One Hat Cyber Team

Edit File: cpu-features.html

<!DOCTYPE html>

 <html class="no-js" lang="en" > 
<head>
 <meta charset="utf-8">
 
 <meta name="viewport" content="width=device-width, initial-scale=1.0">
 
 <title>Arm CPU Features &mdash; QEMU qemu-kvm-6.2.0-53.module+el8.10.0+2055+8eb7870b.4 documentation</title>

</head>

<a href="../../index.html" class="icon icon-home"> QEMU

</div>

<div class="wy-menu wy-menu-vertical" data-spy="affix" role="navigation" aria-label="main navigation">
 
 
 
 
 
 
 Contents:
<ul class="current">
<li class="toctree-l1"><a class="reference internal" href="../../about/index.html">About QEMU</a></li>
<li class="toctree-l1 current"><a class="reference internal" href="../index.html">System Emulation</a><ul class="current">
<li class="toctree-l2"><a class="reference internal" href="../quickstart.html">Quick Start</a></li>
<li class="toctree-l2"><a class="reference internal" href="../invocation.html">Invocation</a></li>
<li class="toctree-l2"><a class="reference internal" href="../device-emulation.html">Device Emulation</a></li>
<li class="toctree-l2"><a class="reference internal" href="../keys.html">Keys in the graphical frontends</a></li>
<li class="toctree-l2"><a class="reference internal" href="../mux-chardev.html">Keys in the character backend multiplexer</a></li>
<li class="toctree-l2"><a class="reference internal" href="../monitor.html">QEMU Monitor</a></li>
<li class="toctree-l2"><a class="reference internal" href="../images.html">Disk Images</a></li>
<li class="toctree-l2"><a class="reference internal" href="../virtio-net-failover.html">QEMU virtio-net standby (net_failover)</a></li>
<li class="toctree-l2"><a class="reference internal" href="../linuxboot.html">Direct Linux Boot</a></li>
<li class="toctree-l2"><a class="reference internal" href="../generic-loader.html">Generic Loader</a></li>
<li class="toctree-l2"><a class="reference internal" href="../guest-loader.html">Guest Loader</a></li>
<li class="toctree-l2"><a class="reference internal" href="../barrier.html">QEMU Barrier Client</a></li>
<li class="toctree-l2"><a class="reference internal" href="../vnc-security.html">VNC security</a></li>
<li class="toctree-l2"><a class="reference internal" href="../tls.html">TLS setup for network services</a></li>
<li class="toctree-l2"><a class="reference internal" href="../secrets.html">Providing secret data to QEMU</a></li>
<li class="toctree-l2"><a class="reference internal" href="../authz.html">Client authorization</a></li>
<li class="toctree-l2"><a class="reference internal" href="../gdb.html">GDB usage</a></li>
<li class="toctree-l2"><a class="reference internal" href="../managed-startup.html">Managed start up options</a></li>
<li class="toctree-l2"><a class="reference internal" href="../bootindex.html">Managing device boot order with bootindex properties</a></li>
<li class="toctree-l2"><a class="reference internal" href="../cpu-hotplug.html">Virtual CPU hotplug</a></li>
<li class="toctree-l2"><a class="reference internal" href="../pr-manager.html">Persistent reservation managers</a></li>
<li class="toctree-l2 current"><a class="reference internal" href="../targets.html">QEMU System Emulator Targets</a><ul class="current">
<li class="toctree-l3 current"><a class="reference internal" href="../target-arm.html">Arm System emulator</a><ul class="current">
<li class="toctree-l4"><a class="reference internal" href="../target-arm.html#choosing-a-board-model">Choosing a board model</a></li>
<li class="toctree-l4"><a class="reference internal" href="../target-arm.html#board-specific-documentation">Board-specific documentation</a></li>
<li class="toctree-l4"><a class="reference internal" href="../target-arm.html#emulated-cpu-architecture-support">Emulated CPU architecture support</a></li>
<li class="toctree-l4 current"><a class="reference internal" href="../target-arm.html#arm-cpu-features">Arm CPU features</a></li>
</ul>
</li>
<li class="toctree-l3"><a class="reference internal" href="../target-avr.html">AVR System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-m68k.html">ColdFire System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-mips.html">MIPS System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-ppc.html">PowerPC System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-riscv.html">RISC-V System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-rx.html">RX System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-s390x.html">s390x System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-sparc.html">Sparc32 System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-sparc64.html">Sparc64 System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-i386.html">x86 System emulator</a></li>
<li class="toctree-l3"><a class="reference internal" href="../target-xtensa.html">Xtensa System emulator</a></li>
</ul>
</li>
<li class="toctree-l2"><a class="reference internal" href="../security.html">Security</a></li>
<li class="toctree-l2"><a class="reference internal" href="../multi-process.html">Multi-process QEMU</a></li>
</ul>
</li>
<li class="toctree-l1"><a class="reference internal" href="../../user/index.html">User Mode Emulation</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../tools/index.html">Tools</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../interop/index.html">System Emulation Management and Interoperability</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../specs/index.html">System Emulation Guest Hardware Specifications</a></li>
<li class="toctree-l1"><a class="reference internal" href="../../devel/index.html">Developer Information</a></li>
</ul>

