xref: /openbmc/qemu/docs/about/deprecated.rst (revision 4b3520fd)
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
23Build options
24-------------
25
26``gprof`` builds (since 8.0)
27''''''''''''''''''''''''''''
28
29The ``--enable-gprof`` configure setting relies on compiler
30instrumentation to gather its data which can distort the generated
31profile. As other non-instrumenting tools are available that give a
32more holistic view of the system with non-instrumented binaries we are
33deprecating the build option and no longer defend it in CI. The
34``--enable-gcov`` build option remains for analysis test case
35coverage.
36
37System emulator command line arguments
38--------------------------------------
39
40``QEMU_AUDIO_`` environment variables and ``-audio-help`` (since 4.0)
41'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
42
43The ``-audiodev`` argument is now the preferred way to specify audio
44backend settings instead of environment variables.  To ease migration to
45the new format, the ``-audiodev-help`` option can be used to convert
46the current values of the environment variables to ``-audiodev`` options.
47
48Creating sound card devices and vnc without ``audiodev=`` property (since 4.2)
49''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
50
51When not using the deprecated legacy audio config, each sound card
52should specify an ``audiodev=`` property.  Additionally, when using
53vnc, you should specify an ``audiodev=`` property if you plan to
54transmit audio through the VNC protocol.
55
56Short-form boolean options (since 6.0)
57''''''''''''''''''''''''''''''''''''''
58
59Boolean options such as ``share=on``/``share=off`` could be written
60in short form as ``share`` and ``noshare``.  This is now deprecated
61and will cause a warning.
62
63``delay`` option for socket character devices (since 6.0)
64'''''''''''''''''''''''''''''''''''''''''''''''''''''''''
65
66The replacement for the ``nodelay`` short-form boolean option is ``nodelay=on``
67rather than ``delay=off``.
68
69``-smp`` ("parameter=0" SMP configurations) (since 6.2)
70'''''''''''''''''''''''''''''''''''''''''''''''''''''''
71
72Specified CPU topology parameters must be greater than zero.
73
74In the SMP configuration, users should either provide a CPU topology
75parameter with a reasonable value (greater than zero) or just omit it
76and QEMU will compute the missing value.
77
78However, historically it was implicitly allowed for users to provide
79a parameter with zero value, which is meaningless and could also possibly
80cause unexpected results in the -smp parsing. So support for this kind of
81configurations (e.g. -smp 8,sockets=0) is deprecated since 6.2 and will
82be removed in the near future, users have to ensure that all the topology
83members described with -smp are greater than zero.
84
85Plugin argument passing through ``arg=<string>`` (since 6.1)
86''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
87
88Passing TCG plugins arguments through ``arg=`` is redundant is makes the
89command-line less readable, especially when the argument itself consist of a
90name and a value, e.g. ``-plugin plugin_name,arg="arg_name=arg_value"``.
91Therefore, the usage of ``arg`` is redundant. Single-word arguments are treated
92as short-form boolean values, and passed to plugins as ``arg_name=on``.
93However, short-form booleans are deprecated and full explicit ``arg_name=on``
94form is preferred.
95
96``-no-hpet`` (since 8.0)
97''''''''''''''''''''''''
98
99The HPET setting has been turned into a machine property.
100Use ``-machine hpet=off`` instead.
101
102``-no-acpi`` (since 8.0)
103''''''''''''''''''''''''
104
105The ``-no-acpi`` setting has been turned into a machine property.
106Use ``-machine acpi=off`` instead.
107
108``-accel hax`` (since 8.0)
109''''''''''''''''''''''''''
110
111The HAXM project has been retired (see https://github.com/intel/haxm#status).
112Use "whpx" (on Windows) or "hvf" (on macOS) instead.
113
114``-async-teardown`` (since 8.1)
115'''''''''''''''''''''''''''''''
116
117Use ``-run-with async-teardown=on`` instead.
118
119``-chroot`` (since 8.1)
120'''''''''''''''''''''''
121
122Use ``-run-with chroot=dir`` instead.
123
124``-singlestep`` (since 8.1)
125'''''''''''''''''''''''''''
126
127The ``-singlestep`` option has been turned into an accelerator property,
128and given a name that better reflects what it actually does.
129Use ``-accel tcg,one-insn-per-tb=on`` instead.
