xref: /qemu/docs/system/arm/virt.rst (revision 923710b6d5b21d9b3fcecc7e6719cfa5a53de268)
1'virt' generic virtual platform (``virt``)
2==========================================
3
4The ``virt`` board is a platform which does not correspond to any
5real hardware; it is designed for use in virtual machines.
6It is the recommended board type if you simply want to run
7a guest such as Linux and do not care about reproducing the
8idiosyncrasies and limitations of a particular bit of real-world
9hardware.
10
11This is a "versioned" board model, so as well as the ``virt`` machine
12type itself (which may have improvements, bugfixes and other minor
13changes between QEMU versions) a version is provided that guarantees
14to have the same behaviour as that of previous QEMU releases, so
15that VM migration will work between QEMU versions. For instance the
16``virt-5.0`` machine type will behave like the ``virt`` machine from
17the QEMU 5.0 release, and migration should work between ``virt-5.0``
18of the 5.0 release and ``virt-5.0`` of the 5.1 release. Migration
19is not guaranteed to work between different QEMU releases for
20the non-versioned ``virt`` machine type.
21
22VM migration is not guaranteed when using ``-cpu max``, as features
23supported may change between QEMU versions.  To ensure your VM can be
24migrated, it is recommended to use another cpu model instead.
25
26Supported devices
27"""""""""""""""""
28
29The virt board supports:
30
31- PCI/PCIe devices
32- Flash memory
33- Either one or two PL011 UARTs for the NonSecure World
34- An RTC
35- The fw_cfg device that allows a guest to obtain data from QEMU
36- A PL061 GPIO controller
37- An optional SMMUv3 IOMMU
38- hotpluggable DIMMs
39- hotpluggable NVDIMMs
40- An MSI controller (GICv2M or ITS). GICv2M is selected by default along
41  with GICv2. ITS is selected by default with GICv3 (>= virt-2.7). Note
42  that ITS is not modeled in TCG mode.
43- 32 virtio-mmio transport devices
44- running guests using the KVM accelerator on aarch64 hardware
45- large amounts of RAM (at least 255GB, and more if using highmem)
46- many CPUs (up to 512 if using a GICv3 and highmem)
47- Secure-World-only devices if the CPU has TrustZone:
48
49  - A second PL011 UART
50  - A second PL061 GPIO controller, with GPIO lines for triggering
51    a system reset or system poweroff
52  - A secure flash memory
53  - 16MB of secure RAM
54
55The second NonSecure UART only exists if a backend is configured
56explicitly (e.g. with a second -serial command line option) and
57TrustZone emulation is not enabled.
58
59Supported guest CPU types:
60
61- ``cortex-a7`` (32-bit)
62- ``cortex-a15`` (32-bit; the default)
63- ``cortex-a35`` (64-bit)
64- ``cortex-a53`` (64-bit)
65- ``cortex-a55`` (64-bit)
66- ``cortex-a57`` (64-bit)
67- ``cortex-a72`` (64-bit)
68- ``cortex-a76`` (64-bit)
69- ``cortex-a710`` (64-bit)
70- ``a64fx`` (64-bit)
71- ``host`` (with KVM only)
72- ``neoverse-n1`` (64-bit)
73- ``neoverse-v1`` (64-bit)
74- ``neoverse-n2`` (64-bit)
75- ``max`` (same as ``host`` for KVM; best possible emulation with TCG)
76
77Note that the default is ``cortex-a15``, so for an AArch64 guest you must
78specify a CPU type.
79
80Also, please note that passing ``max`` CPU (i.e. ``-cpu max``) won't
81enable all the CPU features for a given ``virt`` machine. Where a CPU
82architectural feature requires support in both the CPU itself and in the
83wider system (e.g. the MTE feature), it may not be enabled by default,
84but instead requires a machine option to enable it.
85
86For example, MTE support must be enabled with ``-machine virt,mte=on``,
87as well as by selecting an MTE-capable CPU (e.g., ``max``) with the
88``-cpu`` option.
89
90See the machine-specific options below, or check them for a given machine
91by passing the ``help`` suboption, like: ``-machine virt-9.0,help``.
92
93Graphics output is available, but unlike the x86 PC machine types
94there is no default display device enabled: you should select one from
95the Display devices section of "-device help". The recommended option
96is ``virtio-gpu-pci``; this is the only one which will work correctly
97with KVM. You may also need to ensure your guest kernel is configured
98with support for this; see below.
99
100Machine-specific options
101""""""""""""""""""""""""
102
103The following machine-specific options are supported:
104
105secure
106  Set ``on``/``off`` to enable/disable emulating a guest CPU which implements the
107  Arm Security Extensions (TrustZone). The default is ``off``.
108
109virtualization
110  Set ``on``/``off`` to enable/disable emulating a guest CPU which implements the
111  Arm Virtualization Extensions. The default is ``off``.
112
113mte
114  Set ``on``/``off`` to enable/disable emulating a guest CPU which implements the
115  Arm Memory Tagging Extensions. The default is ``off``.
116
117highmem
118  Set ``on``/``off`` to enable/disable placing devices and RAM in physical
119  address space above 32 bits. The default is ``on`` for machine types
120  later than ``virt-2.12`` when the CPU supports an address space
121  bigger than 32 bits (i.e. 64-bit CPUs, and 32-bit CPUs with the
122  Large Physical Address Extension (LPAE) feature). If you want to
123  boot a 32-bit kernel which does not have ``CONFIG_LPAE`` enabled on
124  a CPU type which implements LPAE, you will need to manually set
125  this to ``off``; otherwise some devices, such as the PCI controller,
126  will not be accessible.
