOMMU
  1496. @item
  1497. TCX or cgthree Frame buffer
  1498. @item
  1499. Lance (Am7990) Ethernet
  1500. @item
  1501. Non Volatile RAM M48T02/M48T08
  1502. @item
  1503. Slave I/O: timers, interrupt controllers, Zilog serial ports, keyboard
  1504. and power/reset logic
  1505. @item
  1506. ESP SCSI controller with hard disk and CD-ROM support
  1507. @item
  1508. Floppy drive (not on SS-600MP)
  1509. @item
  1510. CS4231 sound device (only on SS-5, not working yet)
  1511. @end itemize
  1512. The number of peripherals is fixed in the architecture. Maximum
  1513. memory size depends on the machine type, for SS-5 it is 256MB and for
  1514. others 2047MB.
  1515. Since version 0.8.2, QEMU uses OpenBIOS
  1516. @url{}. OpenBIOS is a free (GPL v2) portable
  1517. firmware implementation. The goal is to implement a 100% IEEE
  1518. 1275-1994 (referred to as Open Firmware) compliant firmware.
  1519. A sample Linux 2.6 series kernel and ram disk image are available on
  1520. the QEMU web site. There are still issues with NetBSD and OpenBSD, but
  1521. most kernel versions work. Please note that currently older Solaris kernels
  1522. don't work probably due to interface issues between OpenBIOS and
  1523. Solaris.
  1524. @c man begin OPTIONS
  1525. The following options are specific to the Sparc32 emulation:
  1526. @table @option
  1527. @item -g @var{W}x@var{H}x[x@var{DEPTH}]
  1528. Set the initial graphics mode. For TCX, the default is 1024x768x8 with the
  1529. option of 1024x768x24. For cgthree, the default is 1024x768x8 with the option
  1530. of 1152x900x8 for people who wish to use OBP.
  1531. @item -prom-env @var{string}
  1532. Set OpenBIOS variables in NVRAM, for example:
  1533. @example
  1534. qemu-system-sparc -prom-env 'auto-boot?=false' \
  1535. -prom-env 'boot-device=sd(0,2,0):d' -prom-env 'boot-args=linux single'
  1536. @end example
  1537. @item -M [SS-4|SS-5|SS-10|SS-20|SS-600MP|LX|Voyager|SPARCClassic] [|SPARCbook]
  1538. Set the emulated machine type. Default is SS-5.
  1539. @end table
  1540. @c man end
  1541. @node Sparc64 System emulator
  1542. @section Sparc64 System emulator
  1543. @cindex system emulation (Sparc64)
  1544. Use the executable @file{qemu-system-sparc64} to simulate a Sun4u
  1545. (UltraSPARC PC-like machine), Sun4v (T1 PC-like machine), or generic
  1546. Niagara (T1) machine. The Sun4u emulator is mostly complete, being
  1547. able to run Linux, NetBSD and OpenBSD in headless (-nographic) mode. The
  1548. Sun4v emulator is still a work in progress.
  1549. The Niagara T1 emulator makes use of firmware and OS binaries supplied in the S10image/ directory
  1550. of the OpenSPARC T1 project @url{}
  1551. and is able to boot the disk.s10hw2 Solaris image.
  1552. @example
  1553. qemu-system-sparc64 -M niagara -L /path-to/S10image/ \
  1554. -nographic -m 256 \
  1555. -drive if=pflash,readonly=on,file=/S10image/disk.s10hw2
  1556. @end example
  1557. QEMU emulates the following peripherals:
  1558. @itemize @minus
  1559. @item
  1560. UltraSparc IIi APB PCI Bridge
  1561. @item
  1562. PCI VGA compatible card with VESA Bochs Extensions
  1563. @item
  1564. PS/2 mouse and keyboard
  1565. @item
  1566. Non Volatile RAM M48T59
  1567. @item
  1568. PC-compatible serial ports
  1569. @item
  1570. 2 PCI IDE interfaces with hard disk and CD-ROM support
  1571. @item
  1572. Floppy disk
  1573. @end itemize
  1574. @c man begin OPTIONS
  1575. The following options are specific to the Sparc64 emulation:
  1576. @table @option
  1577. @item -prom-env @var{string}
  1578. Set OpenBIOS variables in NVRAM, for example:
  1579. @example
  1580. qemu-system-sparc64 -prom-env 'auto-boot?=false'
  1581. @end example
  1582. @item -M [sun4u|sun4v|niagara]
  1583. Set the emulated machine type. The default is sun4u.
  1584. @end table
  1585. @c man end
  1586. @node MIPS System emulator
  1587. @section MIPS System emulator
  1588. @cindex system emulation (MIPS)
  1589. @menu
  1590. * nanoMIPS System emulator ::
  1591. @end menu
  1592. Four executables cover simulation of 32 and 64-bit MIPS systems in
  1593. both endian options, @file{qemu-system-mips}, @file{qemu-system-mipsel}
  1594. @file{qemu-system-mips64} and @file{qemu-system-mips64el}.
