master
rst 1,554 lines 58.5 KB
Raw
1
2 Removed features
3 ================
4
5 What follows is a record of recently removed, formerly deprecated
6 features that serves as a record for users who have encountered
7 trouble after a recent upgrade.
8
9 System emulator command line arguments
10 --------------------------------------
11
12 ``-hdachs`` (removed in 2.12)
13 '''''''''''''''''''''''''''''
14
15 The geometry defined by ``-hdachs c,h,s,t`` should now be specified via
16 ``-device ide-hd,drive=dr,cyls=c,heads=h,secs=s,bios-chs-trans=t``
17 (together with ``-drive if=none,id=dr,...``).
18
19 ``-net channel`` (removed in 2.12)
20 ''''''''''''''''''''''''''''''''''
21
22 This option has been replaced by ``-net user,guestfwd=...``.
23
24 ``-net dump`` (removed in 2.12)
25 '''''''''''''''''''''''''''''''
26
27 ``-net dump[,vlan=n][,file=filename][,len=maxlen]`` has been replaced by
28 ``-object filter-dump,id=id,netdev=dev[,file=filename][,maxlen=maxlen]``.
29 Note that the new syntax works with netdev IDs instead of the old "vlan" hubs.
30
31 ``-no-kvm-pit`` (removed in 2.12)
32 '''''''''''''''''''''''''''''''''
33
34 This was just a dummy option that has been ignored, since the in-kernel PIT
35 cannot be disabled separately from the irqchip anymore. A similar effect
36 (which also disables the KVM IOAPIC) can be obtained with
37 ``-M kernel_irqchip=split``.
38
39 ``-tdf`` (removed in 2.12)
40 ''''''''''''''''''''''''''
41
42 There is no replacement, the ``-tdf`` option has just been ignored since the
43 behaviour that could be changed by this option in qemu-kvm is now the default
44 when using the KVM PIT. It still can be requested explicitly using
45 ``-global kvm-pit.lost_tick_policy=delay``.
46
47 ``-drive secs=s``, ``-drive heads=h`` & ``-drive cyls=c`` (removed in 3.0)
48 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
49
50 The drive geometry should now be specified via
51 ``-device ...,drive=dr,cyls=c,heads=h,secs=s`` (together with
52 ``-drive if=none,id=dr,...``).
53
54 ``-drive serial=``, ``-drive trans=`` & ``-drive addr=`` (removed in 3.0)
55 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
56
57 Use ``-device ...,drive=dr,serial=r,bios-chs-trans=t,addr=a`` instead
58 (together with ``-drive if=none,id=dr,...``).
59
60 ``-net ...,vlan=x`` (removed in 3.0)
61 ''''''''''''''''''''''''''''''''''''
62
63 The term "vlan" was very confusing for most users in this context (it's about
64 specifying a hub ID, not about IEEE 802.1Q or something similar), so this
65 has been removed. To connect one NIC frontend with a network backend, either
66 use ``-nic ...`` (e.g. for on-board NICs) or use ``-netdev ...,id=n`` together
67 with ``-device ...,netdev=n`` (for full control over pluggable NICs). To
68 connect multiple NICs or network backends via a hub device (which is what
69 vlan did), use ``-nic hubport,hubid=x,...`` or
70 ``-netdev hubport,id=n,hubid=x,...`` (with ``-device ...,netdev=n``) instead.
71
72 ``-no-kvm-irqchip`` (removed in 3.0)
73 ''''''''''''''''''''''''''''''''''''
74
75 Use ``-machine kernel_irqchip=off`` instead.
76
77 ``-no-kvm-pit-reinjection`` (removed in 3.0)
78 ''''''''''''''''''''''''''''''''''''''''''''
79
80 Use ``-global kvm-pit.lost_tick_policy=discard`` instead.
81
82 ``-balloon`` (removed in 3.1)
83 '''''''''''''''''''''''''''''
84
85 The ``-balloon virtio`` option has been replaced by ``-device virtio-balloon``.
86 The ``-balloon none`` option was a no-op and has no replacement.
87
88 ``-bootp`` (removed in 3.1)
89 '''''''''''''''''''''''''''
90
91 The ``-bootp /some/file`` argument is replaced by either
92 ``-netdev user,id=x,bootp=/some/file`` (for pluggable NICs, accompanied with
93 ``-device ...,netdev=x``), or ``-nic user,bootp=/some/file`` (for on-board NICs).
94 The new syntax allows different settings to be provided per NIC.
95
96 ``-redir`` (removed in 3.1)
97 '''''''''''''''''''''''''''
98
99 The ``-redir [tcp|udp]:hostport:[guestaddr]:guestport`` option is replaced
100 by either ``-netdev
101 user,id=x,hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport``
102 (for pluggable NICs, accompanied with ``-device ...,netdev=x``) or by the option
103 ``-nic user,hostfwd=[tcp|udp]:[hostaddr]:hostport-[guestaddr]:guestport``
104 (for on-board NICs). The new syntax allows different settings to be provided
105 per NIC.
106
107 ``-smb`` (removed in 3.1)
108 '''''''''''''''''''''''''
109
110 The ``-smb /some/dir`` argument is replaced by either
111 ``-netdev user,id=x,smb=/some/dir`` (for pluggable NICs, accompanied with
112 ``-device ...,netdev=x``), or ``-nic user,smb=/some/dir`` (for on-board NICs).
113 The new syntax allows different settings to be provided per NIC.
114
115 ``-tftp`` (removed in 3.1)
116 ''''''''''''''''''''''''''
117
118 The ``-tftp /some/dir`` argument is replaced by either
119 ``-netdev user,id=x,tftp=/some/dir`` (for pluggable NICs, accompanied with
120 ``-device ...,netdev=x``), or ``-nic user,tftp=/some/dir`` (for embedded NICs).
121 The new syntax allows different settings to be provided per NIC.
122
123 ``-localtime`` (removed in 3.1)
124 '''''''''''''''''''''''''''''''
125
126 Replaced by ``-rtc base=localtime``.
127
128 ``-nodefconfig`` (removed in 3.1)
129 '''''''''''''''''''''''''''''''''
130
131 Use ``-no-user-config`` instead.
132
133 ``-rtc-td-hack`` (removed in 3.1)
134 '''''''''''''''''''''''''''''''''
135
136 Use ``-rtc driftfix=slew`` instead.
137
138 ``-startdate`` (removed in 3.1)
139 '''''''''''''''''''''''''''''''
140
141 Replaced by ``-rtc base=date``.
142
143 ``-vnc ...,tls=...``, ``-vnc ...,x509=...`` & ``-vnc ...,x509verify=...`` (removed in 3.1)
144 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
145
146 The "tls-creds" option should be used instead to point to a "tls-creds-x509"
147 object created using "-object".
148
149 ``-mem-path`` fallback to RAM (removed in 5.0)
150 ''''''''''''''''''''''''''''''''''''''''''''''
151
152 If guest RAM allocation from file pointed by ``mem-path`` failed,
153 QEMU was falling back to allocating from RAM, which might have resulted
154 in unpredictable behavior since the backing file specified by the user
155 as ignored. Currently, users are responsible for making sure the backing storage
156 specified with ``-mem-path`` can actually provide the guest RAM configured with
157 ``-m`` and QEMU fails to start up if RAM allocation is unsuccessful.
158
159 ``-net ...,name=...`` (removed in 5.1)
160 ''''''''''''''''''''''''''''''''''''''
161
162 The ``name`` parameter of the ``-net`` option was a synonym
163 for the ``id`` parameter, which should now be used instead.
164
165 RISC-V firmware not booted by default (removed in 5.1)
166 ''''''''''''''''''''''''''''''''''''''''''''''''''''''
167
168 QEMU 5.1 changes the default behaviour from ``-bios none`` to ``-bios default``
169 for the RISC-V ``virt`` machine and ``sifive_u`` machine.
170
171 ``-numa node,mem=...`` (removed in 5.1)
172 '''''''''''''''''''''''''''''''''''''''
173
174 The parameter ``mem`` of ``-numa node`` was used to assign a part of guest RAM
175 to a NUMA node. But when using it, it's impossible to manage a specified RAM
176 chunk on the host side (like bind it to a host node, setting bind policy, ...),
177 so the guest ends up with the fake NUMA configuration with suboptiomal
178 performance.