130
131User-mode emulator command line arguments
132-----------------------------------------
133
134``-singlestep`` (since 8.1)
135'''''''''''''''''''''''''''
136
137The ``-singlestep`` option has been given a name that better reflects
138what it actually does. For both linux-user and bsd-user, use the
139new ``-one-insn-per-tb`` option instead.
140
141QEMU Machine Protocol (QMP) commands
142------------------------------------
143
144``blockdev-open-tray``, ``blockdev-close-tray`` argument ``device`` (since 2.8)
145'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
146
147Use argument ``id`` instead.
148
149``eject`` argument ``device`` (since 2.8)
150'''''''''''''''''''''''''''''''''''''''''
151
152Use argument ``id`` instead.
153
154``blockdev-change-medium`` argument ``device`` (since 2.8)
155''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
156
157Use argument ``id`` instead.
158
159``block_set_io_throttle`` argument ``device`` (since 2.8)
160'''''''''''''''''''''''''''''''''''''''''''''''''''''''''
161
162Use argument ``id`` instead.
163
164``blockdev-add`` empty string argument ``backing`` (since 2.10)
165'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
166
167Use argument value ``null`` instead.
168
169``block-commit`` arguments ``base`` and ``top`` (since 3.1)
170'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
171
172Use arguments ``base-node`` and ``top-node`` instead.
173
174``nbd-server-add`` and ``nbd-server-remove`` (since 5.2)
175''''''''''''''''''''''''''''''''''''''''''''''''''''''''
176
177Use the more generic commands ``block-export-add`` and ``block-export-del``
178instead.  As part of this deprecation, where ``nbd-server-add`` used a
179single ``bitmap``, the new ``block-export-add`` uses a list of ``bitmaps``.
180
181``query-qmp-schema`` return value member ``values`` (since 6.2)
182'''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
183
184Member ``values`` in return value elements with meta-type ``enum`` is
185deprecated.  Use ``members`` instead.
186
187``drive-backup`` (since 6.2)
188''''''''''''''''''''''''''''
189
190Use ``blockdev-backup`` in combination with ``blockdev-add`` instead.
191This change primarily separates the creation/opening process of the backup
192target with explicit, separate steps. ``blockdev-backup`` uses mostly the
193same arguments as ``drive-backup``, except the ``format`` and ``mode``
194options are removed in favor of using explicit ``blockdev-create`` and
195``blockdev-add`` calls. See :doc:`/interop/live-block-operations` for
196details.
197
198Incorrectly typed ``device_add`` arguments (since 6.2)
199''''''''''''''''''''''''''''''''''''''''''''''''''''''
200
201Due to shortcomings in the internal implementation of ``device_add``, QEMU
202incorrectly accepts certain invalid arguments: Any object or list arguments are
203silently ignored. Other argument types are not checked, but an implicit
204conversion happens, so that e.g. string values can be assigned to integer
205device properties or vice versa.
206
207This is a bug in QEMU that will be fixed in the future so that previously
208accepted incorrect commands will return an error. Users should make sure that
209all arguments passed to ``device_add`` are consistent with the documented
210property types.
211
212``StatusInfo`` member ``singlestep`` (since 8.1)
213''''''''''''''''''''''''''''''''''''''''''''''''
214
215The ``singlestep`` member of the ``StatusInfo`` returned from the
216``query-status`` command is deprecated. This member has a confusing
217name and it never did what the documentation claimed or what its name
218suggests. We do not believe that anybody is actually using the
219information provided in this member.
220
221The information it reports is whether the TCG JIT is in "one
222instruction per translated block" mode (which can be set on the
223command line or via the HMP, but not via QMP). The information remains
224available via the HMP 'info jit' command.
225
226QEMU Machine Protocol (QMP) events
227----------------------------------
228
229``MEM_UNPLUG_ERROR`` (since 6.2)
230''''''''''''''''''''''''''''''''''''''''''''''''''''''''
231
232Use the more generic event ``DEVICE_UNPLUG_GUEST_ERROR`` instead.
233
234``vcpu`` trace events (since 8.1)
235'''''''''''''''''''''''''''''''''
236
237The ability to instrument QEMU helper functions with vCPU-aware trace
238points was removed in 7.0. However QMP still exposed the vcpu
239parameter. This argument has now been deprecated and the remaining
240remaining trace points that used it are selected just by name.