  1595. Five different machine types are emulated:
  1596. @itemize @minus
  1597. @item
  1598. A generic ISA PC-like machine "mips"
  1599. @item
  1600. The MIPS Malta prototype board "malta"
  1601. @item
  1602. An ACER Pica "pica61". This machine needs the 64-bit emulator.
  1603. @item
  1604. MIPS emulator pseudo board "mipssim"
  1605. @item
  1606. A MIPS Magnum R4000 machine "magnum". This machine needs the 64-bit emulator.
  1607. @end itemize
  1608. The generic emulation is supported by Debian 'Etch' and is able to
  1609. install Debian into a virtual disk image. The following devices are
  1610. emulated:
  1611. @itemize @minus
  1612. @item
  1613. A range of MIPS CPUs, default is the 24Kf
  1614. @item
  1615. PC style serial port
  1616. @item
  1617. PC style IDE disk
  1618. @item
  1619. NE2000 network card
  1620. @end itemize
  1621. The Malta emulation supports the following devices:
  1622. @itemize @minus
  1623. @item
  1624. Core board with MIPS 24Kf CPU and Galileo system controller
  1625. @item
  1626. PIIX4 PCI/USB/SMbus controller
  1627. @item
  1628. The Multi-I/O chip's serial device
  1629. @item
  1630. PCI network cards (PCnet32 and others)
  1631. @item
  1632. Malta FPGA serial device
  1633. @item
  1634. Cirrus (default) or any other PCI VGA graphics card
  1635. @end itemize
  1636. The Boston board emulation supports the following devices:
  1637. @itemize @minus
  1638. @item
  1639. Xilinx FPGA, which includes a PCIe root port and an UART
  1640. @item
  1641. Intel EG20T PCH connects the I/O peripherals, but only the SATA bus is emulated
  1642. @end itemize
  1643. The ACER Pica emulation supports:
  1644. @itemize @minus
  1645. @item
  1646. MIPS R4000 CPU
  1647. @item
  1648. PC-style IRQ and DMA controllers
  1649. @item
  1650. PC Keyboard
  1651. @item
  1652. IDE controller
  1653. @end itemize
  1654. The MIPS Magnum R4000 emulation supports:
  1655. @itemize @minus
  1656. @item
  1657. MIPS R4000 CPU
  1658. @item
  1659. PC-style IRQ controller
  1660. @item
  1661. PC Keyboard
  1662. @item
  1663. SCSI controller
  1664. @item
  1665. G364 framebuffer
  1666. @end itemize
  1667. The Fulong 2E emulation supports:
  1668. @itemize @minus
  1669. @item
  1670. Loongson 2E CPU
  1671. @item
  1672. Bonito64 system controller as North Bridge
  1673. @item
  1674. VT82C686 chipset as South Bridge
  1675. @item
  1676. RTL8139D as a network card chipset
  1677. @end itemize
  1678. The mipssim pseudo board emulation provides an environment similar
  1679. to what the proprietary MIPS emulator uses for running Linux.
  1680. It supports:
  1681. @itemize @minus
  1682. @item
  1683. A range of MIPS CPUs, default is the 24Kf
  1684. @item
  1685. PC style serial port
  1686. @item
  1687. MIPSnet network emulation
  1688. @end itemize
  1689. @node nanoMIPS System emulator
  1690. @subsection nanoMIPS System emulator
  1691. @cindex system emulation (nanoMIPS)
  1692. Executable @file{qemu-system-mipsel} also covers simulation of
  1693. 32-bit nanoMIPS system in little endian mode:
  1694. @itemize @minus
  1695. @item
  1696. nanoMIPS I7200 CPU
  1697. @end itemize
  1698. Example of @file{qemu-system-mipsel} usage for nanoMIPS is shown below:
  1699. Download @code{<disk_image_file>} from @url{}.
  1700. Download @code{<kernel_image_file>} from @url{}.
  1701. Start system emulation of Malta board with nanoMIPS I7200 CPU:
  1702. @example
  1703. qemu-system-mipsel -cpu I7200 -kernel @code{<kernel_image_file>} \
  1704. -M malta -serial stdio -m @code{<memory_size>} -hda @code{<disk_image_file>} \
  1705. -append "mem=256m@@0x0 rw console=ttyS0 vga=cirrus vesa=0x111 root=/dev/sda"
  1706. @end example
  1707. @node ARM System emulator
  1708. @section ARM System emulator
  1709. @cindex system emulation (ARM)
  1710. Use the executable @file{qemu-system-arm} to simulate a ARM
  1711. machine. The ARM Integrator/CP board is emulated with the following
  1712. devices:
  1713. @itemize @minus
  1714. @item
  1715. ARM926E, ARM1026E, ARM946E, ARM1136 or Cortex-A8 CPU
  1716. @item
  1717. Two PL011 UARTs
  1718. @item
  1719. SMC 91c111 Ethernet adapter
  1720. @item
  1721. PL110 LCD controller
  1722. @item
  1723. PL050 KMI with PS/2 keyboard and mouse.
  1724. @item
  1725. PL181 MultiMedia Card Interface with SD card.
  1726. @end itemize
  1727. The ARM Versatile baseboard is emulated with the following devices:
  1728. @itemize @minus
  1729. @item
  1730. ARM926E, ARM1136 or Cortex-A8 CPU
  1731. @item
  1732. PL190 Vectored Interrupt Controller
  1733. @item
  1734. Four PL011 UARTs
  1735. @item
  1736. SMC 91c111 Ethernet adapter
  1737. @item
  1738. PL110 LCD controller
  1739. @item
  1740. PL050 KMI with PS/2 keyboard and mouse.
  1741. @item
  1742. PCI host bridge. Note the emulated PCI bridge only provides access to
  1743. PCI memory space. It does not provide access to PCI IO space.
  1744. This means some devices (eg. ne2k_pci NIC) are not usable, and others
  1745. (eg. rtl8139 NIC) are only usable when the guest drivers use the memory
  1746. mapped control registers.
  1747. @item
  1748. PCI OHCI USB controller.
  1749. @item
  1750. LSI53C895A PCI SCSI Host Bus Adapter with hard disk and CD-ROM devices.
  1751. @item
  1752. PL181 MultiMedia Card Interface with SD card.
  1753. @end itemize
  1754. Several variants of the ARM RealView baseboard are emulated,
  1755. including the EB, PB-A8 and PBX-A9. Due to interactions with the
  1756. bootloader, only certain Linux kernel configurations work out
  1757. of the box on these boards.
  1758. Kernels for the PB-A8 board should have CONFIG_REALVIEW_HIGH_PHYS_OFFSET
  1759. enabled in the kernel, and expect 512M RAM. Kernels for The PBX-A9 board
  1761. disabled and expect 1024M RAM.
  1762. The following devices are emulated:
  1763. @itemize @minus
  1764. @item
  1765. ARM926E, ARM1136, ARM11MPCore, Cortex-A8 or Cortex-A9 MPCore CPU
  1766. @item
  1767. ARM AMBA Generic/Distributed Interrupt Controller
  1768. @item
  1769. Four PL011 UARTs
  1770. @item
  1771. SMC 91c111 or SMSC LAN9118 Ethernet adapter
  1772. @item
  1773. PL110 LCD controller
  1774. @item
  1775. PL050 KMI with PS/2 keyboard and mouse
  1776. @item
  1777. PCI host bridge
  1778. @item
  1779. PCI OHCI USB controller
  1780. @item
  1781. LSI53C895A PCI SCSI Host Bus Adapter with hard disk and CD-ROM devices
  1782. @item
  1783. PL181 MultiMedia Card Interface with SD card.
  1784. @end itemize
  1785. The XScale-based clamshell PDA models ("Spitz", "Akita", "Borzoi"
  1786. and "Terrier") emulation includes the following peripherals:
  1787. @itemize @minus
  1788. @item
  1789. Intel PXA270 System-on-chip (ARM V5TE core)
  1790. @item
  1791. NAND Flash memory
  1792. @item
  1793. IBM/Hitachi DSCM microdrive in a PXA PCMCIA slot - not in "Akita"
  1794. @item
  1795. On-chip OHCI USB controller
  1796. @item
  1797. On-chip LCD controller
  1798. @item
  1799. On-chip Real Time Clock
  1800. @item
  1801. TI ADS7846 touchscreen controller on SSP bus
  1802. @item
  1803. Maxim MAX1111 analog-digital converter on I@math{^2}C bus
  1804. @item
  1805. GPIO-connected keyboard controller and LEDs
  1806. @item
  1807. Secure Digital card connected to PXA MMC/SD host
  1808. @item
  1809. Three on-chip UARTs
  1810. @item
  1811. WM8750 audio CODEC on I@math{^2}C and I@math{^2}S busses
  1812. @end itemize
  1813. The Palm Tungsten|E PDA (codename "Cheetah") emulation includes the
  1814. following elements:
  1815. @itemize @minus
  1816. @item
  1817. Texas Instruments OMAP310 System-on-chip (ARM 925T core)
  1818. @item
  1819. ROM and RAM memories (ROM firmware image can be loaded with -option-rom)
  1820. @item
  1821. On-chip LCD controller
  1822. @item
  1823. On-chip Real Time Clock
  1824. @item
  1825. TI TSC2102i touchscreen controller / analog-digital converter / Audio
  1826. CODEC, connected through MicroWire and I@math{^2}S busses
  1827. @item
  1828. GPIO-connected matrix keypad
  1829. @item
  1830. Secure Digital card connected to OMAP MMC/SD host
  1831. @item
  1832. Three on-chip UARTs
  1833. @end itemize
  1834. Nokia N800 and N810 internet tablets (known also as RX-34 and RX-44 / 48)
  1835. emulation supports the following elements:
  1836. @itemize @minus
  1837. @item
  1838. Texas Instruments OMAP2420 System-on-chip (ARM 1136 core)
  1839. @item
  1840. RAM and non-volatile OneNAND Flash memories
  1841. @item
  1842. Display connected to EPSON remote framebuffer chip and OMAP on-chip
  1843. display controller and a LS041y3 MIPI DBI-C controller
  1844. @item
  1845. TI TSC2301 (in N800) and TI TSC2005 (in N810) touchscreen controllers
  1846. driven through SPI bus
  1847. @item
  1848. National Semiconductor LM8323-controlled qwerty keyboard driven
  1849. through I@math{^2}C bus
  1850. @item
  1851. Secure Digital card connected to OMAP MMC/SD host
  1852. @item
  1853. Three OMAP on-chip UARTs and on-chip STI debugging console
  1854. @item
  1855. A Bluetooth(R) transceiver and HCI connected to an UART
  1856. @item
  1857. Mentor Graphics "Inventra" dual-role USB controller embedded in a TI
  1858. TUSB6010 chip - only USB host mode is supported
  1859. @item
  1860. TI TMP105 temperature sensor driven through I@math{^2}C bus
  1861. @item
  1862. TI TWL92230C power management companion with an RTC on I@math{^2}C bus
  1863. @item
  1864. Nokia RETU and TAHVO multi-purpose chips with an RTC, connected
  1865. through CBUS
  1866. @end itemize
  1867. The Luminary Micro Stellaris LM3S811EVB emulation includes the following
  1868. devices:
  1869. @itemize @minus
  1870. @item
  1871. Cortex-M3 CPU core.
  1872. @item
  1873. 64k Flash and 8k SRAM.
  1874. @item
  1875. Timers, UARTs, ADC and I@math{^2}C interface.
  1876. @item
  1877. OSRAM Pictiva 96x16 OLED with SSD0303 controller on I@math{^2}C bus.
  1878. @end itemize
  1879. The Luminary Micro Stellaris LM3S6965EVB emulation includes the following
  1880. devices:
  1881. @itemize @minus
  1882. @item
  1883. Cortex-M3 CPU core.
  1884. @item
  1885. 256k Flash and 64k SRAM.
  1886. @item
  1887. Timers, UARTs, ADC, I@math{^2}C and SSI interfaces.
  1888. @item
  1889. OSRAM Pictiva 128x64 OLED with SSD0323 controller connected via SSI.
  1890. @end itemize
  1891. The Freecom MusicPal internet radio emulation includes the following
  1892. elements:
  1893. @itemize @minus
  1894. @item
  1895. Marvell MV88W8618 ARM core.
  1896. @item
  1897. 32 MB RAM, 256 KB SRAM, 8 MB flash.
  1898. @item
  1899. Up to 2 16550 UARTs
  1900. @item
  1901. MV88W8xx8 Ethernet controller
  1902. @item
  1903. MV88W8618 audio controller, WM8750 CODEC and mixer
  1904. @item
  1905. 128×64 display with brightness control
  1906. @item
  1907. 2 buttons, 2 navigation wheels with button function
  1908. @end itemize
  1909. The Siemens SX1 models v1 and v2 (default) basic emulation.
  1910. The emulation includes the following elements:
  1911. @itemize @minus
  1912. @item
  1913. Texas Instruments OMAP310 System-on-chip (ARM 925T core)
  1914. @item
  1915. ROM and RAM memories (ROM firmware image can be loaded with -pflash)
  1916. V1
  1917. 1 Flash of 16MB and 1 Flash of 8MB
  1918. V2
  1919. 1 Flash of 32MB
  1920. @item
  1921. On-chip LCD controller
  1922. @item
  1923. On-chip Real Time Clock
  1924. @item
  1925. Secure Digital card connected to OMAP MMC/SD host
  1926. @item
  1927. Three on-chip UARTs
  1928. @end itemize
  1929. A Linux 2.6 test image is available on the QEMU web site. More
  1930. information is available in the QEMU mailing-list archive.
  1931. @c man begin OPTIONS
  1932. The following options are specific to the ARM emulation:
  1933. @table @option
  1934. @item -semihosting
  1935. Enable semihosting syscall emulation.
  1936. On ARM this implements the "Angel" interface.
  1937. Note that this allows guest direct access to the host filesystem,
  1938. so should only be used with trusted guest OS.
  1939. @end table
  1940. @c man end
  1941. @node ColdFire System emulator
  1942. @section ColdFire System emulator
  1943. @cindex system emulation (ColdFire)
  1944. @cindex system emulation (M68K)
  1945. Use the executable @file{qemu-system-m68k} to simulate a ColdFire machine.
  1946. The emulator is able to boot a uClinux kernel.
  1947. The M5208EVB emulation includes the following devices:
  1948. @itemize @minus
  1949. @item
  1950. MCF5208 ColdFire V2 Microprocessor (ISA A+ with EMAC).
  1951. @item
  1952. Three Two on-chip UARTs.
  1953. @item
  1954. Fast Ethernet Controller (FEC)
  1955. @end itemize
  1956. The AN5206 emulation includes the following devices:
  1957. @itemize @minus
  1958. @item
  1959. MCF5206 ColdFire V2 Microprocessor.
  1960. @item
  1961. Two on-chip UARTs.
  1962. @end itemize
  1963. @c man begin OPTIONS
  1964. The following options are specific to the ColdFire emulation:
  1965. @table @option
  1966. @item -semihosting
  1967. Enable semihosting syscall emulation.
  1968. On M68K this implements the "ColdFire GDB" interface used by libgloss.
  1969. Note that this allows guest direct access to the host filesystem,
  1970. so should only be used with trusted guest OS.