179 However since 2014 there is an alternative way to assign RAM to a NUMA node
180 using parameter ``memdev``, which does the same as ``mem`` and adds
181 means to actually manage node RAM on the host side. Use parameter ``memdev``
182 with *memory-backend-ram* backend as replacement for parameter ``mem``
183 to achieve the same fake NUMA effect or a properly configured
184 *memory-backend-file* backend to actually benefit from NUMA configuration.
185 New machine versions (since 5.1) will not accept the option but it will still
186 work with old machine types. User can check the QAPI schema to see if the legacy
187 option is supported by looking at MachineInfo::numa-mem-supported property.
188
189 ``-numa`` node (without memory specified) (removed in 5.2)
190 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
191
192 Splitting RAM by default between NUMA nodes had the same issues as ``mem``
193 parameter with the difference that the role of the user plays QEMU using
194 implicit generic or board specific splitting rule.
195 Use ``memdev`` with *memory-backend-ram* backend or ``mem`` (if
196 it's supported by used machine type) to define mapping explicitly instead.
197 Users of existing VMs, wishing to preserve the same RAM distribution, should
198 configure it explicitly using ``-numa node,memdev`` options. Current RAM
199 distribution can be retrieved using HMP command ``info numa`` and if separate
200 memory devices (pc|nv-dimm) are present use ``info memory-device`` and subtract
201 device memory from output of ``info numa``.
202
203 ``-smp`` (invalid topologies) (removed in 5.2)
204 ''''''''''''''''''''''''''''''''''''''''''''''
205
206 CPU topology properties should describe whole machine topology including
207 possible CPUs.
208
209 However, historically it was possible to start QEMU with an incorrect topology
210 where *n* <= *sockets* * *cores* * *threads* < *maxcpus*,
211 which could lead to an incorrect topology enumeration by the guest.
212 Support for invalid topologies is removed, the user must ensure
213 topologies described with -smp include all possible cpus, i.e.
214 *sockets* * *cores* * *threads* = *maxcpus*.
215
216 ``-machine enforce-config-section=on|off`` (removed in 5.2)
217 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
218
219 The ``enforce-config-section`` property was replaced by the
220 ``-global migration.send-configuration={on|off}`` option.
221
222 ``-no-kvm`` (removed in 5.2)
223 ''''''''''''''''''''''''''''
224
225 The ``-no-kvm`` argument was a synonym for setting ``-machine accel=tcg``.
226
227 ``-realtime`` (removed in 6.0)
228 ''''''''''''''''''''''''''''''
229
230 The ``-realtime mlock=on|off`` argument has been replaced by the
231 ``-overcommit mem-lock=on|off`` argument.
232
233 ``-show-cursor`` option (removed in 6.0)
234 ''''''''''''''''''''''''''''''''''''''''
235
236 Use ``-display sdl,show-cursor=on``, ``-display gtk,show-cursor=on``
237 or ``-display default,show-cursor=on`` instead.
238
239 ``-tb-size`` option (removed in 6.0)
240 ''''''''''''''''''''''''''''''''''''
241
242 QEMU 5.0 introduced an alternative syntax to specify the size of the translation
243 block cache, ``-accel tcg,tb-size=``.
244
245 ``-usbdevice audio`` (removed in 6.0)
246 '''''''''''''''''''''''''''''''''''''
247
248 This option lacked the possibility to specify an audio backend device.
249 Use ``-device usb-audio`` now instead (and specify a corresponding USB
250 host controller or ``-usb`` if necessary).
251
252 ``-vnc acl`` (removed in 6.0)
253 '''''''''''''''''''''''''''''
254
255 The ``acl`` option to the ``-vnc`` argument has been replaced
256 by the ``tls-authz`` and ``sasl-authz`` options.
257
258 ``-mon ...,control=readline,pretty=on|off`` (removed in 6.0)
259 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
260
261 The ``pretty=on|off`` switch has no effect for HMP monitors and
262 its use is rejected.
263
264 ``-drive file=json:{...{'driver':'file'}}`` (removed in 6.0)
265 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
266
267 The 'file' driver for drives is no longer appropriate for character or host
268 devices and will only accept regular files (S_IFREG). The correct driver
269 for these file types is 'host_cdrom' or 'host_device' as appropriate.
270
271 Floppy controllers' drive properties (removed in 6.0)
272 '''''''''''''''''''''''''''''''''''''''''''''''''''''
273
274 Use ``-device floppy,...`` instead. When configuring onboard floppy
275 controllers
276 ::
277
278 -global isa-fdc.driveA=...
279 -global sysbus-fdc.driveA=...
280 -global SUNW,fdtwo.drive=...
281
282 become
283 ::
284
285 -device floppy,unit=0,drive=...
286
287 and
288 ::
289
290 -global isa-fdc.driveB=...
291 -global sysbus-fdc.driveB=...
292
293 become
294 ::
295
296 -device floppy,unit=1,drive=...
297
298 When plugging in a floppy controller
299 ::
300
301 -device isa-fdc,...,driveA=...
302
303 becomes
304 ::
305
306 -device isa-fdc,...
307 -device floppy,unit=0,drive=...
308
309 and
310 ::
311
312 -device isa-fdc,...,driveB=...
313
314 becomes
315 ::
316
317 -device isa-fdc,...
318 -device floppy,unit=1,drive=...
319
320 ``-drive`` with bogus interface type (removed in 6.0)
321 '''''''''''''''''''''''''''''''''''''''''''''''''''''
322
323 Drives with interface types other than ``if=none`` are for onboard
324 devices. Drives the board doesn't pick up can no longer be used with
325 -device. Use ``if=none`` instead.
326
327 ``-usbdevice ccid`` (removed in 6.0)
328 '''''''''''''''''''''''''''''''''''''
329
330 This option was undocumented and not used in the field.
331 Use ``-device usb-ccid`` instead.
332
333 ``-no-quit`` (removed in 7.0)
334 '''''''''''''''''''''''''''''
335
336 The ``-no-quit`` was a synonym for ``-display ...,window-close=off`` which
337 should be used instead.
338
339 ``--enable-fips`` (removed in 7.1)
340 ''''''''''''''''''''''''''''''''''
341
342 This option restricted usage of certain cryptographic algorithms when
343 the host is operating in FIPS mode.
344
345 If FIPS compliance is required, QEMU should be built with the ``libgcrypt``
346 or ``gnutls`` library enabled as a cryptography provider.
347
348 Neither the ``nettle`` library, or the built-in cryptography provider are
349 supported on FIPS enabled hosts.
350
351 ``-writeconfig`` (removed in 7.1)
352 '''''''''''''''''''''''''''''''''
353
354 The ``-writeconfig`` option was not able to serialize the entire contents
355 of the QEMU command line. It is thus considered a failed experiment
356 and removed without a replacement.
357
358 ``loaded`` property of secret and TLS credential objects (removed in 9.2)
359 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
360
361 The ``loaded=on`` option in the command line or QMP ``object-add`` either had
362 no effect (if ``loaded`` was the last option) or caused options to be
363 effectively ignored as if they were not given. The property is therefore
364 useless and has been removed.
365
366 ``opened`` property of ``rng-*`` objects (removed in 7.1)
367 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''
368
369 The ``opened=on`` option in the command line or QMP ``object-add`` either had
370 no effect (if ``opened`` was the last option) or caused errors. The property
371 is therefore useless and should simply be removed.
372
373 ``-display sdl,window_close=...`` (removed in 7.1)
374 ''''''''''''''''''''''''''''''''''''''''''''''''''
375
376 Use ``-display sdl,window-close=...`` instead (i.e. with a minus instead of
377 an underscore between "window" and "close").
378
379 ``-alt-grab`` and ``-display sdl,alt_grab=on`` (removed in 7.1)
380 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
381
382 Use ``-display sdl,grab-mod=lshift-lctrl-lalt`` instead.
383
384 ``-ctrl-grab`` and ``-display sdl,ctrl_grab=on`` (removed in 7.1)
385 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
386
387 Use ``-display sdl,grab-mod=rctrl`` instead.
388
389 ``-sdl`` (removed in 7.1)
390 '''''''''''''''''''''''''
391
392 Use ``-display sdl`` instead.
393
394 ``-curses`` (removed in 7.1)
395 ''''''''''''''''''''''''''''
396
397 Use ``-display curses`` instead.
398
399 Creating sound card devices using ``-soundhw`` (removed in 7.1)
400 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
401
402 Sound card devices should be created using ``-device`` or ``-audio``.
403 The exception is ``pcspk`` which can be activated using ``-machine
404 pcspk-audiodev=<name>``.