241
242Human Monitor Protocol (HMP) commands
243-------------------------------------
244
245``singlestep`` (since 8.1)
246''''''''''''''''''''''''''
247
248The ``singlestep`` command has been replaced by the ``one-insn-per-tb``
249command, which has the same behaviour but a less misleading name.
250
251Host Architectures
252------------------
253
254BE MIPS (since 7.2)
255'''''''''''''''''''
256
257As Debian 10 ("Buster") moved into LTS the big endian 32 bit version of
258MIPS moved out of support making it hard to maintain our
259cross-compilation CI tests of the architecture. As we no longer have
260CI coverage support may bitrot away before the deprecation process
261completes. The little endian variants of MIPS (both 32 and 64 bit) are
262still a supported host architecture.
263
264System emulation on 32-bit x86 hosts (since 8.0)
265''''''''''''''''''''''''''''''''''''''''''''''''
266
267Support for 32-bit x86 host deployments is increasingly uncommon in mainstream
268OS distributions given the widespread availability of 64-bit x86 hardware.
269The QEMU project no longer considers 32-bit x86 support for system emulation to
270be an effective use of its limited resources, and thus intends to discontinue
271it. Since all recent x86 hardware from the past >10 years is capable of the
27264-bit x86 extensions, a corresponding 64-bit OS should be used instead.
273
274
275System emulator machines
276------------------------
277
278Arm ``virt`` machine ``dtb-kaslr-seed`` property (since 7.1)
279''''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
280
281The ``dtb-kaslr-seed`` property on the ``virt`` board has been
282deprecated; use the new name ``dtb-randomness`` instead. The new name
283better reflects the way this property affects all random data within
284the device tree blob, not just the ``kaslr-seed`` node.
285
286``pc-i440fx-1.4`` up to ``pc-i440fx-1.7`` (since 7.0)
287'''''''''''''''''''''''''''''''''''''''''''''''''''''
288
289These old machine types are quite neglected nowadays and thus might have
290various pitfalls with regards to live migration. Use a newer machine type
291instead.
292
293
294Backend options
295---------------
296
297Using non-persistent backing file with pmem=on (since 6.1)
298''''''''''''''''''''''''''''''''''''''''''''''''''''''''''
299
300This option is used when ``memory-backend-file`` is consumed by emulated NVDIMM
301device. However enabling ``memory-backend-file.pmem`` option, when backing file
302is (a) not DAX capable or (b) not on a filesystem that support direct mapping
303of persistent memory, is not safe and may lead to data loss or corruption in case
304of host crash.
305Options are:
306
307    - modify VM configuration to set ``pmem=off`` to continue using fake NVDIMM
308      (without persistence guaranties) with backing file on non DAX storage
309    - move backing file to NVDIMM storage and keep ``pmem=on``
310      (to have NVDIMM with persistence guaranties).
311
312Device options
313--------------
314
315Emulated device options
316'''''''''''''''''''''''
317
318``-device virtio-blk,scsi=on|off`` (since 5.0)
319^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
320
321The virtio-blk SCSI passthrough feature is a legacy VIRTIO feature.  VIRTIO 1.0
322and later do not support it because the virtio-scsi device was introduced for
323full SCSI support.  Use virtio-scsi instead when SCSI passthrough is required.
324
325Note this also applies to ``-device virtio-blk-pci,scsi=on|off``, which is an
326alias.
327
328``-device nvme-ns,eui64-default=on|off`` (since 7.1)
329^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
330
331In QEMU versions 6.1, 6.2 and 7.0, the ``nvme-ns`` generates an EUI-64
332identifier that is not globally unique. If an EUI-64 identifier is required, the
333user must set it explicitly using the ``nvme-ns`` device parameter ``eui64``.
334
335``-device nvme,use-intel-id=on|off`` (since 7.1)
336^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
337
338The ``nvme`` device originally used a PCI Vendor/Device Identifier combination
339from Intel that was not properly allocated. Since version 5.2, the controller
340has used a properly allocated identifier. Deprecate the ``use-intel-id``
341machine compatibility parameter.
342
343``-device cxl-type3,memdev=xxxx`` (since 8.0)
344^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
345
346The ``cxl-type3`` device initially only used a single memory backend.  With
347the addition of volatile memory support, it is now necessary to distinguish
348between persistent and volatile memory backends.  As such, memdev is deprecated
349in favor of persistent-memdev.