  1971. @end table
  1972. @c man end
  1973. @node Cris System emulator
  1974. @section Cris System emulator
  1975. @cindex system emulation (Cris)
  1976. TODO
  1977. @node Microblaze System emulator
  1978. @section Microblaze System emulator
  1979. @cindex system emulation (Microblaze)
  1980. TODO
  1981. @node SH4 System emulator
  1982. @section SH4 System emulator
  1983. @cindex system emulation (SH4)
  1984. TODO
  1985. @node Xtensa System emulator
  1986. @section Xtensa System emulator
  1987. @cindex system emulation (Xtensa)
  1988. Two executables cover simulation of both Xtensa endian options,
  1989. @file{qemu-system-xtensa} and @file{qemu-system-xtensaeb}.
  1990. Two different machine types are emulated:
  1991. @itemize @minus
  1992. @item
  1993. Xtensa emulator pseudo board "sim"
  1994. @item
  1995. Avnet LX60/LX110/LX200 board
  1996. @end itemize
  1997. The sim pseudo board emulation provides an environment similar
  1998. to one provided by the proprietary Tensilica ISS.
  1999. It supports:
  2000. @itemize @minus
  2001. @item
  2002. A range of Xtensa CPUs, default is the DC232B
  2003. @item
  2004. Console and filesystem access via semihosting calls
  2005. @end itemize
  2006. The Avnet LX60/LX110/LX200 emulation supports:
  2007. @itemize @minus
  2008. @item
  2009. A range of Xtensa CPUs, default is the DC232B
  2010. @item
  2011. 16550 UART
  2012. @item
  2013. OpenCores 10/100 Mbps Ethernet MAC
  2014. @end itemize
  2015. @c man begin OPTIONS
  2016. The following options are specific to the Xtensa emulation:
  2017. @table @option
  2018. @item -semihosting
  2019. Enable semihosting syscall emulation.
  2020. Xtensa semihosting provides basic file IO calls, such as open/read/write/seek/select.
  2021. Tensilica baremetal libc for ISS and linux platform "sim" use this interface.
  2022. Note that this allows guest direct access to the host filesystem,
  2023. so should only be used with trusted guest OS.
  2024. @end table
  2025. @c man end
  2026. @node QEMU Guest Agent
  2027. @chapter QEMU Guest Agent invocation
  2028. @include qemu-ga.texi
  2029. @node QEMU User space emulator
  2030. @chapter QEMU User space emulator
  2031. @menu
  2032. * Supported Operating Systems ::
  2033. * Features::
  2034. * Linux User space emulator::
  2035. * BSD User space emulator ::
  2036. @end menu
  2037. @node Supported Operating Systems
  2038. @section Supported Operating Systems
  2039. The following OS are supported in user space emulation:
  2040. @itemize @minus
  2041. @item
  2042. Linux (referred as qemu-linux-user)
  2043. @item
  2044. BSD (referred as qemu-bsd-user)
  2045. @end itemize
  2046. @node Features
  2047. @section Features
  2048. QEMU user space emulation has the following notable features:
  2049. @table @strong
  2050. @item System call translation:
  2051. QEMU includes a generic system call translator. This means that
  2052. the parameters of the system calls can be converted to fix
  2053. endianness and 32/64-bit mismatches between hosts and targets.
  2054. IOCTLs can be converted too.
  2055. @item POSIX signal handling:
  2056. QEMU can redirect to the running program all signals coming from
  2057. the host (such as @code{SIGALRM}), as well as synthesize signals from
  2058. virtual CPU exceptions (for example @code{SIGFPE} when the program
  2059. executes a division by zero).
  2060. QEMU relies on the host kernel to emulate most signal system
  2061. calls, for example to emulate the signal mask. On Linux, QEMU
  2062. supports both normal and real-time signals.
  2063. @item Threading:
  2064. On Linux, QEMU can emulate the @code{clone} syscall and create a real
  2065. host thread (with a separate virtual CPU) for each emulated thread.
  2066. Note that not all targets currently emulate atomic operations correctly.
  2067. x86 and ARM use a global lock in order to preserve their semantics.
  2068. @end table
  2069. QEMU was conceived so that ultimately it can emulate itself. Although
  2070. it is not very useful, it is an important test to show the power of the
  2071. emulator.