</div>
 </div>
 </nav>

<ul class="wy-breadcrumbs">
 
 <li><a href="../../index.html">Docs</a> &raquo;</li>
 
 <li><a href="../index.html">System Emulation</a> &raquo;</li>
 
 <li><a href="../targets.html">QEMU System Emulator Targets</a> &raquo;</li>
 
 <li><a href="../target-arm.html">Arm System emulator</a> &raquo;</li>
 
 <li>Arm CPU Features</li>
 
 
 <li class="wy-breadcrumbs-aside">
 
 
 
 <a href="https://gitlab.com/qemu-project/qemu/blob/master/docs/system/arm/cpu-features.rst" class="fa fa-gitlab"> Edit on GitLab</a>
 
 
 
 </li>
 
 </ul>

<hr/>
</div>
 <div role="main" class="document" itemscope="itemscope" itemtype="http://schema.org/Article">
 <div itemprop="articleBody">
 
 <div class="section" id="arm-cpu-features">
<h1>Arm CPU Features<a class="headerlink" href="#arm-cpu-features" title="Permalink to this headline">¶</a></h1>
CPU features are optional features that a CPU of supporting type may
choose to implement or not. In QEMU, optional CPU features have
corresponding boolean CPU proprieties that, when enabled, indicate
that the feature is implemented, and, conversely, when disabled,
indicate that it is not implemented. An example of an Arm CPU feature
is the Performance Monitoring Unit (PMU). CPU types such as the
Cortex-A15 and the Cortex-A57, which respectively implement Arm
architecture reference manuals ARMv7-A and ARMv8-A, may both optionally
implement PMUs. For example, if a user wants to use a Cortex-A15 without
a PMU, then the <code class="docutils literal notranslate">-cpu</code> parameter should contain <code class="docutils literal notranslate">pmu=off</code> on the QEMU
command line, i.e. <code class="docutils literal notranslate">-cpu cortex-a15,pmu=off</code>.
As not all CPU types support all optional CPU features, then whether or
not a CPU property exists depends on the CPU type. For example, CPUs
that implement the ARMv8-A architecture reference manual may optionally
support the AArch32 CPU feature, which may be enabled by disabling the
<code class="docutils literal notranslate">aarch64</code> CPU property. A CPU type such as the Cortex-A15, which does
not implement ARMv8-A, will not have the <code class="docutils literal notranslate">aarch64</code> CPU property.
QEMU’s support may be limited for some CPU features, only partially
supporting the feature or only supporting the feature under certain
configurations. For example, the <code class="docutils literal notranslate">aarch64</code> CPU feature, which, when
disabled, enables the optional AArch32 CPU feature, is only supported
when using the KVM accelerator and when running on a host CPU type that
supports the feature. While <code class="docutils literal notranslate">aarch64</code> currently only works with KVM,
it could work with TCG. CPU features that are specific to KVM are
prefixed with “kvm-” and are described in “KVM VCPU Features”.
</div>
<div class="section" id="cpu-feature-probing">
<h1>CPU Feature Probing<a class="headerlink" href="#cpu-feature-probing" title="Permalink to this headline">¶</a></h1>
Determining which CPU features are available and functional for a given
CPU type is possible with the <code class="docutils literal notranslate">query-cpu-model-expansion</code> QMP command.
Below are some examples where <code class="docutils literal notranslate">scripts/qmp/qmp-shell</code> (see the top comment
block in the script for usage) is used to issue the QMP commands.
<ol class="arabic">
<li>Determine which CPU features are available for the <code class="docutils literal notranslate">max</code> CPU type
(Note, we started QEMU with qemu-system-aarch64, so <code class="docutils literal notranslate">max</code> is
implementing the ARMv8-A reference manual in this case):
<div class="highlight-default notranslate"><div class="highlight"><pre>(QEMU) query-cpu-model-expansion type=full model={&quot;name&quot;:&quot;max&quot;}
