<h2 id="cargo_bench_name">NAME</h2>
<div class="sectionbody">
<p>cargo-bench - Execute benchmarks of a package</p>
</div>
<div class="sect1">
<h2 id="cargo_bench_synopsis">SYNOPSIS</h2>
<div class="sectionbody">
<div class="paragraph">
<p><code>cargo bench [<em>OPTIONS</em>] [BENCHNAME] [-- <em>BENCH-OPTIONS</em>]</code></p>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_description">DESCRIPTION</h2>
<div class="sectionbody">
<div class="paragraph">
<p>Compile and execute benchmarks.</p>
</div>
<div class="paragraph">
<p>The benchmark filtering argument <code>BENCHNAME</code> and all the arguments following
the two dashes (<code>--</code>) are passed to the benchmark binaries and thus to
<em>libtest</em> (rustc’s built in unit-test and micro-benchmarking framework). If
you’re passing arguments to both Cargo and the binary, the ones after <code>--</code> go
to the binary, the ones before go to Cargo. For details about libtest’s
arguments see the output of <code>cargo bench — --help</code>. As an example, this will
run only the benchmark named <code>foo</code> (and skip other similarly named benchmarks
like <code>foobar</code>):</p>
</div>
<div class="literalblock">
<div class="content">
<pre>cargo bench -- foo --exact</pre>
</div>
</div>
<div class="paragraph">
<p>Benchmarks are built with the <code>--test</code> option to <code>rustc</code> which creates an
executable with a <code>main</code> function that automatically runs all functions
annotated with the <code>#[bench]</code> attribute. Cargo passes the <code>--bench</code> flag to
the test harness to tell it to run only benchmarks.</p>
</div>
<div class="paragraph">
<p>The libtest harness may be disabled by setting <code>harness = false</code> in the target
manifest settings, in which case your code will need to provide its own <code>main</code>
function to handle running benchmarks.</p>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_options">OPTIONS</h2>
<div class="sectionbody">
<div class="sect2">
<h3 id="cargo_bench_benchmark_options">Benchmark Options</h3>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--no-run</strong></dt>
<dd>
<p>Compile, but don’t run benchmarks.</p>
</dd>
<dt class="hdlist1"><strong>--no-fail-fast</strong></dt>
<dd>
<p>Run all benchmarks regardless of failure. Without this flag, Cargo will exit
after the first executable fails. The Rust test harness will run all
benchmarks within the executable to completion, this flag only applies to
the executable as a whole.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_package_selection">Package Selection</h3>
<div class="paragraph">
<p>By default, when no package selection options are given, the packages selected
depend on the current working directory. In the root of a virtual workspace,
all workspace members are selected (<code>--all</code> is implied). Otherwise, only the
package in the current directory will be selected. The default packages may be
overridden with the <code>workspace.default-members</code> key in the root <code>Cargo.toml</code>
manifest.</p>
</div>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>-p</strong> <em>SPEC</em>…​</dt>
<dt class="hdlist1"><strong>--package</strong> <em>SPEC</em>…​</dt>
<dd>
<p>Benchmark only the specified packages. See <a href="commands/cargo-pkgid.html">cargo-pkgid(1)</a> for the
SPEC format. This flag may be specified multiple times.</p>
</dd>
<dt class="hdlist1"><strong>--all</strong></dt>
<dd>
<p>Benchmark all members in the workspace.</p>
</dd>
<dt class="hdlist1"><strong>--exclude</strong> <em>SPEC</em>…​</dt>
<dd>
<p>Exclude the specified packages. Must be used in conjunction with the
<code>--all</code> flag. This flag may be specified multiple times.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_target_selection">Target Selection</h3>
<div class="paragraph">
<p>When no target selection options are given, <code>cargo bench</code> will build the
following targets of the selected packages:</p>
</div>
<div class="ulist">
<ul>
<li>
<p>lib — used to link with binaries and benchmarks</p>
</li>
<li>
<p>bins (only if benchmark targets are built and required features are
available)</p>
</li>
<li>
<p>lib as a benchmark</p>
</li>
<li>
<p>bins as benchmarks</p>
</li>
<li>
<p>benchmark targets</p>
</li>
</ul>
</div>
<div class="paragraph">
<p>The default behavior can be changed by setting the <code>bench</code> flag for the target
in the manifest settings. Setting examples to <code>bench = true</code> will build and
run the example as a benchmark. Setting targets to <code>bench = false</code> will stop
them from being benchmarked by default. Target selection options that take a
target by name ignore the <code>bench</code> flag and will always benchmark the given
target.</p>
</div>
<div class="paragraph">
<p>Passing target selection flags will benchmark only the
specified targets.</p>
</div>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--lib</strong></dt>
<dd>
<p>Benchmark the package’s library.</p>
</dd>
<dt class="hdlist1"><strong>--bin</strong> <em>NAME</em>…​</dt>
<dd>
<p>Benchmark the specified binary. This flag may be specified multiple times.</p>
</dd>
<dt class="hdlist1"><strong>--bins</strong></dt>
<dd>
<p>Benchmark all binary targets.</p>
</dd>
<dt class="hdlist1"><strong>--example</strong> <em>NAME</em>…​</dt>
<dd>
<p>Benchmark the specified example. This flag may be specified multiple times.</p>
</dd>
<dt class="hdlist1"><strong>--examples</strong></dt>
<dd>
<p>Benchmark all example targets.</p>
</dd>
<dt class="hdlist1"><strong>--test</strong> <em>NAME</em>…​</dt>
<dd>
<p>Benchmark the specified integration test. This flag may be specified multiple
times.</p>
</dd>
<dt class="hdlist1"><strong>--tests</strong></dt>
<dd>
<p>Benchmark all targets in test mode that have the <code>test = true</code> manifest
flag set. By default this includes the library and binaries built as
unittests, and integration tests. Be aware that this will also build any
required dependencies, so the lib target may be built twice (once as a
unittest, and once as a dependency for binaries, integration tests, etc.).