  2072. @node Linux User space emulator
  2073. @section Linux User space emulator
  2074. @menu
  2075. * Quick Start::
  2076. * Wine launch::
  2077. * Command line options::
  2078. * Other binaries::
  2079. @end menu
  2080. @node Quick Start
  2081. @subsection Quick Start
  2082. In order to launch a Linux process, QEMU needs the process executable
  2083. itself and all the target (x86) dynamic libraries used by it.
  2084. @itemize
  2085. @item On x86, you can just try to launch any process by using the native
  2086. libraries:
  2087. @example
  2088. qemu-i386 -L / /bin/ls
  2089. @end example
  2090. @code{-L /} tells that the x86 dynamic linker must be searched with a
  2091. @file{/} prefix.
  2092. @item Since QEMU is also a linux process, you can launch QEMU with
  2093. QEMU (NOTE: you can only do that if you compiled QEMU from the sources):
  2094. @example
  2095. qemu-i386 -L / qemu-i386 -L / /bin/ls
  2096. @end example
  2097. @item On non x86 CPUs, you need first to download at least an x86 glibc
  2098. (@file{qemu-runtime-i386-XXX-.tar.gz} on the QEMU web page). Ensure that
  2099. @code{LD_LIBRARY_PATH} is not set:
  2100. @example
  2101. unset LD_LIBRARY_PATH
  2102. @end example
  2103. Then you can launch the precompiled @file{ls} x86 executable:
  2104. @example
  2105. qemu-i386 tests/i386/ls
  2106. @end example
  2107. You can look at @file{scripts/} so that
  2108. QEMU is automatically launched by the Linux kernel when you try to
  2109. launch x86 executables. It requires the @code{binfmt_misc} module in the
  2110. Linux kernel.
  2111. @item The x86 version of QEMU is also included. You can try weird things such as:
  2112. @example
  2113. qemu-i386 /usr/local/qemu-i386/bin/qemu-i386 \
  2114. /usr/local/qemu-i386/bin/ls-i386
  2115. @end example
  2116. @end itemize
  2117. @node Wine launch
  2118. @subsection Wine launch
  2119. @itemize
  2120. @item Ensure that you have a working QEMU with the x86 glibc
  2121. distribution (see previous section). In order to verify it, you must be
  2122. able to do:
  2123. @example
  2124. qemu-i386 /usr/local/qemu-i386/bin/ls-i386
  2125. @end example
  2126. @item Download the binary x86 Wine install
  2127. (@file{qemu-XXX-i386-wine.tar.gz} on the QEMU web page).
  2128. @item Configure Wine on your account. Look at the provided script
  2129. @file{/usr/local/qemu-i386/@/bin/}. Your previous
  2130. @code{$@{HOME@}/.wine} directory is saved to @code{$@{HOME@}/}.
  2131. @item Then you can try the example @file{putty.exe}:
  2132. @example
  2133. qemu-i386 /usr/local/qemu-i386/wine/bin/wine \
  2134. /usr/local/qemu-i386/wine/c/Program\ Files/putty.exe
  2135. @end example
  2136. @end itemize
  2137. @node Command line options
  2138. @subsection Command line options
  2139. @example
  2140. @command{qemu-i386} [@option{-h]} [@option{-d]} [@option{-L} @var{path}] [@option{-s} @var{size}] [@option{-cpu} @var{model}] [@option{-g} @var{port}] [@option{-B} @var{offset}] [@option{-R} @var{size}] @var{program} [@var{arguments}...]
  2141. @end example
  2142. @table @option
  2143. @item -h
  2144. Print the help
  2145. @item -L path
  2146. Set the x86 elf interpreter prefix (default=/usr/local/qemu-i386)
  2147. @item -s size
  2148. Set the x86 stack size in bytes (default=524288)
  2149. @item -cpu model
  2150. Select CPU model (-cpu help for list and additional feature selection)
  2151. @item -E @var{var}=@var{value}
  2152. Set environment @var{var} to @var{value}.
  2153. @item -U @var{var}
  2154. Remove @var{var} from the environment.
  2155. @item -B offset
  2156. Offset guest address by the specified number of bytes. This is useful when
  2157. the address region required by guest applications is reserved on the host.
  2158. This option is currently only supported on some hosts.
  2159. @item -R size
  2160. Pre-allocate a guest virtual address space of the given size (in bytes).
  2161. "G", "M", and "k" suffixes may be used when specifying the size.
  2162. @end table
  2163. Debug options:
  2164. @table @option
  2165. @item -d item1,...
  2166. Activate logging of the specified items (use '-d help' for a list of log items)
  2167. @item -p pagesize
  2168. Act as if the host page size was 'pagesize' bytes
  2169. @item -g port
  2170. Wait gdb connection to port
  2171. @item -singlestep
  2172. Run the emulation in single step mode.
  2173. @end table
  2174. Environment variables:
  2175. @table @env
  2176. @item QEMU_STRACE
  2177. Print system calls and arguments similar to the 'strace' program
  2178. (NOTE: the actual 'strace' program will not work because the user
  2179. space emulator hasn't implemented ptrace). At the moment this is
  2180. incomplete. All system calls that don't have a specific argument
  2181. format are printed with information for six arguments. Many
  2182. flag-style arguments don't have decoders and will show up as numbers.