{ &quot;return&quot;: {
 &quot;model&quot;: { &quot;name&quot;: &quot;max&quot;, &quot;props&quot;: {
 &quot;sve1664&quot;: true, &quot;pmu&quot;: true, &quot;sve1792&quot;: true, &quot;sve1920&quot;: true,
 &quot;sve128&quot;: true, &quot;aarch64&quot;: true, &quot;sve1024&quot;: true, &quot;sve&quot;: true,
 &quot;sve640&quot;: true, &quot;sve768&quot;: true, &quot;sve1408&quot;: true, &quot;sve256&quot;: true,
 &quot;sve1152&quot;: true, &quot;sve512&quot;: true, &quot;sve384&quot;: true, &quot;sve1536&quot;: true,
 &quot;sve896&quot;: true, &quot;sve1280&quot;: true, &quot;sve2048&quot;: true
}}}}
</pre></div>
</div>
</li>
</ol>
We see that the <code class="docutils literal notranslate">max</code> CPU type has the <code class="docutils literal notranslate">pmu</code>, <code class="docutils literal notranslate">aarch64</code>, <code class="docutils literal notranslate">sve</code>, and many
<code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU features. We also see that all the CPU features are
enabled, as they are all <code class="docutils literal notranslate">true</code>. (The <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU features are all
optional SVE vector lengths (see “SVE CPU Properties”). While with TCG
all SVE vector lengths can be supported, when KVM is in use it’s more
likely that only a few lengths will be supported, if SVE is supported at
all.)
<ol class="arabic" start="2">
<li>Let’s try to disable the PMU:
<div class="highlight-default notranslate"><div class="highlight"><pre>(QEMU) query-cpu-model-expansion type=full model={&quot;name&quot;:&quot;max&quot;,&quot;props&quot;:{&quot;pmu&quot;:false}}
{ &quot;return&quot;: {
 &quot;model&quot;: { &quot;name&quot;: &quot;max&quot;, &quot;props&quot;: {
 &quot;sve1664&quot;: true, &quot;pmu&quot;: false, &quot;sve1792&quot;: true, &quot;sve1920&quot;: true,
 &quot;sve128&quot;: true, &quot;aarch64&quot;: true, &quot;sve1024&quot;: true, &quot;sve&quot;: true,
 &quot;sve640&quot;: true, &quot;sve768&quot;: true, &quot;sve1408&quot;: true, &quot;sve256&quot;: true,
 &quot;sve1152&quot;: true, &quot;sve512&quot;: true, &quot;sve384&quot;: true, &quot;sve1536&quot;: true,
 &quot;sve896&quot;: true, &quot;sve1280&quot;: true, &quot;sve2048&quot;: true
}}}}
</pre></div>
</div>
</li>
</ol>
We see it worked, as <code class="docutils literal notranslate">pmu</code> is now <code class="docutils literal notranslate">false</code>.
<ol class="arabic" start="3">
<li>Let’s try to disable <code class="docutils literal notranslate">aarch64</code>, which enables the AArch32 CPU feature:
<div class="highlight-default notranslate"><div class="highlight"><pre>(QEMU) query-cpu-model-expansion type=full model={&quot;name&quot;:&quot;max&quot;,&quot;props&quot;:{&quot;aarch64&quot;:false}}
{&quot;error&quot;: {
 &quot;class&quot;: &quot;GenericError&quot;, &quot;desc&quot;:
 &quot;&#39;aarch64&#39; feature cannot be disabled unless KVM is enabled and 32-bit EL1 is supported&quot;
}}
</pre></div>
</div>
</li>
</ol>
It looks like this feature is limited to a configuration we do not
currently have.
<ol class="arabic" start="4">
<li>Let’s disable <code class="docutils literal notranslate">sve</code> and see what happens to all the optional SVE
vector lengths:
<div class="highlight-default notranslate"><div class="highlight"><pre>(QEMU) query-cpu-model-expansion type=full model={&quot;name&quot;:&quot;max&quot;,&quot;props&quot;:{&quot;sve&quot;:false}}
{ &quot;return&quot;: {
 &quot;model&quot;: { &quot;name&quot;: &quot;max&quot;, &quot;props&quot;: {
 &quot;sve1664&quot;: false, &quot;pmu&quot;: true, &quot;sve1792&quot;: false, &quot;sve1920&quot;: false,
 &quot;sve128&quot;: false, &quot;aarch64&quot;: true, &quot;sve1024&quot;: false, &quot;sve&quot;: false,
 &quot;sve640&quot;: false, &quot;sve768&quot;: false, &quot;sve1408&quot;: false, &quot;sve256&quot;: false,