405
406 ``-watchdog`` (removed in 7.2)
407 ''''''''''''''''''''''''''''''
408
409 Use ``-device`` instead.
410
411 Hexadecimal sizes with scaling multipliers (removed in 8.0)
412 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
413
414 Input parameters that take a size value should only use a size suffix
415 (such as 'k' or 'M') when the base is written in decimal, and not when
416 the value is hexadecimal. That is, '0x20M' should be written either as
417 '32M' or as '0x2000000'.
418
419 ``-chardev`` backend aliases ``tty`` and ``parport`` (removed in 8.0)
420 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
421
422 ``tty`` and ``parport`` used to be aliases for ``serial`` and ``parallel``
423 respectively. The actual backend names should be used instead.
424
425 ``-drive if=none`` for the sifive_u OTP device (removed in 8.0)
426 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
427
428 Use ``-drive if=pflash`` to configure the OTP device of the sifive_u
429 RISC-V machine instead.
430
431 ``-spice password=string`` (removed in 8.0)
432 '''''''''''''''''''''''''''''''''''''''''''
433
434 This option was insecure because the SPICE password remained visible in
435 the process listing. This was replaced by the new ``password-secret``
436 option which lets the password be securely provided on the command
437 line using a ``secret`` object instance.
438
439 ``QEMU_AUDIO_`` environment variables and ``-audio-help`` (removed in 8.2)
440 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
441
442 The ``-audiodev`` and ``-audio`` command line options are now the only
443 way to specify audio backend settings.
444
445 Using ``-audiodev`` to define the default audio backend (removed in 8.2)
446 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
447
448 If no audiodev property is specified, previous versions would use the
449 first ``-audiodev`` command line option as a fallback. Starting with
450 version 8.2, audio backends created with ``-audiodev`` will only be
451 used by clients (sound cards, machines with embedded sound hardware, VNC)
452 that refer to it in an ``audiodev=`` property.
453
454 In order to configure a default audio backend, use the ``-audio``
455 command line option without specifying a ``model``; while previous
456 versions of QEMU required a model, starting with version 8.2
457 QEMU does not require a model and will not create any sound card
458 in this case.
459
460 Note that the default audio backend must be configured on the command
461 line if the ``-nodefaults`` options is used.
462
463 ``-no-hpet`` (removed in 9.0)
464 '''''''''''''''''''''''''''''
465
466 The HPET setting has been turned into a machine property.
467 Use ``-machine hpet=off`` instead.
468
469 ``-no-acpi`` (removed in 9.0)
470 '''''''''''''''''''''''''''''
471
472 The ``-no-acpi`` setting has been turned into a machine property.
473 Use ``-machine acpi=off`` instead.
474
475 ``-async-teardown`` (removed in 9.0)
476 ''''''''''''''''''''''''''''''''''''
477
478 Use ``-run-with async-teardown=on`` instead.
479
480 ``-chroot`` (removed in 9.0)
481 ''''''''''''''''''''''''''''
482
483 Use ``-run-with chroot=dir`` instead.
484
485 ``-singlestep`` (removed in 9.0)
486 ''''''''''''''''''''''''''''''''
487
488 The ``-singlestep`` option has been turned into an accelerator property,
489 and given a name that better reflects what it actually does.
490 Use ``-accel tcg,one-insn-per-tb=on`` instead.
491
492 ``-smp`` ("parameter=0" SMP configurations) (removed in 9.0)
493 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
494
495 Specified CPU topology parameters must be greater than zero.
496
497 In the SMP configuration, users should either provide a CPU topology
498 parameter with a reasonable value (greater than zero) or just omit it
499 and QEMU will compute the missing value.
500
501 However, historically it was implicitly allowed for users to provide
502 a parameter with zero value, which is meaningless and could also possibly
503 cause unexpected results in the -smp parsing. So support for this kind of
504 configurations (e.g. -smp 8,sockets=0) is removed since 9.0, users have
505 to ensure that all the topology members described with -smp are greater
506 than zero.
507
508 ``-global migration.decompress-error-check`` (removed in 9.1)
509 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
510
511 Removed along with the ``compression`` migration capability.
512
513 ``-device virtio-blk,scsi=on|off`` (removed in 9.1)
514 '''''''''''''''''''''''''''''''''''''''''''''''''''
515
516 The virtio-blk SCSI passthrough feature is a legacy VIRTIO feature. VIRTIO 1.0
517 and later do not support it because the virtio-scsi device was introduced for
518 full SCSI support. Use virtio-scsi instead when SCSI passthrough is required.
519
520 ``-fsdev proxy`` and ``-virtfs proxy`` (removed in 9.2)
521 '''''''''''''''''''''''''''''''''''''''''''''''''''''''
522
523 The 9p ``proxy`` filesystem backend driver was originally developed to
524 enhance security by dispatching low level filesystem operations from 9p
525 server (QEMU process) over to a separate process (the virtfs-proxy-helper
526 binary). However the proxy backend was much slower than the local backend,
527 didn't see any development in years, and showed to be less secure,
528 especially due to the fact that its helper daemon must be run as root.
529
530 Use ``local``, possibly mapping permissions et al by using its 'mapped'
531 security model option, or switch to ``virtiofs``. The virtiofs daemon
532 ``virtiofsd`` uses vhost to eliminate the high latency costs of the 9p
533 ``proxy`` backend.
534
535 ``-portrait`` and ``-rotate`` (removed in 9.2)
536 ''''''''''''''''''''''''''''''''''''''''''''''
537
538 The ``-portrait`` and ``-rotate`` options were documented as only
539 working with the PXA LCD device, and all the machine types using
540 that display device were removed in 9.2, so these options also
541 have been dropped.
542
543 These options were intended to simulate a mobile device being
544 rotated by the user, and had three effects:
545
546 * the display output was rotated by 90, 180 or 270 degrees
547 * the mouse/trackpad input was rotated the opposite way
548 * the machine model would signal to the guest about its
549 orientation
550
551 Of these three things, the input-rotation was coded without being
552 restricted to boards which supported the full set of device-rotation
553 handling, so in theory the options were usable on other machine models
554 to produce an odd effect (rotating input but not display output). But
555 this was never intended or documented behaviour, so we have dropped
556 the options along with the machine models they were intended for.
557
558 ``-runas`` (removed in 10.0)
559 ''''''''''''''''''''''''''''
560
561 Use ``-run-with user=..`` instead.
562
563 ``-old-param`` option for booting Arm kernels via param_struct (removed in 10.2)
564 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
565
566 The ``-old-param`` command line option was specific to Arm targets:
567 it was used when directly booting a guest kernel to pass it the
568 command line and other information via the old ``param_struct`` ABI,
569 rather than the newer ATAGS or DTB mechanisms. This option was only
570 ever needed to support ancient kernels on some old board types
571 like the ``akita`` or ``terrier``; it has been deprecated in the
572 kernel since 2001. None of the board types QEMU supports need
573 ``param_struct`` support, so this option has been removed.
574
575 32-bit host operating systems (removed in 11.0)
576 '''''''''''''''''''''''''''''''''''''''''''''''
577
578 Keeping 32-bit host support alive was a substantial burden for the
579 QEMU project. Thus QEMU dropped all support for all 32-bit host systems.
580
581
582 User-mode emulator command line arguments
583 -----------------------------------------
584
585 ``-singlestep`` (removed in 9.0)
586 ''''''''''''''''''''''''''''''''
587
588 The ``-singlestep`` option has been given a name that better reflects
589 what it actually does. For both linux-user and bsd-user, use the
590 ``-one-insn-per-tb`` option instead.
591
592 ``-p`` (removed in 10.2)
593 ''''''''''''''''''''''''
594
595 The ``-p`` option pretends to control the host page size. However,
596 it is not possible to change the host page size; we stopped trying
597 to do anything with the option except print a warning from 9.0,
598 and now the option is removed entirely.
599
600
601 QEMU Machine Protocol (QMP) commands
602 ------------------------------------
603
604 ``block-dirty-bitmap-add`` "autoload" parameter (removed in 4.2)
605 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
606
607 The "autoload" parameter has been ignored since 2.12.0. All bitmaps
608 are automatically loaded from qcow2 images.