350
351``-fsdev proxy`` and ``-virtfs proxy`` (since 8.1)
352^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
353
354The 9p ``proxy`` filesystem backend driver has been deprecated and will be
355removed (along with its proxy helper daemon) in a future version of QEMU. Please
356use ``-fsdev local`` or ``-virtfs local`` for using the 9p ``local`` filesystem
357backend, or alternatively consider deploying virtiofsd instead.
358
359The 9p ``proxy`` backend was originally developed as an alternative to the 9p
360``local`` backend. The idea was to enhance security by dispatching actual low
361level filesystem operations from 9p server (QEMU process) over to a separate
362process (the virtfs-proxy-helper binary). However this alternative never gained
363momentum. The proxy backend is much slower than the local backend, hasn't seen
364any development in years, and showed to be less secure, especially due to the
365fact that its helper daemon must be run as root, whereas with the local backend
366QEMU is typically run as unprivileged user and allows to tighten behaviour by
367mapping permissions et al by using its 'mapped' security model option.
368
369Nowadays it would make sense to reimplement the ``proxy`` backend by using
370QEMU's ``vhost`` feature, which would eliminate the high latency costs under
371which the 9p ``proxy`` backend currently suffers. However as of to date nobody
372has indicated plans for such kind of reimplementation unfortunately.
373
374
375Block device options
376''''''''''''''''''''
377
378``"backing": ""`` (since 2.12)
379^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
380
381In order to prevent QEMU from automatically opening an image's backing
382chain, use ``"backing": null`` instead.
383
384``rbd`` keyvalue pair encoded filenames: ``""`` (since 3.1)
385^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
386
387Options for ``rbd`` should be specified according to its runtime options,
388like other block drivers.  Legacy parsing of keyvalue pair encoded
389filenames is useful to open images with the old format for backing files;
390These image files should be updated to use the current format.
391
392Example of legacy encoding::
393
394  json:{"file.driver":"rbd", "file.filename":"rbd:rbd/name"}
395
396The above, converted to the current supported format::
397
398  json:{"file.driver":"rbd", "file.pool":"rbd", "file.image":"name"}
399
400``iscsi,password=xxx`` (since 8.0)
401^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
402
403Specifying the iSCSI password in plain text on the command line using the
404``password`` option is insecure. The ``password-secret`` option should be
405used instead, to refer to a ``--object secret...`` instance that provides
406a password via a file, or encrypted.
407
408Backwards compatibility
409-----------------------
410
411Runnability guarantee of CPU models (since 4.1)
412'''''''''''''''''''''''''''''''''''''''''''''''
413
414Previous versions of QEMU never changed existing CPU models in
415ways that introduced additional host software or hardware
416requirements to the VM.  This allowed management software to
417safely change the machine type of an existing VM without
418introducing new requirements ("runnability guarantee").  This
419prevented CPU models from being updated to include CPU
420vulnerability mitigations, leaving guests vulnerable in the
421default configuration.
422
423The CPU model runnability guarantee won't apply anymore to
424existing CPU models.  Management software that needs runnability
425guarantees must resolve the CPU model aliases using the
426``alias-of`` field returned by the ``query-cpu-definitions`` QMP
427command.
428
429While those guarantees are kept, the return value of
430``query-cpu-definitions`` will have existing CPU model aliases
431point to a version that doesn't break runnability guarantees
432(specifically, version 1 of those CPU models).  In future QEMU
433versions, aliases will point to newer CPU model versions
434depending on the machine type, so management software must
435resolve CPU model aliases before starting a virtual machine.
436
437QEMU guest agent
438----------------
439
440``--blacklist`` command line option (since 7.2)
441'''''''''''''''''''''''''''''''''''''''''''''''
442
443``--blacklist`` has been replaced by ``--block-rpcs`` (which is a better
444wording for what this option does). The short form ``-b`` still stays
445the same and thus is the preferred way for scripts that should run with
446both, older and future versions of QEMU.
447
448``blacklist`` config file option (since 7.2)
449''''''''''''''''''''''''''''''''''''''''''''
450
451The ``blacklist`` config file option has been renamed to ``block-rpcs``
452(to be in sync with the renaming of the corresponding command line
453option).
454
455Migration
456---------
457
458``skipped`` MigrationStats field (since 8.1)
459''''''''''''''''''''''''''''''''''''''''''''
460
461``skipped`` field in Migration stats has been deprecated.  It hasn't
462been used for more than 10 years.
463
464