 &quot;sve1152&quot;: false, &quot;sve512&quot;: false, &quot;sve384&quot;: false, &quot;sve1536&quot;: false,
 &quot;sve896&quot;: false, &quot;sve1280&quot;: false, &quot;sve2048&quot;: false
}}}}
</pre></div>
</div>
</li>
</ol>
As expected they are now all <code class="docutils literal notranslate">false</code>.
<ol class="arabic" start="5">
<li>Let’s try probing CPU features for the Cortex-A15 CPU type:
<div class="highlight-default notranslate"><div class="highlight"><pre>(QEMU) query-cpu-model-expansion type=full model={&quot;name&quot;:&quot;cortex-a15&quot;}
{&quot;return&quot;: {&quot;model&quot;: {&quot;name&quot;: &quot;cortex-a15&quot;, &quot;props&quot;: {&quot;pmu&quot;: true}}}}
</pre></div>
</div>
</li>
</ol>
Only the <code class="docutils literal notranslate">pmu</code> CPU feature is available.
<div class="section" id="a-note-about-cpu-feature-dependencies">
<h2>A note about CPU feature dependencies<a class="headerlink" href="#a-note-about-cpu-feature-dependencies" title="Permalink to this headline">¶</a></h2>
It’s possible for features to have dependencies on other features. I.e.
it may be possible to change one feature at a time without error, but
when attempting to change all features at once an error could occur
depending on the order they are processed. It’s also possible changing
all at once doesn’t generate an error, because a feature’s dependencies
are satisfied with other features, but the same feature cannot be changed
independently without error. For these reasons callers should always
attempt to make their desired changes all at once in order to ensure the
collection is valid.
</div>
<div class="section" id="a-note-about-cpu-models-and-kvm">
<h2>A note about CPU models and KVM<a class="headerlink" href="#a-note-about-cpu-models-and-kvm" title="Permalink to this headline">¶</a></h2>
Named CPU models generally do not work with KVM. There are a few cases
that do work, e.g. using the named CPU model <code class="docutils literal notranslate">cortex-a57</code> with KVM on a
seattle host, but mostly if KVM is enabled the <code class="docutils literal notranslate">host</code> CPU type must be
used. This means the guest is provided all the same CPU features as the
host CPU type has. And, for this reason, the <code class="docutils literal notranslate">host</code> CPU type should
enable all CPU features that the host has by default. Indeed it’s even
a bit strange to allow disabling CPU features that the host has when using
the <code class="docutils literal notranslate">host</code> CPU type, but in the absence of CPU models it’s the best we can
do if we want to launch guests without all the host’s CPU features enabled.
Enabling KVM also affects the <code class="docutils literal notranslate">query-cpu-model-expansion</code> QMP command. The
affect is not only limited to specific features, as pointed out in example
(3) of “CPU Feature Probing”, but also to which CPU types may be expanded.
When KVM is enabled, only the <code class="docutils literal notranslate">max</code>, <code class="docutils literal notranslate">host</code>, and current CPU type may be
expanded. This restriction is necessary as it’s not possible to know all
CPU types that may work with KVM, but it does impose a small risk of users
experiencing unexpected errors. For example on a seattle, as mentioned
above, the <code class="docutils literal notranslate">cortex-a57</code> CPU type is also valid when KVM is enabled.
Therefore a user could use the <code class="docutils literal notranslate">host</code> CPU type for the current type, but
then attempt to query <code class="docutils literal notranslate">cortex-a57</code>, however that query will fail with our
restrictions. This shouldn’t be an issue though as management layers and