127
128compact-highmem
129  Set ``on``/``off`` to enable/disable the compact layout for high memory regions.
130  The default is ``on`` for machine types later than ``virt-7.2``.
131
132highmem-redists
133  Set ``on``/``off`` to enable/disable the high memory region for GICv3 or
134  GICv4 redistributor. The default is ``on``. Setting this to ``off`` will
135  limit the maximum number of CPUs when GICv3 or GICv4 is used.
136
137highmem-ecam
138  Set ``on``/``off`` to enable/disable the high memory region for PCI ECAM.
139  The default is ``on`` for machine types later than ``virt-3.0``.
140
141highmem-mmio
142  Set ``on``/``off`` to enable/disable the high memory region for PCI MMIO.
143  The default is ``on``.
144
145gic-version
146  Specify the version of the Generic Interrupt Controller (GIC) to provide.
147  Valid values are:
148
149  ``2``
150    GICv2. Note that this limits the number of CPUs to 8.
151  ``3``
152    GICv3. This allows up to 512 CPUs.
153  ``4``
154    GICv4. Requires ``virtualization`` to be ``on``; allows up to 317 CPUs.
155  ``host``
156    Use the same GIC version the host provides, when using KVM
157  ``max``
158    Use the best GIC version possible (same as host when using KVM;
159    with TCG this is currently ``3`` if ``virtualization`` is ``off`` and
160    ``4`` if ``virtualization`` is ``on``, but this may change in future)
161
162its
163  Set ``on``/``off`` to enable/disable ITS instantiation. The default is ``on``
164  for machine types later than ``virt-2.7``.
165
166iommu
167  Set the IOMMU type to create for the guest. Valid values are:
168
169  ``none``
170    Don't create an IOMMU (the default)
171  ``smmuv3``
172    Create an SMMUv3
173
174default-bus-bypass-iommu
175  Set ``on``/``off`` to enable/disable `bypass_iommu
176  <https://gitlab.com/qemu-project/qemu/-/blob/master/docs/bypass-iommu.txt>`_
177  for default root bus.
178
179ras
180  Set ``on``/``off`` to enable/disable reporting host memory errors to a guest
181  using ACPI and guest external abort exceptions. The default is off.
182
183acpi
184  Set ``on``/``off``/``auto`` to enable/disable ACPI.
185
186dtb-randomness
187  Set ``on``/``off`` to pass random seeds via the guest DTB
188  rng-seed and kaslr-seed nodes (in both "/chosen" and
189  "/secure-chosen") to use for features like the random number
190  generator and address space randomisation. The default is
191  ``on``. You will want to disable it if your trusted boot chain
192  will verify the DTB it is passed, since this option causes the
193  DTB to be non-deterministic. It would be the responsibility of
194  the firmware to come up with a seed and pass it on if it wants to.
195
196dtb-kaslr-seed
197  A deprecated synonym for dtb-randomness.
198
199x-oem-id
200  Set string (up to 6 bytes) to override the default value of field OEMID in ACPI
201  table header.
202
203x-oem-table-id
204  Set string (up to 8 bytes) to override the default value of field OEM Table ID
205  in ACPI table header.
206
207Linux guest kernel configuration
208""""""""""""""""""""""""""""""""
209
210The 'defconfig' for Linux arm and arm64 kernels should include the
211right device drivers for virtio and the PCI controller; however some older
212kernel versions, especially for 32-bit Arm, did not have everything
213enabled by default. If you're not seeing PCI devices that you expect,
214then check that your guest config has::
215
216  CONFIG_PCI=y
217  CONFIG_VIRTIO_PCI=y
218  CONFIG_PCI_HOST_GENERIC=y
219
220If you want to use the ``virtio-gpu-pci`` graphics device you will also
221need::
222
223  CONFIG_DRM=y
224  CONFIG_DRM_VIRTIO_GPU=y
225
226Hardware configuration information for bare-metal programming
227"""""""""""""""""""""""""""""""""""""""""""""""""""""""""""""
228
229The ``virt`` board automatically generates a device tree blob ("dtb")
230which it passes to the guest. This provides information about the
231addresses, interrupt lines and other configuration of the various devices
232in the system. Guest code can rely on and hard-code the following
233addresses:
234
235- Flash memory starts at address 0x0000_0000
236
237- RAM starts at 0x4000_0000
238
239All other information about device locations may change between
240QEMU versions, so guest code must look in the DTB.
241
242QEMU supports two types of guest image boot for ``virt``, and
243the way for the guest code to locate the dtb binary differs:
244
245- For guests using the Linux kernel boot protocol (this means any
246  non-ELF file passed to the QEMU ``-kernel`` option) the address
247  of the DTB is passed in a register (``r2`` for 32-bit guests,
248  or ``x0`` for 64-bit guests)
249
250- For guests booting as "bare-metal" (any other kind of boot),
251  the DTB is at the start of RAM (0x4000_0000)
252