609
610 ``cpu-add`` (removed in 5.2)
611 ''''''''''''''''''''''''''''
612
613 Use ``device_add`` for hotplugging vCPUs instead of ``cpu-add``. See
614 documentation of ``query-hotpluggable-cpus`` for additional details.
615
616 ``change`` (removed in 6.0)
617 '''''''''''''''''''''''''''
618
619 Use ``blockdev-change-medium`` or ``change-vnc-password`` or
620 ``display-update`` instead.
621
622 ``query-events`` (removed in 6.0)
623 '''''''''''''''''''''''''''''''''
624
625 The ``query-events`` command has been superseded by the more powerful
626 and accurate ``query-qmp-schema`` command.
627
628 ``migrate_set_cache_size`` and ``query-migrate-cache-size`` (removed in 6.0)
629 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
630
631 Use ``migrate_set_parameter`` and ``info migrate_parameters`` instead.
632
633 ``migrate_set_downtime`` and ``migrate_set_speed`` (removed in 6.0)
634 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
635
636 Use ``migrate_set_parameter`` instead.
637
638 ``query-cpus`` (removed in 6.0)
639 '''''''''''''''''''''''''''''''
640
641 The ``query-cpus`` command is replaced by the ``query-cpus-fast`` command.
642
643 ``query-cpus-fast`` ``arch`` output member (removed in 6.0)
644 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
645
646 The ``arch`` output member of the ``query-cpus-fast`` command is
647 replaced by the ``target`` output member.
648
649 chardev client socket with ``wait`` option (removed in 6.0)
650 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
651
652 Character devices creating sockets in client mode should not specify
653 the 'wait' field, which is only applicable to sockets in server mode
654
655 ``query-named-block-nodes`` result ``encryption_key_missing`` (removed in 6.0)
656 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
657
658 Removed with no replacement.
659
660 ``query-block`` result ``inserted.encryption_key_missing`` (removed in 6.0)
661 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
662
663 Removed with no replacement.
664
665 ``query-named-block-nodes`` and ``query-block`` result dirty-bitmaps[i].status (removed in 6.0)
666 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
667
668 The ``status`` field of the ``BlockDirtyInfo`` structure, returned by
669 these commands is removed. Two new boolean fields, ``recording`` and
670 ``busy`` effectively replace it.
671
672 ``query-block`` result field ``dirty-bitmaps`` (removed in 6.0)
673 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
674
675 The ``dirty-bitmaps`` field of the ``BlockInfo`` structure, returned by
676 the query-block command is itself now removed. The ``dirty-bitmaps``
677 field of the ``BlockDeviceInfo`` struct should be used instead, which is the
678 type of the ``inserted`` field in query-block replies, as well as the
679 type of array items in query-named-block-nodes.
680
681 ``object-add`` option ``props`` (removed in 6.0)
682 ''''''''''''''''''''''''''''''''''''''''''''''''
683
684 Specify the properties for the object as top-level arguments instead.
685
686 ``query-sgx`` return value member ``section-size`` (removed in 8.0)
687 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
688
689 Member ``section-size`` in the return value of ``query-sgx``
690 was superseded by ``sections``.
691
692
693 ``query-sgx-capabilities`` return value member ``section-size`` (removed in 8.0)
694 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
695
696 Member ``section-size`` in the return value of ``query-sgx-capabilities``
697 was superseded by ``sections``.
698
699 ``query-migrate`` return value member ``skipped`` (removed in 9.1)
700 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
701
702 Member ``skipped`` of the ``MigrationRAMStats`` struct hasn't been used
703 for more than 10 years. Removed with no replacement.
704
705 ``migrate`` command option ``inc`` (removed in 9.1)
706 '''''''''''''''''''''''''''''''''''''''''''''''''''
707
708 Use blockdev-mirror with NBD instead. See "QMP invocation for live
709 storage migration with ``blockdev-mirror`` + NBD" in
710 docs/interop/live-block-operations.rst for a detailed explanation.
711
712 ``migrate`` command option ``blk`` (removed in 9.1)
713 '''''''''''''''''''''''''''''''''''''''''''''''''''
714
715 Use blockdev-mirror with NBD instead. See "QMP invocation for live
716 storage migration with ``blockdev-mirror`` + NBD" in
717 docs/interop/live-block-operations.rst for a detailed explanation.
718
719 ``migrate`` command with file-based ``fd:`` URI (removed in 11.0)
720 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
721
722 In order to reduce ambiguity, the ``fd:`` URI usage of providing a
723 file descriptor to a plain file has been removed in favor of
724 explicitly using the ``file:`` URI with the file descriptor being
725 passed as an ``fdset``. Refer to the ``add-fd`` command documentation
726 for details on the ``fdset`` usage.
727
728 ``migrate-set-capabilities`` ``block`` option (removed in 9.1)
729 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
730
731 Block migration has been removed. For a replacement, see "QMP
732 invocation for live storage migration with ``blockdev-mirror`` + NBD"
733 in docs/interop/live-block-operations.rst.
734
735 ``migrate-set-capabilities`` ``zero-blocks`` option (removed in 11.0)
736 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
737
738 The ``zero-blocks`` capability was accidentally left behind when
739 block migration capability got removed in 9.1. Removed with no
740 replacement.
741
742 ``migrate-set-parameter`` ``compress-level`` option (removed in 9.1)
743 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
744
745 Use ``multifd-zlib-level`` or ``multifd-zstd-level`` instead.
746
747 ``migrate-set-parameter`` ``compress-threads`` option (removed in 9.1)
748 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
749
750 Use ``multifd-channels`` instead.
751
752 ``migrate-set-parameter`` ``compress-wait-thread`` option (removed in 9.1)
753 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
754
755 Removed with no replacement.
756
757 ``migrate-set-parameter`` ``decompress-threads`` option (removed in 9.1)
758 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
759
760 Use ``multifd-channels`` instead.
761
762 ``migrate-set-capability`` ``compress`` option (removed in 9.1)
763 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
764
765 Use ``multifd-compression`` instead.
766
767 Incorrectly typed ``device_add`` arguments (removed in 9.2)
768 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
769
770 Due to shortcomings in the internal implementation of ``device_add``,
771 QEMU used to incorrectly accept certain invalid arguments. Any object
772 or list arguments were silently ignored. Other argument types were not
773 checked, but an implicit conversion happened, so that e.g. string
774 values could be assigned to integer device properties or vice versa.
775
776 ``query-migrationthreads`` (removed in 11.0)
777 ''''''''''''''''''''''''''''''''''''''''''''
778
779 Removed with no replacement, as it reported only a limited set of
780 threads (for example, it only reported source side of multifd threads,
781 without reporting any destination threads, or non-multifd source
782 threads). For debugging purpose, please use ``-name
783 $VM,debug-threads=on`` instead.
784
785 ``migrate`` argument ``detach`` (removed in 11.0)
786 ''''''''''''''''''''''''''''''''''''''''''''''''''
787
788 This argument has always been ignored.
789
790 QEMU Machine Protocol (QMP) events
791 ----------------------------------
792
793 ``MEM_UNPLUG_ERROR`` (removed in 9.1)
794 '''''''''''''''''''''''''''''''''''''
795
796 MEM_UNPLUG_ERROR has been replaced by the more generic ``DEVICE_UNPLUG_GUEST_ERROR`` event.
797
798 ``vcpu`` trace events (removed in 9.1)
799 ''''''''''''''''''''''''''''''''''''''
800
801 The ability to instrument QEMU helper functions with vCPU-aware trace
802 points was removed in 7.0.
803
804
805 Human Monitor Protocol (HMP) commands
806 -------------------------------------
807
808 ``usb_add`` and ``usb_remove`` (removed in 2.12)
809 ''''''''''''''''''''''''''''''''''''''''''''''''
810
811 Replaced by ``device_add`` and ``device_del`` (use ``device_add help`` for a
812 list of available devices).
813
814 ``host_net_add`` and ``host_net_remove`` (removed in 2.12)
815 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
816
817 Replaced by ``netdev_add`` and ``netdev_del``.
818
819 The ``hub_id`` parameter of ``hostfwd_add`` / ``hostfwd_remove`` (removed in 5.0)
820 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
821
822 The ``[hub_id name]`` parameter tuple of the 'hostfwd_add' and
823 'hostfwd_remove' HMP commands has been replaced by ``netdev_id``.