users have been preferring the <code class="docutils literal notranslate">host</code> CPU type for use with KVM for quite
some time. Additionally, if the KVM-enabled QEMU instance running on a
seattle host is using the <code class="docutils literal notranslate">cortex-a57</code> CPU type, then querying <code class="docutils literal notranslate">cortex-a57</code>
will work.
</div>
</div>
<div class="section" id="using-cpu-features">
<h1>Using CPU Features<a class="headerlink" href="#using-cpu-features" title="Permalink to this headline">¶</a></h1>
After determining which CPU features are available and supported for a
given CPU type, then they may be selectively enabled or disabled on the
QEMU command line with that CPU type:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,pmu=off,sve=on,sve128=on,sve256=on
</pre></div>
</div>
The example above disables the PMU and enables the first two SVE vector
lengths for the <code class="docutils literal notranslate">max</code> CPU type. Note, the <code class="docutils literal notranslate">sve=on</code> isn’t actually
necessary, because, as we observed above with our probe of the <code class="docutils literal notranslate">max</code> CPU
type, <code class="docutils literal notranslate">sve</code> is already on by default. Also, based on our probe of
defaults, it would seem we need to disable many SVE vector lengths, rather
than only enabling the two we want. This isn’t the case, because, as
disabling many SVE vector lengths would be quite verbose, the <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU
properties have special semantics (see “SVE CPU Property Parsing
Semantics”).
</div>
<div class="section" id="kvm-vcpu-features">
<h1>KVM VCPU Features<a class="headerlink" href="#kvm-vcpu-features" title="Permalink to this headline">¶</a></h1>
KVM VCPU features are CPU features that are specific to KVM, such as
paravirt features or features that enable CPU virtualization extensions.
The features’ CPU properties are only available when KVM is enabled and
are named with the prefix “kvm-“. KVM VCPU features may be probed,
enabled, and disabled in the same way as other CPU features. Below is
the list of KVM VCPU features and their descriptions.
<blockquote>
<div><dl class="docutils">
<dt>kvm-no-adjvtime By default kvm-no-adjvtime is disabled. This</dt>
<dd>means that by default the virtual time
adjustment is enabled (vtime is not not
adjusted).
When virtual time adjustment is enabled each
time the VM transitions back to running state
the VCPU’s virtual counter is updated to ensure
stopped time is not counted. This avoids time
jumps surprising guest OSes and applications,
as long as they use the virtual counter for
timekeeping. However it has the side effect of
the virtual and physical counters diverging.
All timekeeping based on the virtual counter
will appear to lag behind any timekeeping that
does not subtract VM stopped time. The guest
may resynchronize its virtual counter with
other time sources as needed.
Enable kvm-no-adjvtime to disable virtual time
adjustment, also restoring the legacy (pre-5.0)
behavior.
</dd>
<dt>kvm-steal-time Since v5.2, kvm-steal-time is enabled by</dt>
<dd>default when KVM is enabled, the feature is
supported, and the guest is 64-bit.
When kvm-steal-time is enabled a 64-bit guest
can account for time its CPUs were not running
due to the host not scheduling the corresponding
VCPU threads. The accounting statistics may
influence the guest scheduler behavior and/or be
exposed to the guest userspace.
</dd>
</dl>
</div></blockquote>
</div>
<div class="section" id="tcg-vcpu-features">
<h1>TCG VCPU Features<a class="headerlink" href="#tcg-vcpu-features" title="Permalink to this headline">¶</a></h1>
TCG VCPU features are CPU features that are specific to TCG.
Below is the list of TCG VCPU features and their descriptions.
<blockquote>
<div><dl class="docutils">
