xref: /openbmc/qemu/docs/about/deprecated.rst (revision 7200fb21)
1.. _Deprecated features:
2
3Deprecated features
4===================
5
6In general features are intended to be supported indefinitely once
7introduced into QEMU. In the event that a feature needs to be removed,
8it will be listed in this section. The feature will remain functional for the
9release in which it was deprecated and one further release. After these two
10releases, the feature is liable to be removed. Deprecated features may also
11generate warnings on the console when QEMU starts up, or if activated via a
12monitor command, however, this is not a mandatory requirement.
13
14Prior to the 2.10.0 release there was no official policy on how
15long features would be deprecated prior to their removal, nor
16any documented list of which features were deprecated. Thus
17any features deprecated prior to 2.10.0 will be treated as if
18they were first deprecated in the 2.10.0 release.
19
20What follows is a list of all features currently marked as
21deprecated.
22
23System emulator command line arguments
24--------------------------------------
25
26Short-form boolean options (since 6.0)
27''''''''''''''''''''''''''''''''''''''
28
29Boolean options such as ``share=on``/``share=off`` could be written
30in short form as ``share`` and ``noshare``.  This is now deprecated
31and will cause a warning.
32
33``delay`` option for socket character devices (since 6.0)
34'''''''''''''''''''''''''''''''''''''''''''''''''''''''''
35
36The replacement for the ``nodelay`` short-form boolean option is ``nodelay=on``
37rather than ``delay=off``.
38
39``-smp`` ("parameter=0" SMP configurations) (since 6.2)
40'''''''''''''''''''''''''''''''''''''''''''''''''''''''
41
42Specified CPU topology parameters must be greater than zero.
43
44In the SMP configuration, users should either provide a CPU topology
45parameter with a reasonable value (greater than zero) or just omit it
46and QEMU will compute the missing value.
47
48However, historically it was implicitly allowed for users to provide
49a parameter with zero value, which is meaningless and could also possibly
50cause unexpected results in the -smp parsing. So support for this kind of
51configurations (e.g. -smp 8,sockets=0) is deprecated since 6.2 and will
52be removed in the near future, users have to ensure that all the topology
53members described with -smp are greater than zero.
54
55Plugin argument passing through ``arg=<string>`` (since 6.1)
56''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
57
58Passing TCG plugins arguments through ``arg=`` is redundant is makes the
59command-line less readable, especially when the argument itself consist of a
60name and a value, e.g. ``-plugin plugin_name,arg="arg_name=arg_value"``.
61Therefore, the usage of ``arg`` is redundant. Single-word arguments are treated
62as short-form boolean values, and passed to plugins as ``arg_name=on``.
63However, short-form booleans are deprecated and full explicit ``arg_name=on``
64form is preferred.
65
66``-no-hpet`` (since 8.0)
67''''''''''''''''''''''''
68
69The HPET setting has been turned into a machine property.
70Use ``-machine hpet=off`` instead.
71
72``-no-acpi`` (since 8.0)
73''''''''''''''''''''''''
74
75The ``-no-acpi`` setting has been turned into a machine property.
76Use ``-machine acpi=off`` instead.
77
78``-async-teardown`` (since 8.1)
79'''''''''''''''''''''''''''''''
80
81Use ``-run-with async-teardown=on`` instead.
82
83``-chroot`` (since 8.1)
84'''''''''''''''''''''''
85
86Use ``-run-with chroot=dir`` instead.
87
88``-singlestep`` (since 8.1)
89'''''''''''''''''''''''''''
90
91The ``-singlestep`` option has been turned into an accelerator property,
92and given a name that better reflects what it actually does.
93Use ``-accel tcg,one-insn-per-tb=on`` instead.
94
95User-mode emulator command line arguments
96-----------------------------------------
97
98``-singlestep`` (since 8.1)
99'''''''''''''''''''''''''''
100
101The ``-singlestep`` option has been given a name that better reflects
102what it actually does. For both linux-user and bsd-user, use the
103new ``-one-insn-per-tb`` option instead.