Targets may be enabled or disabled by setting the <code>test</code> flag in the
manifest settings for the target.</p>
</dd>
<dt class="hdlist1"><strong>--bench</strong> <em>NAME</em>…​</dt>
<dd>
<p>Benchmark the specified benchmark. This flag may be specified multiple times.</p>
</dd>
<dt class="hdlist1"><strong>--benches</strong></dt>
<dd>
<p>Benchmark all targets in benchmark mode that have the <code>bench = true</code>
manifest flag set. By default this includes the library and binaries built
as benchmarks, and bench targets. Be aware that this will also build any
required dependencies, so the lib target may be built twice (once as a
benchmark, and once as a dependency for binaries, benchmarks, etc.).
Targets may be enabled or disabled by setting the <code>bench</code> flag in the
manifest settings for the target.</p>
</dd>
<dt class="hdlist1"><strong>--all-targets</strong></dt>
<dd>
<p>Benchmark all targets. This is equivalent to specifying <code>--lib --bins
--tests --benches --examples</code>.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_feature_selection">Feature Selection</h3>
<div class="paragraph">
<p>When no feature options are given, the <code>default</code> feature is activated for
every selected package.</p>
</div>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--features</strong> <em>FEATURES</em></dt>
<dd>
<p>Space or comma separated list of features to activate. These features only
apply to the current directory’s package. Features of direct dependencies
may be enabled with <code><dep-name>/<feature-name></code> syntax.</p>
</dd>
<dt class="hdlist1"><strong>--all-features</strong></dt>
<dd>
<p>Activate all available features of all selected packages.</p>
</dd>
<dt class="hdlist1"><strong>--no-default-features</strong></dt>
<dd>
<p>Do not activate the <code>default</code> feature of the current directory’s
package.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_compilation_options">Compilation Options</h3>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--target</strong> <em>TRIPLE</em></dt>
<dd>
<p>Benchmark for the given architecture. The default is the host
architecture. The general format of the triple is
<code><arch><sub>-<vendor>-<sys>-<abi></code>. Run <code>rustc --print target-list</code> for a
list of supported targets.</p>
<div class="paragraph">
<p>This may also be specified with the <code>build.target</code>
<a href="reference/config.html">config value</a>.</p>
</div>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_output_options">Output Options</h3>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--target-dir</strong> <em>DIRECTORY</em></dt>
<dd>
<p>Directory for all generated artifacts and intermediate files. May also be
specified with the <code>CARGO_TARGET_DIR</code> environment variable, or the
<code>build.target-dir</code> <a href="reference/config.html">config value</a>. Defaults
to <code>target</code> in the root of the workspace.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_display_options">Display Options</h3>
<div class="paragraph">
<p>By default the Rust test harness hides output from benchmark execution to keep
results readable. Benchmark output can be recovered (e.g., for debugging) by
passing <code>--nocapture</code> to the benchmark binaries:</p>
</div>
<div class="literalblock">
<div class="content">
<pre>cargo bench -- --nocapture</pre>
</div>
</div>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>-v</strong></dt>
<dt class="hdlist1"><strong>--verbose</strong></dt>
<dd>
<p>Use verbose output. May be specified twice for "very verbose" output which
includes extra output such as dependency warnings and build script output.