<dt>pauth Enable or disable <code class="docutils literal notranslate">FEAT_Pauth</code>, pointer</dt>
<dd>authentication. By default, the feature is
enabled with <code class="docutils literal notranslate">-cpu max</code>.</dd>
<dt>pauth-impdef When <code class="docutils literal notranslate">FEAT_Pauth</code> is enabled, either the</dt>
<dd>impdef (Implementation Defined) algorithm
is enabled or the architected QARMA algorithm
is enabled. By default the impdef algorithm
is disabled, and QARMA is enabled.
The architected QARMA algorithm has good
cryptographic properties, but can be quite slow
to emulate. The impdef algorithm used by QEMU
is non-cryptographic but significantly faster.
</dd>
</dl>
</div></blockquote>
</div>
<div class="section" id="sve-cpu-properties">
<h1>SVE CPU Properties<a class="headerlink" href="#sve-cpu-properties" title="Permalink to this headline">¶</a></h1>
There are two types of SVE CPU properties: <code class="docutils literal notranslate">sve</code> and <code class="docutils literal notranslate">sve&lt;N&gt;</code>. The first
is used to enable or disable the entire SVE feature, just as the <code class="docutils literal notranslate">pmu</code>
CPU property completely enables or disables the PMU. The second type
is used to enable or disable specific vector lengths, where <code class="docutils literal notranslate">N</code> is the
number of bits of the length. The <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU properties have special
dependencies and constraints, see “SVE CPU Property Dependencies and
Constraints” below. Additionally, as we want all supported vector lengths
to be enabled by default, then, in order to avoid overly verbose command
lines (command lines full of <code class="docutils literal notranslate">sve&lt;N&gt;=off</code>, for all <code class="docutils literal notranslate">N</code> not wanted), we
provide the parsing semantics listed in “SVE CPU Property Parsing
Semantics”.
<div class="section" id="sve-cpu-property-dependencies-and-constraints">
<h2>SVE CPU Property Dependencies and Constraints<a class="headerlink" href="#sve-cpu-property-dependencies-and-constraints" title="Permalink to this headline">¶</a></h2>
<blockquote>
<div><ol class="arabic simple">
<li>At least one vector length must be enabled when <code class="docutils literal notranslate">sve</code> is enabled.</li>
<li>If a vector length <code class="docutils literal notranslate">N</code> is enabled, then, when KVM is enabled, all
smaller, host supported vector lengths must also be enabled. If
KVM is not enabled, then only all the smaller, power-of-two vector
lengths must be enabled. E.g. with KVM if the host supports all
vector lengths up to 512-bits (128, 256, 384, 512), then if <code class="docutils literal notranslate">sve512</code>
is enabled, the 128-bit vector length, 256-bit vector length, and
384-bit vector length must also be enabled. Without KVM, the 384-bit
vector length would not be required.</li>
<li>If KVM is enabled then only vector lengths that the host CPU type
support may be enabled. If SVE is not supported by the host, then
no <code class="docutils literal notranslate">sve*</code> properties may be enabled.</li>
</ol>
</div></blockquote>
</div>
<div class="section" id="sve-cpu-property-parsing-semantics">
<h2>SVE CPU Property Parsing Semantics<a class="headerlink" href="#sve-cpu-property-parsing-semantics" title="Permalink to this headline">¶</a></h2>
<blockquote>
<div><ol class="arabic simple">
<li>If SVE is disabled (<code class="docutils literal notranslate">sve=off</code>), then which SVE vector lengths
are enabled or disabled is irrelevant to the guest, as the entire
SVE feature is disabled and that disables all vector lengths for
the guest. However QEMU will still track any <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU
properties provided by the user. If later an <code class="docutils literal notranslate">sve=on</code> is provided,