824
825 ``cpu-add`` (removed in 5.2)
826 ''''''''''''''''''''''''''''
827
828 Use ``device_add`` for hotplugging vCPUs instead of ``cpu-add``. See
829 documentation of ``query-hotpluggable-cpus`` for additional details.
830
831 ``change vnc TARGET`` (removed in 6.0)
832 ''''''''''''''''''''''''''''''''''''''
833
834 No replacement. The ``change vnc password`` and ``change DEVICE MEDIUM``
835 commands are not affected.
836
837 ``acl_show``, ``acl_reset``, ``acl_policy``, ``acl_add``, ``acl_remove`` (removed in 6.0)
838 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
839
840 The ``acl_show``, ``acl_reset``, ``acl_policy``, ``acl_add``, and
841 ``acl_remove`` commands were removed with no replacement. Authorization
842 for VNC should be performed using the pluggable QAuthZ objects.
843
844 ``migrate-set-cache-size`` and ``info migrate-cache-size`` (removed in 6.0)
845 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
846
847 Use ``migrate-set-parameters`` and ``info migrate-parameters`` instead.
848
849 ``migrate_set_downtime`` and ``migrate_set_speed`` (removed in 6.0)
850 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
851
852 Use ``migrate-set-parameters`` instead.
853
854 ``info cpustats`` (removed in 6.1)
855 ''''''''''''''''''''''''''''''''''
856
857 This command didn't produce any output already. Removed with no replacement.
858
859 ``singlestep`` (removed in 9.0)
860 '''''''''''''''''''''''''''''''
861
862 The ``singlestep`` command has been replaced by the ``one-insn-per-tb``
863 command, which has the same behaviour but a less misleading name.
864
865 ``migrate`` command ``-i`` option (removed in 9.1)
866 ''''''''''''''''''''''''''''''''''''''''''''''''''
867
868 Use blockdev-mirror with NBD instead. See "QMP invocation for live
869 storage migration with ``blockdev-mirror`` + NBD" in
870 docs/interop/live-block-operations.rst for a detailed explanation.
871
872 ``migrate`` command ``-b`` option (removed in 9.1)
873 ''''''''''''''''''''''''''''''''''''''''''''''''''
874
875 Use blockdev-mirror with NBD instead. See "QMP invocation for live
876 storage migration with ``blockdev-mirror`` + NBD" in
877 docs/interop/live-block-operations.rst for a detailed explanation.
878
879 ``migrate_set_capability`` ``block`` option (removed in 9.1)
880 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
881
882 Block migration has been removed. For a replacement, see "QMP
883 invocation for live storage migration with ``blockdev-mirror`` + NBD"
884 in docs/interop/live-block-operations.rst.
885
886 ``migrate_set_parameter`` ``compress-level`` option (removed in 9.1)
887 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
888
889 Use ``multifd-zlib-level`` or ``multifd-zstd-level`` instead.
890
891 ``migrate_set_parameter`` ``compress-threads`` option (removed in 9.1)
892 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
893
894 Use ``multifd-channels`` instead.
895
896 ``migrate_set_parameter`` ``compress-wait-thread`` option (removed in 9.1)
897 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
898
899 Removed with no replacement.
900
901 ``migrate_set_parameter`` ``decompress-threads`` option (removed in 9.1)
902 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
903
904 Use ``multifd-channels`` instead.
905
906 ``migrate_set_capability`` ``compress`` option (removed in 9.1)
907 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
908
909 Use ``multifd-compression`` instead.
910
911 Host Architectures
912 ------------------
913
914 System emulation on 32-bit Windows hosts (removed in 9.0)
915 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''
916
917 Windows 11 has no support for 32-bit host installs, and Windows 10 did
918 not support new 32-bit installs, only upgrades. 32-bit Windows support
919 has now been dropped by the MSYS2 project. QEMU also is deprecating
920 and dropping support for 32-bit x86 host deployments in
921 general. 32-bit Windows is therefore no longer a supported host for
922 QEMU. Since all recent x86 hardware from the past >10 years is
923 capable of the 64-bit x86 extensions, a corresponding 64-bit OS should
924 be used instead.
925
926 32-bit hosts for 64-bit guests (removed in 10.0)
927 ''''''''''''''''''''''''''''''''''''''''''''''''
928
929 In general, 32-bit hosts cannot support the memory space or atomicity
930 requirements of 64-bit guests. Prior to 10.0, QEMU attempted to
931 work around the atomicity issues in system mode by running all vCPUs
932 in a single thread context; in user mode atomicity was simply broken.
933 From 10.0, QEMU has disabled configuration of 64-bit guests on 32-bit hosts.
934
935 32-bit MIPS (removed in 10.2)
936 '''''''''''''''''''''''''''''
937
938 Debian 12 "Bookworm" removed support for 32-bit MIPS, making it hard to
939 maintain our cross-compilation CI tests of the architecture.
940
941 64-bit MIPS (removed in 11.1)
942 '''''''''''''''''''''''''''''
943
944 Debian 13 "Trixie" removed support for MIPS, making it hard to maintain our
945 cross-compilation CI tests of the architecture.
946
947 32-bit PPC (removed in 10.2)
948 ''''''''''''''''''''''''''''
949
950 The QEMU project no longer supports 32-bit host builds.
951
952 Guest Emulator ISAs
953 -------------------
954
955 RISC-V ISA privilege specification version 1.09.1 (removed in 5.1)
956 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
957
958 The RISC-V ISA privilege specification version 1.09.1 has been removed.
959 QEMU supports both the newer version 1.10.0 and the ratified version 1.11.0, these
960 should be used instead of the 1.09.1 version.
961
962 System emulator CPUS
963 --------------------
964
965 KVM guest support on 32-bit Arm hosts (removed in 5.2)
966 ''''''''''''''''''''''''''''''''''''''''''''''''''''''
967
968 The Linux kernel has dropped support for allowing 32-bit Arm systems
969 to host KVM guests as of the 5.7 kernel, and was thus removed from QEMU
970 as well. Running 32-bit guests on a 64-bit Arm host remains supported.
971
972 RISC-V ISA Specific CPUs (removed in 5.1)
973 '''''''''''''''''''''''''''''''''''''''''
974
975 The RISC-V cpus with the ISA version in the CPU name have been removed. The
976 four CPUs are: ``rv32gcsu-v1.9.1``, ``rv32gcsu-v1.10.0``, ``rv64gcsu-v1.9.1`` and
977 ``rv64gcsu-v1.10.0``. Instead the version can be specified via the CPU ``priv_spec``
978 option when using the ``rv32`` or ``rv64`` CPUs.
979
980 RISC-V no MMU CPUs (removed in 5.1)
981 '''''''''''''''''''''''''''''''''''
982
983 The RISC-V no MMU cpus have been removed. The two CPUs: ``rv32imacu-nommu`` and
984 ``rv64imacu-nommu`` can no longer be used. Instead the MMU status can be specified
985 via the CPU ``mmu`` option when using the ``rv32`` or ``rv64`` CPUs.
986
987 ``compat`` property of server class POWER CPUs (removed in 6.0)
988 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
989
990 The ``max-cpu-compat`` property of the ``pseries`` machine type should be used
991 instead.
992
993 ``moxie`` CPU (removed in 6.1)
994 ''''''''''''''''''''''''''''''
995
996 Nobody was using this CPU emulation in QEMU, and there were no test images
997 available to make sure that the code is still working, so it has been removed
998 without replacement.
999
1000 ``lm32`` CPUs (removed in 6.1)
1001 ''''''''''''''''''''''''''''''
1002
1003 The only public user of this architecture was the milkymist project,
1004 which has been dead for years; there was never an upstream Linux
1005 port. Removed without replacement.
1006
1007 ``unicore32`` CPUs (removed in 6.1)
1008 '''''''''''''''''''''''''''''''''''
1009
1010 Support for this CPU was removed from the upstream Linux kernel, and
1011 there is no available upstream toolchain to build binaries for it.