104
105QEMU Machine Protocol (QMP) commands
106------------------------------------
107
108``blockdev-open-tray``, ``blockdev-close-tray`` argument ``device`` (since 2.8)
109'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
110
111Use argument ``id`` instead.
112
113``eject`` argument ``device`` (since 2.8)
114'''''''''''''''''''''''''''''''''''''''''
115
116Use argument ``id`` instead.
117
118``blockdev-change-medium`` argument ``device`` (since 2.8)
119''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
120
121Use argument ``id`` instead.
122
123``block_set_io_throttle`` argument ``device`` (since 2.8)
124'''''''''''''''''''''''''''''''''''''''''''''''''''''''''
125
126Use argument ``id`` instead.
127
128``blockdev-add`` empty string argument ``backing`` (since 2.10)
129'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
130
131Use argument value ``null`` instead.
132
133``block-commit`` arguments ``base`` and ``top`` (since 3.1)
134'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
135
136Use arguments ``base-node`` and ``top-node`` instead.
137
138``nbd-server-add`` and ``nbd-server-remove`` (since 5.2)
139''''''''''''''''''''''''''''''''''''''''''''''''''''''''
140
141Use the more generic commands ``block-export-add`` and ``block-export-del``
142instead.  As part of this deprecation, where ``nbd-server-add`` used a
143single ``bitmap``, the new ``block-export-add`` uses a list of ``bitmaps``.
144
145``query-qmp-schema`` return value member ``values`` (since 6.2)
146'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
147
148Member ``values`` in return value elements with meta-type ``enum`` is
149deprecated.  Use ``members`` instead.
150
151``drive-backup`` (since 6.2)
152''''''''''''''''''''''''''''
153
154Use ``blockdev-backup`` in combination with ``blockdev-add`` instead.
155This change primarily separates the creation/opening process of the backup
156target with explicit, separate steps. ``blockdev-backup`` uses mostly the
157same arguments as ``drive-backup``, except the ``format`` and ``mode``
158options are removed in favor of using explicit ``blockdev-create`` and
159``blockdev-add`` calls. See :doc:`/interop/live-block-operations` for
160details.
161
162Incorrectly typed ``device_add`` arguments (since 6.2)
163''''''''''''''''''''''''''''''''''''''''''''''''''''''
164
165Due to shortcomings in the internal implementation of ``device_add``, QEMU
166incorrectly accepts certain invalid arguments: Any object or list arguments are
167silently ignored. Other argument types are not checked, but an implicit
168conversion happens, so that e.g. string values can be assigned to integer
169device properties or vice versa.
170
171This is a bug in QEMU that will be fixed in the future so that previously
172accepted incorrect commands will return an error. Users should make sure that
173all arguments passed to ``device_add`` are consistent with the documented
174property types.
175
176``StatusInfo`` member ``singlestep`` (since 8.1)
177''''''''''''''''''''''''''''''''''''''''''''''''
178
179The ``singlestep`` member of the ``StatusInfo`` returned from the
180``query-status`` command is deprecated. This member has a confusing
181name and it never did what the documentation claimed or what its name
182suggests. We do not believe that anybody is actually using the
183information provided in this member.
184
185The information it reports is whether the TCG JIT is in "one
186instruction per translated block" mode (which can be set on the
187command line or via the HMP, but not via QMP). The information remains
188available via the HMP 'info jit' command.
189
190QEMU Machine Protocol (QMP) events
191----------------------------------
192
193``MEM_UNPLUG_ERROR`` (since 6.2)
194''''''''''''''''''''''''''''''''''''''''''''''''''''''''
195
196Use the more generic event ``DEVICE_UNPLUG_GUEST_ERROR`` instead.
197
198``vcpu`` trace events (since 8.1)
199'''''''''''''''''''''''''''''''''
200
201The ability to instrument QEMU helper functions with vCPU-aware trace
202points was removed in 7.0. However QMP still exposed the vcpu
203parameter. This argument has now been deprecated and the remaining
204remaining trace points that used it are selected just by name.
205
206Human Monitor Protocol (HMP) commands
207-------------------------------------
208
209``singlestep`` (since 8.1)
210''''''''''''''''''''''''''
211
212The ``singlestep`` command has been replaced by the ``one-insn-per-tb``
213command, which has the same behaviour but a less misleading name.