  2183. @end table
  2184. @node Other binaries
  2185. @subsection Other binaries
  2186. @cindex user mode (Alpha)
  2187. @command{qemu-alpha} TODO.
  2188. @cindex user mode (ARM)
  2189. @command{qemu-armeb} TODO.
  2190. @cindex user mode (ARM)
  2191. @command{qemu-arm} is also capable of running ARM "Angel" semihosted ELF
  2192. binaries (as implemented by the arm-elf and arm-eabi Newlib/GDB
  2193. configurations), and arm-uclinux bFLT format binaries.
  2194. @cindex user mode (ColdFire)
  2195. @cindex user mode (M68K)
  2196. @command{qemu-m68k} is capable of running semihosted binaries using the BDM
  2197. (m5xxx-ram-hosted.ld) or m68k-sim (sim.ld) syscall interfaces, and
  2198. coldfire uClinux bFLT format binaries.
  2199. The binary format is detected automatically.
  2200. @cindex user mode (Cris)
  2201. @command{qemu-cris} TODO.
  2202. @cindex user mode (i386)
  2203. @command{qemu-i386} TODO.
  2204. @command{qemu-x86_64} TODO.
  2205. @cindex user mode (Microblaze)
  2206. @command{qemu-microblaze} TODO.
  2207. @cindex user mode (MIPS)
  2208. @command{qemu-mips} executes 32-bit big endian MIPS binaries (MIPS O32 ABI).
  2209. @command{qemu-mipsel} executes 32-bit little endian MIPS binaries (MIPS O32 ABI).
  2210. @command{qemu-mips64} executes 64-bit big endian MIPS binaries (MIPS N64 ABI).
  2211. @command{qemu-mips64el} executes 64-bit little endian MIPS binaries (MIPS N64 ABI).
  2212. @command{qemu-mipsn32} executes 32-bit big endian MIPS binaries (MIPS N32 ABI).
  2213. @command{qemu-mipsn32el} executes 32-bit little endian MIPS binaries (MIPS N32 ABI).
  2214. @cindex user mode (NiosII)
  2215. @command{qemu-nios2} TODO.
  2216. @cindex user mode (PowerPC)
  2217. @command{qemu-ppc64abi32} TODO.
  2218. @command{qemu-ppc64} TODO.
  2219. @command{qemu-ppc} TODO.
  2220. @cindex user mode (SH4)
  2221. @command{qemu-sh4eb} TODO.
  2222. @command{qemu-sh4} TODO.
  2223. @cindex user mode (SPARC)
  2224. @command{qemu-sparc} can execute Sparc32 binaries (Sparc32 CPU, 32 bit ABI).
  2225. @command{qemu-sparc32plus} can execute Sparc32 and SPARC32PLUS binaries
  2226. (Sparc64 CPU, 32 bit ABI).
  2227. @command{qemu-sparc64} can execute some Sparc64 (Sparc64 CPU, 64 bit ABI) and
  2228. SPARC32PLUS binaries (Sparc64 CPU, 32 bit ABI).
  2229. @node BSD User space emulator
  2230. @section BSD User space emulator
  2231. @menu
  2232. * BSD Status::
  2233. * BSD Quick Start::
  2234. * BSD Command line options::
  2235. @end menu
  2236. @node BSD Status
  2237. @subsection BSD Status
  2238. @itemize @minus
  2239. @item
  2240. target Sparc64 on Sparc64: Some trivial programs work.
  2241. @end itemize
  2242. @node BSD Quick Start
  2243. @subsection Quick Start
  2244. In order to launch a BSD process, QEMU needs the process executable
  2245. itself and all the target dynamic libraries used by it.
  2246. @itemize
  2247. @item On Sparc64, you can just try to launch any process by using the native
  2248. libraries:
  2249. @example
  2250. qemu-sparc64 /bin/ls
  2251. @end example
  2252. @end itemize
  2253. @node BSD Command line options
  2254. @subsection Command line options
  2255. @example
  2256. @command{qemu-sparc64} [@option{-h]} [@option{-d]} [@option{-L} @var{path}] [@option{-s} @var{size}] [@option{-bsd} @var{type}] @var{program} [@var{arguments}...]
  2257. @end example
  2258. @table @option
  2259. @item -h
  2260. Print the help
  2261. @item -L path
  2262. Set the library root path (default=/)
  2263. @item -s size
  2264. Set the stack size in bytes (default=524288)
  2265. @item -ignore-environment
  2266. Start with an empty environment. Without this option,
  2267. the initial environment is a copy of the caller's environment.
  2268. @item -E @var{var}=@var{value}
  2269. Set environment @var{var} to @var{value}.
  2270. @item -U @var{var}
  2271. Remove @var{var} from the environment.
  2272. @item -bsd type
  2273. Set the type of the emulated BSD Operating system. Valid values are
  2274. FreeBSD, NetBSD and OpenBSD (default).
  2275. @end table
  2276. Debug options:
  2277. @table @option
  2278. @item -d item1,...