1012 Removed without replacement.
1013
1014 x86 ``Icelake-Client`` CPU (removed in 7.1)
1015 '''''''''''''''''''''''''''''''''''''''''''
1016
1017 There isn't ever Icelake Client CPU, it is some wrong and imaginary one.
1018 Use ``Icelake-Server`` instead.
1019
1020 Nios II CPU (removed in 9.1)
1021 ''''''''''''''''''''''''''''
1022
1023 QEMU Nios II architecture was orphan; Intel has EOL'ed the Nios II
1024 processor IP (see `Intel discontinuance notification`_).
1025
1026 CRIS CPU architecture (removed in 9.2)
1027 ''''''''''''''''''''''''''''''''''''''
1028
1029 The CRIS architecture was pulled from Linux in 4.17 and the compiler
1030 was no longer packaged in any distro making it harder to run the
1031 ``check-tcg`` tests.
1032
1033 RISC-V 'any' CPU type ``-cpu any`` (removed in 9.2)
1034 '''''''''''''''''''''''''''''''''''''''''''''''''''
1035
1036 The 'any' CPU type was introduced back in 2018 and was around since the
1037 initial RISC-V QEMU port. Its usage was always been unclear: users don't know
1038 what to expect from a CPU called 'any', and in fact the CPU does not do anything
1039 special that isn't already done by the default CPUs rv32/rv64.
1040
1041 Power8E and Power8NVL CPUs (removed in 11.1)
1042 ''''''''''''''''''''''''''''''''''''''''''''
1043
1044 The Power8E and Power8NVL variants of Power8 are not really useful anymore
1045 in qemu, and are old and unmaintained.
1046 Hence, the CPUs as well as corresponding Power8NVL and Power8E PnvChips have
1047 been removed
1048
1049 System accelerators
1050 -------------------
1051
1052 Userspace local APIC with KVM (x86, removed in 8.0)
1053 '''''''''''''''''''''''''''''''''''''''''''''''''''
1054
1055 ``-M kernel-irqchip=off`` cannot be used on KVM if the CPU model includes
1056 a local APIC. The ``split`` setting is supported, as is using ``-M
1057 kernel-irqchip=off`` when the CPU does not have a local APIC.
1058
1059 MIPS "Trap-and-Emulate" KVM support (removed in 8.0)
1060 ''''''''''''''''''''''''''''''''''''''''''''''''''''
1061
1062 The MIPS "Trap-and-Emulate" KVM host and guest support was removed
1063 from Linux in 2021, and is not supported anymore by QEMU either.
1064
1065 HAXM (``-accel hax``) (removed in 8.2)
1066 ''''''''''''''''''''''''''''''''''''''
1067
1068 The HAXM project has been retired (see https://github.com/intel/haxm#status).
1069 Use "whpx" (on Windows) or "hvf" (on macOS) instead.
1070
1071 System emulator machines
1072 ------------------------
1073
1074 Versioned machine types (aarch64, arm, i386, m68k, ppc64, s390x, x86_64)
1075 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1076
1077 In accordance with our versioned machine type deprecation policy, all machine
1078 types with version |VER_MACHINE_DELETION_VERSION|, or older, have been
1079 removed.
1080
1081 ``s390-virtio`` (removed in 2.6)
1082 ''''''''''''''''''''''''''''''''
1083
1084 Use the ``s390-ccw-virtio`` machine instead.
1085
1086 The m68k ``dummy`` machine (removed in 2.9)
1087 '''''''''''''''''''''''''''''''''''''''''''
1088
1089 Use the ``none`` machine with the ``loader`` device instead.
1090
1091 ``xlnx-ep108`` (removed in 3.0)
1092 '''''''''''''''''''''''''''''''
1093
1094 The EP108 was an early access development board that is no longer used.
1095 Use the ``xlnx-zcu102`` machine instead.
1096
1097 ``spike_v1.9.1`` and ``spike_v1.10`` (removed in 5.1)
1098 '''''''''''''''''''''''''''''''''''''''''''''''''''''
1099
1100 The version specific Spike machines have been removed in favour of the
1101 generic ``spike`` machine. If you need to specify an older version of the RISC-V
1102 spec you can use the ``-cpu rv64gcsu,priv_spec=v1.10.0`` command line argument.
1103
1104 mips ``r4k`` platform (removed in 5.2)
1105 ''''''''''''''''''''''''''''''''''''''
1106
1107 This machine type was very old and unmaintained. Users should use the ``malta``
1108 machine type instead.
1109
1110 mips ``fulong2e`` machine alias (removed in 6.0)
1111 ''''''''''''''''''''''''''''''''''''''''''''''''
1112
1113 This machine has been renamed ``fuloong2e``.
1114
1115 Raspberry Pi ``raspi2`` and ``raspi3`` machines (removed in 6.2)
1116 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1117
1118 The Raspberry Pi machines come in various models (A, A+, B, B+). To be able
1119 to distinguish which model QEMU is implementing, the ``raspi2`` and ``raspi3``
1120 machines have been renamed ``raspi2b`` and ``raspi3b``.
1121
1122 Aspeed ``swift-bmc`` machine (removed in 7.0)
1123 '''''''''''''''''''''''''''''''''''''''''''''
1124
1125 This machine was removed because it was unused. Alternative AST2500 based
1126 OpenPOWER machines are ``witherspoon-bmc`` and ``romulus-bmc``.
1127
1128 ppc ``taihu`` machine (removed in 7.2)
1129 ''''''''''''''''''''''''''''''''''''''
1130
1131 This machine was removed because it was partially emulated and 405
1132 machines are very similar. Use the ``ref405ep`` machine instead.
1133
1134 Nios II ``10m50-ghrd`` and ``nios2-generic-nommu`` machines (removed in 9.1)
1135 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1136
1137 The Nios II architecture was orphan.
1138
1139 ``shix`` (removed in 9.2)
1140 '''''''''''''''''''''''''
1141
1142 The machine was unmaintained.
1143
1144 Arm machines ``akita``, ``borzoi``, ``cheetah``, ``connex``, ``mainstone``, ``n800``, ``n810``, ``spitz``, ``terrier``, ``tosa``, ``verdex``, ``z2`` (removed in 9.2)
1145 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1146
1147 QEMU included models of some machine types where the QEMU code that
1148 emulates their SoCs was very old and unmaintained. This code was
1149 blocking our ability to move forward with various changes across
1150 the codebase, and over many years nobody has been interested in
1151 trying to modernise it. We don't expect any of these machines to have
1152 a large number of users, because they're all modelling hardware that
1153 has now passed away into history. We are therefore dropping support
1154 for all machine types using the PXA2xx and OMAP2 SoCs. We are also
1155 dropping the ``cheetah`` OMAP1 board, because we don't have any
1156 test images for it and don't know of anybody who does.
1157
1158 Aspeed ``tacoma-bmc`` machine (removed in 10.0)
1159 '''''''''''''''''''''''''''''''''''''''''''''''
1160
1161 The ``tacoma-bmc`` machine was removed because it didn't bring much
1162 compared to the ``rainier-bmc`` machine. Also, the ``tacoma-bmc`` was
1163 a board used for bring up of the AST2600 SoC that never left the
1164 labs. It can be easily replaced by the ``rainier-bmc`` machine, which
1165 was the actual final product, or by the ``ast2600-evb`` with some
1166 tweaks.
1167
1168 ppc ``ref405ep`` machine (removed in 10.0)
1169 ''''''''''''''''''''''''''''''''''''''''''
1170
1171 This machine was removed because PPC 405 CPU have no known users,
1172 firmware images are not available, OpenWRT dropped support in 2019,
1173 U-Boot in 2017, and Linux in 2024.
1174
1175 Big-Endian variants of ``petalogix-ml605`` and ``xlnx-zynqmp-pmu`` machines (removed in 10.1)
1176 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1177
1178 Both the MicroBlaze ``petalogix-ml605`` and ``xlnx-zynqmp-pmu`` machines
1179 were added for little endian CPUs. Big endian support was never tested
1180 and likely never worked. Starting with QEMU v10.1, the machines are now
1181 only available as little-endian machines.
1182
1183 Mips ``mipssim`` machine (removed in 10.2)
1184 ''''''''''''''''''''''''''''''''''''''''''
1185
1186 Linux dropped support for this virtual machine type in kernel v3.7, and
1187 there was also no binary available online to use with that board.