214
215Host Architectures
216------------------
217
218BE MIPS (since 7.2)
219'''''''''''''''''''
220
221As Debian 10 ("Buster") moved into LTS the big endian 32 bit version of
222MIPS moved out of support making it hard to maintain our
223cross-compilation CI tests of the architecture. As we no longer have
224CI coverage support may bitrot away before the deprecation process
225completes. The little endian variants of MIPS (both 32 and 64 bit) are
226still a supported host architecture.
227
228System emulation on 32-bit x86 hosts (since 8.0)
229''''''''''''''''''''''''''''''''''''''''''''''''
230
231Support for 32-bit x86 host deployments is increasingly uncommon in mainstream
232OS distributions given the widespread availability of 64-bit x86 hardware.
233The QEMU project no longer considers 32-bit x86 support for system emulation to
234be an effective use of its limited resources, and thus intends to discontinue
235it. Since all recent x86 hardware from the past >10 years is capable of the
23664-bit x86 extensions, a corresponding 64-bit OS should be used instead.
237
238
239System emulator machines
240------------------------
241
242Arm ``virt`` machine ``dtb-kaslr-seed`` property (since 7.1)
243''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
244
245The ``dtb-kaslr-seed`` property on the ``virt`` board has been
246deprecated; use the new name ``dtb-randomness`` instead. The new name
247better reflects the way this property affects all random data within
248the device tree blob, not just the ``kaslr-seed`` node.
249
250``pc-i440fx-2.0`` up to ``pc-i440fx-2.3`` (since 8.2)
251'''''''''''''''''''''''''''''''''''''''''''''''''''''
252
253These old machine types are quite neglected nowadays and thus might have
254various pitfalls with regards to live migration. Use a newer machine type
255instead.
256
257
258Backend options
259---------------
260
261Using non-persistent backing file with pmem=on (since 6.1)
262''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
263
264This option is used when ``memory-backend-file`` is consumed by emulated NVDIMM
265device. However enabling ``memory-backend-file.pmem`` option, when backing file
266is (a) not DAX capable or (b) not on a filesystem that support direct mapping
267of persistent memory, is not safe and may lead to data loss or corruption in case
268of host crash.
269Options are:
270
271    - modify VM configuration to set ``pmem=off`` to continue using fake NVDIMM
272      (without persistence guaranties) with backing file on non DAX storage
273    - move backing file to NVDIMM storage and keep ``pmem=on``
274      (to have NVDIMM with persistence guaranties).
275
276Device options
277--------------
278
279Emulated device options
280'''''''''''''''''''''''
281
282``-device virtio-blk,scsi=on|off`` (since 5.0)
283^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
284
285The virtio-blk SCSI passthrough feature is a legacy VIRTIO feature.  VIRTIO 1.0
286and later do not support it because the virtio-scsi device was introduced for
287full SCSI support.  Use virtio-scsi instead when SCSI passthrough is required.
288
289Note this also applies to ``-device virtio-blk-pci,scsi=on|off``, which is an
290alias.
291
292``-device nvme-ns,eui64-default=on|off`` (since 7.1)
293^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
294
295In QEMU versions 6.1, 6.2 and 7.0, the ``nvme-ns`` generates an EUI-64
296identifier that is not globally unique. If an EUI-64 identifier is required, the
297user must set it explicitly using the ``nvme-ns`` device parameter ``eui64``.
298
299``-device nvme,use-intel-id=on|off`` (since 7.1)
300^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
301
302The ``nvme`` device originally used a PCI Vendor/Device Identifier combination
303from Intel that was not properly allocated. Since version 5.2, the controller
304has used a properly allocated identifier. Deprecate the ``use-intel-id``
305machine compatibility parameter.
306
307``-device cxl-type3,memdev=xxxx`` (since 8.0)
308^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
309
310The ``cxl-type3`` device initially only used a single memory backend.  With
311the addition of volatile memory support, it is now necessary to distinguish
312between persistent and volatile memory backends.  As such, memdev is deprecated
313in favor of persistent-memdev.