then the guest will get only the enabled lengths. If no <code class="docutils literal notranslate">sve=on</code>
is provided and there are explicitly enabled vector lengths, then
an error is generated.</li>
<li>If SVE is enabled (<code class="docutils literal notranslate">sve=on</code>), but no <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU properties are
provided, then all supported vector lengths are enabled, which when
KVM is not in use means including the non-power-of-two lengths, and,
when KVM is in use, it means all vector lengths supported by the host
processor.</li>
<li>If SVE is enabled, then an error is generated when attempting to
disable the last enabled vector length (see constraint (1) of “SVE
CPU Property Dependencies and Constraints”).</li>
<li>If one or more vector lengths have been explicitly enabled and at
at least one of the dependency lengths of the maximum enabled length
has been explicitly disabled, then an error is generated (see
constraint (2) of “SVE CPU Property Dependencies and Constraints”).</li>
<li>When KVM is enabled, if the host does not support SVE, then an error
is generated when attempting to enable any <code class="docutils literal notranslate">sve*</code> properties (see
constraint (3) of “SVE CPU Property Dependencies and Constraints”).</li>
<li>When KVM is enabled, if the host does support SVE, then an error is
generated when attempting to enable any vector lengths not supported
by the host (see constraint (3) of “SVE CPU Property Dependencies and
Constraints”).</li>
<li>If one or more <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU properties are set <code class="docutils literal notranslate">off</code>, but no <code class="docutils literal notranslate">sve&lt;N&gt;</code>,
CPU properties are set <code class="docutils literal notranslate">on</code>, then the specified vector lengths are
disabled but the default for any unspecified lengths remains enabled.
When KVM is not enabled, disabling a power-of-two vector length also
disables all vector lengths larger than the power-of-two length.
When KVM is enabled, then disabling any supported vector length also
disables all larger vector lengths (see constraint (2) of “SVE CPU
Property Dependencies and Constraints”).</li>
<li>If one or more <code class="docutils literal notranslate">sve&lt;N&gt;</code> CPU properties are set to <code class="docutils literal notranslate">on</code>, then they
are enabled and all unspecified lengths default to disabled, except
for the required lengths per constraint (2) of “SVE CPU Property
Dependencies and Constraints”, which will even be auto-enabled if
they were not explicitly enabled.</li>
<li>If SVE was disabled (<code class="docutils literal notranslate">sve=off</code>), allowing all vector lengths to be
explicitly disabled (i.e. avoiding the error specified in (3) of
“SVE CPU Property Parsing Semantics”), then if later an <code class="docutils literal notranslate">sve=on</code> is
provided an error will be generated. To avoid this error, one must
enable at least one vector length prior to enabling SVE.</li>
</ol>
</div></blockquote>
</div>
<div class="section" id="sve-cpu-property-examples">
<h2>SVE CPU Property Examples<a class="headerlink" href="#sve-cpu-property-examples" title="Permalink to this headline">¶</a></h2>
<blockquote>
<div><ol class="arabic">
<li>Disable SVE:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve=off
</pre></div>
</div>
</li>
<li>Implicitly enable all vector lengths for the <code class="docutils literal notranslate">max</code> CPU type:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max
</pre></div>
</div>
</li>
<li>When KVM is enabled, implicitly enable all host CPU supported vector
lengths with the <code class="docutils literal notranslate">host</code> CPU type:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt,accel=kvm -cpu host