1188
1189 Arm ``ast2700a0-evb`` machine (removed in 11.0)
1190 '''''''''''''''''''''''''''''''''''''''''''''''
1191
1192 The ``ast2700a0-evb`` machine represents the first revision of the AST2700
1193 and serves as the initial engineering sample rather than a production version.
1194 A newer revision, A1, is now supported, and the ``ast2700a1-evb`` should
1195 replace the older A0 version.
1196
1197 Arm ``highbank`` and ``midway`` machines (removed in 11.0)
1198 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1199
1200 There were no known users left for these machines. If you just want to
1201 boot a Cortex-A15 or Cortex-A9 Linux, use the ``virt`` machine instead.
1202
1203 Arm ``sonorapass-bmc`` machine (removed in 11.1)
1204 ''''''''''''''''''''''''''''''''''''''''''''''''
1205
1206 The ``sonorapass-bmc`` machine represents a lab server that never
1207 entered production. It can be replaced by the ``ast2500-evb`` machine
1208 using the ``fmc-model`` option to specify the flash type. The I2C
1209 devices connected to the board can be defined via the QEMU command
1210 line.
1211
1212 Arm ``qcom-dc-scm-v1-bmc`` and ``qcom-firework-bmc`` machine (removed in 11.1)
1213 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1214
1215 The ``qcom-dc-scm-v1-bmc`` and ``qcom-firework-bmc`` represent lab
1216 servers that never entered production. Since they do not rely on any
1217 specific device models, they can be replaced by the ``ast2600-evb``
1218 machine using the ``fmc-model`` option to specify the flash type. The
1219 I2C devices connected to the board can be defined via the QEMU command
1220 line.
1221
1222 Arm ``fp5280g2-bmc`` machine (removed in 11.1)
1223 ''''''''''''''''''''''''''''''''''''''''''''''
1224
1225 The ``fp5280g2-bmc`` machine does not rely on any specific device
1226 models, it can be replaced by the ``ast2500-evb`` machine using the
1227 ``fmc-model`` option to specify the flash type. The I2C devices
1228 connected to the board can be defined via the QEMU command line.
1229
1230 Arm ``fby35`` machine (removed in 11.1)
1231 '''''''''''''''''''''''''''''''''''''''
1232
1233 The ``fby35`` machine was originally added as an example of a
1234 multi-SoC system, with the expectation the models would evolve over
1235 time in an heterogeneous system. This hasn't happened and no public
1236 firmware is available to boot it. It can be replaced by the
1237 ``ast2700fc``, another multi-SoC machine based on the newer AST2700
1238 SoCs which are excepted to receive better support in the future.
1239
1240 RISC-V default machine (removed in 11.1)
1241 ''''''''''''''''''''''''''''''''''''''''
1242
1243 RISC-V used to define ``spike`` as the default machine if no machine option
1244 was given via the command line. This happened because ``spike`` was the first
1245 RISC-V machine implemented in QEMU and setting it as default was
1246 convenient at that time. Now we have 7 riscv64 and 6 riscv32 machines
1247 and having ``spike`` as a default is no longer justified.
1248
1249 The default machine option has been removed, forcing users to always set the
1250 machine they want to use to avoid confusion. Existing users of the ``spike``
1251 machine must ensure that they're setting the ``spike`` machine in the
1252 command line (``-M spike``).
1253
1254 Arm ``musicpal`` machine (removed in 11.2)
1255 ''''''''''''''''''''''''''''''''''''''''''
1256
1257 The ``musicpal`` machine was an emulation of the Freecom MusicPal, an
1258 "internet radio" device from the mid-2000s. This hardware was never
1259 supported in upstream Linux, was based on a Marvell SoC with no public
1260 documentation, and is now obsolete. We accidentally broke booting this
1261 board in 2024, without anybody noticing; it has therefore been removed
1262 without a deprecation period.
1263
1264 linux-user mode CPUs
1265 --------------------
1266
1267 ``tilegx`` CPUs (removed in 6.0)
1268 ''''''''''''''''''''''''''''''''
1269
1270 The ``tilegx`` guest CPU support has been removed without replacement. It was
1271 only implemented in linux-user mode, but support for this CPU was removed from
1272 the upstream Linux kernel in 2018, and it has also been dropped from glibc, so
1273 there is no new Linux development taking place with this architecture. For
1274 running the old binaries, you can use older versions of QEMU.
1275
1276 ``ppc64abi32`` CPUs (removed in 7.0)
1277 ''''''''''''''''''''''''''''''''''''
1278
1279 The ``ppc64abi32`` architecture has a number of issues which regularly
1280 tripped up the CI testing and was suspected to be quite broken. For that
1281 reason the maintainers strongly suspected no one actually used it.
1282
1283 ``nios2`` CPU (removed in 9.1)
1284 ''''''''''''''''''''''''''''''
1285
1286 QEMU Nios II architecture was orphan; Intel has EOL'ed the Nios II
1287 processor IP (see `Intel discontinuance notification`_).
1288
1289 iwMMXt emulation and the ``pxa`` CPUs (removed in 10.2)
1290 '''''''''''''''''''''''''''''''''''''''''''''''''''''''
1291
1292 The ``pxa`` CPU family (``pxa250``, ``pxa255``, ``pxa260``,
1293 ``pxa261``, ``pxa262``, ``pxa270-a0``, ``pxa270-a1``, ``pxa270``,
1294 ``pxa270-b0``, ``pxa270-b1``, ``pxa270-c0``, ``pxa270-c5``) were
1295 not available in system emulation, because all the machine types which
1296 used these CPUs were removed in the QEMU 9.2 release. We don't
1297 believe that anybody was using the iwMMXt emulation (which you
1298 would have to explicitly enable on the command line), and we did
1299 not have any tests to validate it or any real hardware or similar
1300 known-good implementation to test against. These CPUs have
1301 therefore been removed in linux-user mode as well.
1302
1303 TCG introspection features
1304 --------------------------
1305
1306 TCG trace-events (removed in 7.0)
1307 '''''''''''''''''''''''''''''''''
1308
1309 The ability to add new TCG trace points had bit rotted and as the
1310 feature can be replicated with TCG plugins it was removed. If
1311 any user is currently using this feature and needs help with
1312 converting to using TCG plugins they should contact the qemu-devel
1313 mailing list.
1314
1315
1316 Firmware, ACPI, Device Tree
1317 ---------------------------
1318
1319 RISC-V "virt" board "riscv,delegate" DT property (removed in 11.0)
1320 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1321
1322 The "riscv,delegate" DT property was added in QEMU 7.0 as part of
1323 the AIA APLIC support. The property changed name during the review
1324 process in Linux and the correct name ended up being "riscv,delegation".
1325
1326
1327 System emulator devices
1328 -----------------------
1329
1330 ``spapr-pci-vfio-host-bridge`` (removed in 2.12)
1331 '''''''''''''''''''''''''''''''''''''''''''''''''
1332
1333 The ``spapr-pci-vfio-host-bridge`` device type has been replaced by the
1334 ``spapr-pci-host-bridge`` device type.
1335
1336 ``ivshmem`` (removed in 4.0)
1337 ''''''''''''''''''''''''''''
1338
1339 Replaced by either the ``ivshmem-plain`` or ``ivshmem-doorbell``.
1340
1341 ``ide-drive`` (removed in 6.0)
1342 ''''''''''''''''''''''''''''''
1343
1344 The 'ide-drive' device has been removed. Users should use 'ide-hd' or
1345 'ide-cd' as appropriate to get an IDE hard disk or CD-ROM as needed.
1346
1347 ``scsi-disk`` (removed in 6.0)
1348 ''''''''''''''''''''''''''''''
1349
1350 The 'scsi-disk' device has been removed. Users should use 'scsi-hd' or
1351 'scsi-cd' as appropriate to get a SCSI hard disk or CD-ROM as needed.
1352
1353 ``sga`` (removed in 8.0)
1354 ''''''''''''''''''''''''
1355
1356 The ``sga`` device loaded an option ROM for x86 targets which enabled
1357 SeaBIOS to send messages to the serial console. SeaBIOS 1.11.0 onwards
1358 contains native support for this feature and thus use of the option
1359 ROM approach was obsolete. The native SeaBIOS support can be activated
1360 by using ``-machine graphics=off``.