314
315``-fsdev proxy`` and ``-virtfs proxy`` (since 8.1)
316^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
317
318The 9p ``proxy`` filesystem backend driver has been deprecated and will be
319removed (along with its proxy helper daemon) in a future version of QEMU. Please
320use ``-fsdev local`` or ``-virtfs local`` for using the 9p ``local`` filesystem
321backend, or alternatively consider deploying virtiofsd instead.
322
323The 9p ``proxy`` backend was originally developed as an alternative to the 9p
324``local`` backend. The idea was to enhance security by dispatching actual low
325level filesystem operations from 9p server (QEMU process) over to a separate
326process (the virtfs-proxy-helper binary). However this alternative never gained
327momentum. The proxy backend is much slower than the local backend, hasn't seen
328any development in years, and showed to be less secure, especially due to the
329fact that its helper daemon must be run as root, whereas with the local backend
330QEMU is typically run as unprivileged user and allows to tighten behaviour by
331mapping permissions et al by using its 'mapped' security model option.
332
333Nowadays it would make sense to reimplement the ``proxy`` backend by using
334QEMU's ``vhost`` feature, which would eliminate the high latency costs under
335which the 9p ``proxy`` backend currently suffers. However as of to date nobody
336has indicated plans for such kind of reimplementation unfortunately.
337
338RISC-V 'any' CPU type ``-cpu any`` (since 8.2)
339^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
340
341The 'any' CPU type was introduced back in 2018 and has been around since the
342initial RISC-V QEMU port. Its usage has always been unclear: users don't know
343what to expect from a CPU called 'any', and in fact the CPU does not do anything
344special that isn't already done by the default CPUs rv32/rv64.
345
346After the introduction of the 'max' CPU type, RISC-V now has a good coverage
347of generic CPUs: rv32 and rv64 as default CPUs and 'max' as a feature complete
348CPU for both 32 and 64 bit builds. Users are then discouraged to use the 'any'
349CPU type starting in 8.2.
350
351RISC-V CPU properties which start with capital 'Z' (since 8.2)
352^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
353
354All RISC-V CPU properties which start with capital 'Z' are being deprecated
355starting in 8.2. The reason is that they were wrongly added with capital 'Z'
356in the past. CPU properties were later added with lower-case names, which
357is the format we want to use from now on.
358
359Users which try to use these deprecated properties will receive a warning
360recommending to switch to their stable counterparts:
361
362- "Zifencei" should be replaced with "zifencei"
363- "Zicsr" should be replaced with "zicsr"
364- "Zihintntl" should be replaced with "zihintntl"
365- "Zihintpause" should be replaced with "zihintpause"
366- "Zawrs" should be replaced with "zawrs"
367- "Zfa" should be replaced with "zfa"
368- "Zfh" should be replaced with "zfh"
369- "Zfhmin" should be replaced with "zfhmin"
370- "Zve32f" should be replaced with "zve32f"
371- "Zve64f" should be replaced with "zve64f"
372- "Zve64d" should be replaced with "zve64d"
373
374``-device pvrdma`` and the rdma subsystem (since 8.2)
375^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
376
377The pvrdma device and the whole rdma subsystem are in a bad shape and
378without active maintenance. The QEMU project intends to remove this
379device and subsystem from the code base in a future release without
380replacement unless somebody steps up and improves the situation.
381
382
383Block device options
384''''''''''''''''''''
385
386``"backing": ""`` (since 2.12)
387^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
388
389In order to prevent QEMU from automatically opening an image's backing
390chain, use ``"backing": null`` instead.
391
392``rbd`` keyvalue pair encoded filenames: ``""`` (since 3.1)
393^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
394
395Options for ``rbd`` should be specified according to its runtime options,
396like other block drivers.  Legacy parsing of keyvalue pair encoded
397filenames is useful to open images with the old format for backing files;
398These image files should be updated to use the current format.