</pre></div>
</div>
</li>
<li>Only enable the 128-bit vector length:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve128=on
</pre></div>
</div>
</li>
<li>Disable the 512-bit vector length and all larger vector lengths,
since 512 is a power-of-two. This results in all the smaller,
uninitialized lengths (128, 256, and 384) defaulting to enabled:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve512=off
</pre></div>
</div>
</li>
<li>Enable the 128-bit, 256-bit, and 512-bit vector lengths:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve128=on,sve256=on,sve512=on
</pre></div>
</div>
</li>
<li>The same as (6), but since the 128-bit and 256-bit vector
lengths are required for the 512-bit vector length to be enabled,
then allow them to be auto-enabled:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve512=on
</pre></div>
</div>
</li>
<li>Do the same as (7), but by first disabling SVE and then re-enabling it:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve=off,sve512=on,sve=on
</pre></div>
</div>
</li>
<li>Force errors regarding the last vector length:
<div class="highlight-default notranslate"><div class="highlight"><pre>$ qemu-system-aarch64 -M virt -cpu max,sve128=off
$ qemu-system-aarch64 -M virt -cpu max,sve=off,sve128=off,sve=on
</pre></div>
</div>
</li>
</ol>
</div></blockquote>
</div>
<div class="section" id="sve-cpu-property-recommendations">
<h2>SVE CPU Property Recommendations<a class="headerlink" href="#sve-cpu-property-recommendations" title="Permalink to this headline">¶</a></h2>
The examples in “SVE CPU Property Examples” exhibit many ways to select
vector lengths which developers may find useful in order to avoid overly
verbose command lines. However, the recommended way to select vector
lengths is to explicitly enable each desired length. Therefore only
example’s (1), (4), and (6) exhibit recommended uses of the properties.
</div>
<div class="section" id="sve-user-mode-default-vector-length-property">
<h2>SVE User-mode Default Vector Length Property<a class="headerlink" href="#sve-user-mode-default-vector-length-property" title="Permalink to this headline">¶</a></h2>
For qemu-aarch64, the cpu property <code class="docutils literal notranslate">sve-default-vector-length=N</code> is
defined to mirror the Linux kernel parameter file
<code class="docutils literal notranslate">/proc/sys/abi/sve_default_vector_length</code>. The default length, <code class="docutils literal notranslate">N</code>,
is in units of bytes and must be between 16 and 8192.
If not specified, the default vector length is 64.
If the default length is larger than the maximum vector length enabled,
the actual vector length will be reduced. Note that the maximum vector
length supported by QEMU is 256.
If this property is set to <code class="docutils literal notranslate">-1</code> then the default vector length
is set to the maximum possible length.
</div>
</div>

</div>
 
 </div>
 <footer>
 
 <div class="rst-footer-buttons" role="navigation" aria-label="footer navigation">
 
 <a href="../target-avr.html" class="btn btn-neutral float-right" title="AVR System emulator" accesskey="n" rel="next">Next </a>
 
 
 <a href="emulation.html" class="btn btn-neutral" title="A-profile CPU architecture support" accesskey="p" rel="prev"> Previous</a>
 
 </div>

<hr/>

<div role="contentinfo">
 
 &copy; Copyright 2021, The QEMU Project Developers.

</div>
 Built with <a href="http://sphinx-doc.org/">Sphinx</a> using a <a href="https://github.com/rtfd/sphinx_rtd_theme">theme</a> provided by <a href="https://readthedocs.org">Read the Docs</a>.

This documentation is for QEMU version 6.2.0.

<a href="../../about/license.html">QEMU and this manual are released under the
GNU General Public License, version 2.</a>

</footer>

</div>
 </div>

</section>

</div>

</body>
</html>