  2279. Activate logging of the specified items (use '-d help' for a list of log items)
  2280. @item -p pagesize
  2281. Act as if the host page size was 'pagesize' bytes
  2282. @item -singlestep
  2283. Run the emulation in single step mode.
  2284. @end table
  2285. @node System requirements
  2286. @chapter System requirements
  2287. @section KVM kernel module
  2288. On x86_64 hosts, the default set of CPU features enabled by the KVM accelerator
  2289. require the host to be running Linux v4.5 or newer.
  2290. The OpteronG[345] CPU models require KVM support for RDTSCP, which was
  2291. added with Linux 4.5 which is supported by the major distros. And even
  2292. if RHEL7 has kernel 3.10, KVM there has the required functionality there
  2293. to make it close to a 4.5 or newer kernel.
  2294. @include docs/security.texi
  2295. @include qemu-tech.texi
  2296. @include qemu-deprecated.texi
  2297. @node Supported build platforms
  2298. @appendix Supported build platforms
  2299. QEMU aims to support building and executing on multiple host OS platforms.
  2300. This appendix outlines which platforms are the major build targets. These
  2301. platforms are used as the basis for deciding upon the minimum required
  2302. versions of 3rd party software QEMU depends on. The supported platforms
  2303. are the targets for automated testing performed by the project when patches
  2304. are submitted for review, and tested before and after merge.
  2305. If a platform is not listed here, it does not imply that QEMU won't work.
  2306. If an unlisted platform has comparable software versions to a listed platform,
  2307. there is every expectation that it will work. Bug reports are welcome for
  2308. problems encountered on unlisted platforms unless they are clearly older
  2309. vintage than what is described here.
  2310. Note that when considering software versions shipped in distros as support
  2311. targets, QEMU considers only the version number, and assumes the features in
  2312. that distro match the upstream release with the same version. In other words,
  2313. if a distro backports extra features to the software in their distro, QEMU
  2314. upstream code will not add explicit support for those backports, unless the
  2315. feature is auto-detectable in a manner that works for the upstream releases
  2316. too.
  2317. The Repology site @url{} is a useful resource to identify
  2318. currently shipped versions of software in various operating systems, though
  2319. it does not cover all distros listed below.
  2320. @section Linux OS
  2321. For distributions with frequent, short-lifetime releases, the project will
  2322. aim to support all versions that are not end of life by their respective
  2323. vendors. For the purposes of identifying supported software versions, the
  2324. project will look at Fedora, Ubuntu, and openSUSE distros. Other short-
  2325. lifetime distros will be assumed to ship similar software versions.
  2326. For distributions with long-lifetime releases, the project will aim to support
  2327. the most recent major version at all times. Support for the previous major
  2328. version will be dropped 2 years after the new major version is released. For
  2329. the purposes of identifying supported software versions, the project will look
  2330. at RHEL, Debian, Ubuntu LTS, and SLES distros. Other long-lifetime distros will
  2331. be assumed to ship similar software versions.
  2332. @section Windows
  2333. The project supports building with current versions of the MinGW toolchain,
  2334. hosted on Linux.
  2335. @section macOS
  2336. The project supports building with the two most recent versions of macOS, with
  2337. the current homebrew package set available.
  2338. @section FreeBSD
  2339. The project aims to support the all the versions which are not end of life.
  2340. @section NetBSD
  2341. The project aims to support the most recent major version at all times. Support
  2342. for the previous major version will be dropped 2 years after the new major
  2343. version is released.
  2344. @section OpenBSD
  2345. The project aims to support the all the versions which are not end of life.
  2346. @node License
  2347. @appendix License
  2348. QEMU is a trademark of Fabrice Bellard.
  2349. QEMU is released under the
  2350. @url{,GNU General Public License},
  2351. version 2. Parts of QEMU have specific licenses, see file
  2352. @url{;a=blob_plain;f=LICENSE,LICENSE}.
  2353. @node Index
  2354. @appendix Index
  2355. @menu
  2356. * Concept Index::
  2357. * Function Index::
  2358. * Keystroke Index::
  2359. * Program Index::
  2360. * Data Type Index::
  2361. * Variable Index::
  2362. @end menu
  2363. @node Concept Index
  2364. @section Concept Index
  2365. This is the main index. Should we combine all keywords in one index? TODO
  2366. @printindex cp
  2367. @node Function Index
  2368. @section Function Index
  2369. This index could be used for command line options and monitor functions.
  2370. @printindex fn
  2371. @node Keystroke Index
  2372. @section Keystroke Index
  2373. This is a list of all keystrokes which have a special function
  2374. in system emulation.
  2375. @printindex ky
  2376. @node Program Index
  2377. @section Program Index
  2378. @printindex pg
  2379. @node Data Type Index
  2380. @section Data Type Index
  2381. This index could be used for qdev device names and options.
  2382. @printindex tp
  2383. @node Variable Index
  2384. @section Variable Index
  2385. @printindex vr
  2386. @bye