399
400Example of legacy encoding::
401
402  json:{"file.driver":"rbd", "file.filename":"rbd:rbd/name"}
403
404The above, converted to the current supported format::
405
406  json:{"file.driver":"rbd", "file.pool":"rbd", "file.image":"name"}
407
408``iscsi,password=xxx`` (since 8.0)
409^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
410
411Specifying the iSCSI password in plain text on the command line using the
412``password`` option is insecure. The ``password-secret`` option should be
413used instead, to refer to a ``--object secret...`` instance that provides
414a password via a file, or encrypted.
415
416CPU device properties
417'''''''''''''''''''''
418
419``pmu-num=n`` on RISC-V CPUs (since 8.2)
420^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
421
422In order to support more flexible counter configurations this has been replaced
423by a ``pmu-mask`` property. If set of counters is continuous then the mask can
424be calculated with ``((2 ^ n) - 1) << 3``. The least significant three bits
425must be left clear.
426
427
428Backwards compatibility
429-----------------------
430
431Runnability guarantee of CPU models (since 4.1)
432'''''''''''''''''''''''''''''''''''''''''''''''
433
434Previous versions of QEMU never changed existing CPU models in
435ways that introduced additional host software or hardware
436requirements to the VM.  This allowed management software to
437safely change the machine type of an existing VM without
438introducing new requirements ("runnability guarantee").  This
439prevented CPU models from being updated to include CPU
440vulnerability mitigations, leaving guests vulnerable in the
441default configuration.
442
443The CPU model runnability guarantee won't apply anymore to
444existing CPU models.  Management software that needs runnability
445guarantees must resolve the CPU model aliases using the
446``alias-of`` field returned by the ``query-cpu-definitions`` QMP
447command.
448
449While those guarantees are kept, the return value of
450``query-cpu-definitions`` will have existing CPU model aliases
451point to a version that doesn't break runnability guarantees
452(specifically, version 1 of those CPU models).  In future QEMU
453versions, aliases will point to newer CPU model versions
454depending on the machine type, so management software must
455resolve CPU model aliases before starting a virtual machine.
456
457QEMU guest agent
458----------------
459
460``--blacklist`` command line option (since 7.2)
461'''''''''''''''''''''''''''''''''''''''''''''''
462
463``--blacklist`` has been replaced by ``--block-rpcs`` (which is a better
464wording for what this option does). The short form ``-b`` still stays
465the same and thus is the preferred way for scripts that should run with
466both, older and future versions of QEMU.
467
468``blacklist`` config file option (since 7.2)
469''''''''''''''''''''''''''''''''''''''''''''
470
471The ``blacklist`` config file option has been renamed to ``block-rpcs``
472(to be in sync with the renaming of the corresponding command line
473option).
474
475Migration
476---------
477
478``skipped`` MigrationStats field (since 8.1)
479''''''''''''''''''''''''''''''''''''''''''''
480
481``skipped`` field in Migration stats has been deprecated.  It hasn't
482been used for more than 10 years.
483
484``inc`` migrate command option (since 8.2)
485''''''''''''''''''''''''''''''''''''''''''
486
487Use blockdev-mirror with NBD instead.
488
489As an intermediate step the ``inc`` functionality can be achieved by
490setting the ``block-incremental`` migration parameter to ``true``.
491But this parameter is also deprecated.
492
493``blk`` migrate command option (since 8.2)
494''''''''''''''''''''''''''''''''''''''''''
495
496Use blockdev-mirror with NBD instead.
497
498As an intermediate step the ``blk`` functionality can be achieved by
499setting the ``block`` migration capability to ``true``.  But this
500capability is also deprecated.
501
502block migration (since 8.2)
503'''''''''''''''''''''''''''
504
505Block migration is too inflexible.  It needs to migrate all block
506devices or none.
507
508Please see "QMP invocation for live storage migration with
509``blockdev-mirror`` + NBD" in docs/interop/live-block-operations.rst
510for a detailed explanation.
511
512old compression method (since 8.2)
513''''''''''''''''''''''''''''''''''
514
515Compression method fails too much.  Too many races.  We are going to
516remove it if nobody fixes it.  For starters, migration-test
517compression tests are disabled becase they fail randomly.  If you need
518compression, use multifd compression methods.
519