1361
1362 ``pvrdma`` and the RDMA subsystem (removed in 9.1)
1363 ''''''''''''''''''''''''''''''''''''''''''''''''''
1364
1365 The 'pvrdma' device and the whole RDMA subsystem have been removed.
1366
1367 ``-device sd-card,spec_version=1`` (removed in 10.2)
1368 ''''''''''''''''''''''''''''''''''''''''''''''''''''
1369
1370 SD physical layer specification v2.00 supersedes the v1.10 one.
1371
1372
1373 System emulator binaries
1374 ------------------------
1375
1376 ``qemu-system-microblazeel`` (removed in 11.0)
1377 ''''''''''''''''''''''''''''''''''''''''''''''
1378
1379 The ``qemu-system-microblaze`` binary can emulate little-endian machines
1380 now, too, so the separate binary ``qemu-system-microblazeel`` (with the
1381 ``el`` suffix) for little-endian targets is not required anymore. The
1382 ``petalogix-s3adsp1800`` machine can now be switched to little endian by
1383 setting its ``endianness`` property to ``little``.
1384
1385
1386 Related binaries
1387 ----------------
1388
1389 ``qemu-nbd --partition`` (removed in 5.0)
1390 '''''''''''''''''''''''''''''''''''''''''
1391
1392 The ``qemu-nbd --partition $digit`` code (also spelled ``-P``)
1393 could only handle MBR partitions, and never correctly handled logical
1394 partitions beyond partition 5. Exporting a partition can still be
1395 done by utilizing the ``--image-opts`` option with a raw blockdev
1396 using the ``offset`` and ``size`` parameters layered on top of
1397 any other existing blockdev. For example, if partition 1 is 100MiB
1398 long starting at 1MiB, the old command::
1399
1400 qemu-nbd -t -P 1 -f qcow2 file.qcow2
1401
1402 can be rewritten as::
1403
1404 qemu-nbd -t --image-opts driver=raw,offset=1M,size=100M,file.driver=qcow2,file.file.driver=file,file.file.filename=file.qcow2
1405
1406 ``qemu-img convert -n -o`` (removed in 5.1)
1407 '''''''''''''''''''''''''''''''''''''''''''
1408
1409 All options specified in ``-o`` are image creation options, so
1410 they are now rejected when used with ``-n`` to skip image creation.
1411
1412
1413 ``qemu-img create -b bad file $size`` (removed in 5.1)
1414 ''''''''''''''''''''''''''''''''''''''''''''''''''''''
1415
1416 When creating an image with a backing file that could not be opened,
1417 ``qemu-img create`` used to issue a warning about the failure but
1418 proceed with the image creation if an explicit size was provided.
1419 However, as the ``-u`` option exists for this purpose, it is safer to
1420 enforce that any failure to open the backing image (including if the
1421 backing file is missing or an incorrect format was specified) is an
1422 error when ``-u`` is not used.
1423
1424 ``qemu-img amend`` to adjust backing file (removed in 6.1)
1425 ''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1426
1427 The use of ``qemu-img amend`` to modify the name or format of a qcow2
1428 backing image was never fully documented or tested, and interferes
1429 with other amend operations that need access to the original backing
1430 image (such as deciding whether a v3 zero cluster may be left
1431 unallocated when converting to a v2 image). Any changes to the
1432 backing chain should be performed with ``qemu-img rebase -u`` either
1433 before or after the remaining changes being performed by amend, as
1434 appropriate.
1435
1436 ``qemu-img`` backing file without format (removed in 6.1)
1437 '''''''''''''''''''''''''''''''''''''''''''''''''''''''''
1438
1439 The use of ``qemu-img create``, ``qemu-img rebase``, or ``qemu-img
1440 convert`` to create or modify an image that depends on a backing file
1441 now requires that an explicit backing format be provided. This is
1442 for safety: if QEMU probes a different format than what you thought,
1443 the data presented to the guest will be corrupt; similarly, presenting
1444 a raw image to a guest allows a potential security exploit if a future
1445 probe sees a non-raw image based on guest writes.
1446
1447 To avoid creating unsafe backing chains, you must pass ``-o
1448 backing_fmt=`` (or the shorthand ``-F`` during create) to specify the
1449 intended backing format. You may use ``qemu-img rebase -u`` to
1450 retroactively add a backing format to an existing image. However, be
1451 aware that there are already potential security risks to blindly using
1452 ``qemu-img info`` to probe the format of an untrusted backing image,
1453 when deciding what format to add into an existing image.
1454
1455 Block devices
1456 -------------
1457
1458 VXHS backend (removed in 5.1)
1459 '''''''''''''''''''''''''''''
1460
1461 The VXHS code did not compile since v2.12.0. It was removed in 5.1.
1462
1463 ``sheepdog`` driver (removed in 6.0)
1464 ''''''''''''''''''''''''''''''''''''
1465
1466 The corresponding upstream server project is no longer maintained.
1467 Users are recommended to switch to an alternative distributed block
1468 device driver such as RBD.
1469
1470 ``gluster`` backend (removed in 11.1)
1471 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1472
1473 According to https://marc.info/?l=fedora-devel-list&m=171934833215726
1474 the GlusterFS development effectively ended.
1475
1476
1477 VFIO devices
1478 ------------
1479
1480 ``-device vfio-calxeda-xgmac`` (removed in 10.2)
1481 ''''''''''''''''''''''''''''''''''''''''''''''''
1482 The vfio-calxeda-xgmac device allows to assign a host Calxeda Highbank
1483 10Gb XGMAC Ethernet controller device ("calxeda,hb-xgmac" compatibility
1484 string) to a guest. Calxeda HW has been ewasted now and there is no point
1485 keeping that device.
1486
1487 ``-device vfio-amd-xgbe`` (removed in 10.2)
1488 '''''''''''''''''''''''''''''''''''''''''''
1489 The vfio-amd-xgbe device allows to assign a host AMD 10GbE controller
1490 to a guest ("amd,xgbe-seattle-v1a" compatibility string). AMD "Seattle"
1491 is not supported anymore and there is no point keeping that device.
1492
1493 ``-device vfio-platform`` (removed in 10.2)
1494 '''''''''''''''''''''''''''''''''''''''''''
1495 The vfio-platform device allows to assign a host platform device
1496 to a guest in a generic manner. Integrating a new device into
1497 the vfio-platform infrastructure requires some adaptation at
1498 both kernel and qemu level. No such attempt has been done for years
1499 and the conclusion is that vfio-platform has not got any traction.
1500 PCIe passthrough shall be the mainline solution.
1501
1502 Tools
1503 -----
1504
1505 virtiofsd (removed in 8.0)
1506 ''''''''''''''''''''''''''
1507
1508 There is a newer Rust implementation of ``virtiofsd`` at
1509 ``https://gitlab.com/virtio-fs/virtiofsd``; this has been
1510 stable for some time and is now widely used.
1511 The command line and feature set is very close to the removed
1512 C implementation.
1513
1514 QEMU guest agent
1515 ----------------
1516
1517 ``--blacklist`` command line option (removed in 9.1)
1518 ''''''''''''''''''''''''''''''''''''''''''''''''''''
1519
1520 ``--blacklist`` has been replaced by ``--block-rpcs`` (which is a better
1521 wording for what this option does). The short form ``-b`` still stays
1522 the same and thus is the preferred way for scripts that should run with
1523 both, older and future versions of QEMU.
1524
1525 ``blacklist`` config file option (removed in 9.1)
1526 '''''''''''''''''''''''''''''''''''''''''''''''''
1527
1528 The ``blacklist`` config file option has been renamed to ``block-rpcs``
1529 (to be in sync with the renaming of the corresponding command line
1530 option).
1531
1532 Device options
1533 --------------
1534
1535 Character device options
1536 ''''''''''''''''''''''''
1537
1538 ``reconnect`` (removed in 10.2)
1539 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1540
1541 The ``reconnect`` has been replaced by ``reconnect-ms``, which provides
1542 better precision.
1543
1544 Net device options
1545 ''''''''''''''''''
1546
1547 Stream ``reconnect`` (removed in 10.2)
1548 ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
1549
1550 The ``reconnect`` has been replaced by ``reconnect-ms``, which provides
1551 better precision.
1552
1553
1554 .. _Intel discontinuance notification: https://www.intel.com/content/www/us/en/content-details/781327/intel-is-discontinuing-ip-ordering-codes-listed-in-pdn2312-for-nios-ii-ip.html