May also be specified with the <code>term.verbose</code>
<a href="reference/config.html">config value</a>.</p>
</dd>
<dt class="hdlist1"><strong>-q</strong></dt>
<dt class="hdlist1"><strong>--quiet</strong></dt>
<dd>
<p>No output printed to stdout.</p>
</dd>
<dt class="hdlist1"><strong>--color</strong> <em>WHEN</em></dt>
<dd>
<p>Control when colored output is used. Valid values:</p>
<div class="ulist">
<ul>
<li>
<p><code>auto</code> (default): Automatically detect if color support is available on the
terminal.</p>
</li>
<li>
<p><code>always</code>: Always display colors.</p>
</li>
<li>
<p><code>never</code>: Never display colors.</p>
</li>
</ul>
</div>
<div class="paragraph">
<p>May also be specified with the <code>term.color</code>
<a href="reference/config.html">config value</a>.</p>
</div>
</dd>
<dt class="hdlist1"><strong>--message-format</strong> <em>FMT</em></dt>
<dd>
<p>The output format for diagnostic messages. Valid values:</p>
<div class="ulist">
<ul>
<li>
<p><code>human</code> (default): Display in a human-readable text format.</p>
</li>
<li>
<p><code>json</code>: Emit JSON messages to stdout.</p>
</li>
<li>
<p><code>short</code>: Emit shorter, human-readable text messages.</p>
</li>
</ul>
</div>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_manifest_options">Manifest Options</h3>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>--manifest-path</strong> <em>PATH</em></dt>
<dd>
<p>Path to the <code>Cargo.toml</code> file. By default, Cargo searches in the current
directory or any parent directory for the <code>Cargo.toml</code> file.</p>
</dd>
<dt class="hdlist1"><strong>--frozen</strong></dt>
<dt class="hdlist1"><strong>--locked</strong></dt>
<dd>
<p>Either of these flags requires that the <code>Cargo.lock</code> file is
up-to-date. If the lock file is missing, or it needs to be updated, Cargo will
exit with an error. The <code>--frozen</code> flag also prevents Cargo from
attempting to access the network to determine if it is out-of-date.</p>
<div class="paragraph">
<p>These may be used in environments where you want to assert that the
<code>Cargo.lock</code> file is up-to-date (such as a CI build) or want to avoid network
access.</p>
</div>
</dd>
<dt class="hdlist1"><strong>--offline</strong></dt>
<dd>
<p>Prevents Cargo from accessing the network for any reason. Without this
flag, Cargo will stop with an error if it needs to access the network and
the network is not available. With this flag, Cargo will attempt to
proceed without the network if possible.</p>
<div class="paragraph">
<p>Beware that this may result in different dependency resolution than online
mode. Cargo will restrict itself to crates that are downloaded locally, even
if there might be a newer version as indicated in the local copy of the index.
See the <a href="commands/cargo-fetch.html">cargo-fetch(1)</a> command to download dependencies before going
offline.</p>
</div>
<div class="paragraph">
<p>May also be specified with the <code>net.offline</code> <a href="reference/config.html">config value</a>.</p>
</div>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_common_options">Common Options</h3>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>-h</strong></dt>
<dt class="hdlist1"><strong>--help</strong></dt>
<dd>
<p>Prints help information.</p>
</dd>
<dt class="hdlist1"><strong>-Z</strong> <em>FLAG</em>…​</dt>
<dd>
<p>Unstable (nightly-only) flags to Cargo. Run <code>cargo -Z help</code> for
details.</p>
</dd>
</dl>
</div>
</div>
<div class="sect2">
<h3 id="cargo_bench_miscellaneous_options">Miscellaneous Options</h3>
<div class="paragraph">
<p>The <code>--jobs</code> argument affects the building of the benchmark executable but
does not affect how many threads are used when running the benchmarks. The
Rust test harness runs benchmarks serially in a single thread.</p>
</div>
<div class="dlist">
<dl>
<dt class="hdlist1"><strong>-j</strong> <em>N</em></dt>
<dt class="hdlist1"><strong>--jobs</strong> <em>N</em></dt>
<dd>
<p>Number of parallel jobs to run. May also be specified with the
<code>build.jobs</code> <a href="reference/config.html">config value</a>. Defaults to
the number of CPUs.</p>
</dd>
</dl>
</div>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_profiles">PROFILES</h2>
<div class="sectionbody">
<div class="paragraph">
<p>Profiles may be used to configure compiler options such as optimization levels
and debug settings. See
<a href="reference/manifest.html#the-profile-sections">the reference</a>
for more details.</p>
</div>
<div class="paragraph">
<p>Benchmarks are always built with the <code>bench</code> profile. Binary and lib targets
are built separately as benchmarks with the <code>bench</code> profile. Library targets
are built with the <code>release</code> profiles when linked to binaries and benchmarks.
Dependencies use the <code>release</code> profile.</p>
</div>
<div class="paragraph">
<p>If you need a debug build of a benchmark, try building it with
<a href="commands/cargo-build.html">cargo-build(1)</a> which will use the <code>test</code> profile which is by default
unoptimized and includes debug information. You can then run the debug-enabled
benchmark manually.</p>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_environment">ENVIRONMENT</h2>
<div class="sectionbody">
<div class="paragraph">
<p>See <a href="reference/environment-variables.html">the reference</a> for
details on environment variables that Cargo reads.</p>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_exit_status">Exit Status</h2>
<div class="sectionbody">
<div class="dlist">
<dl>
<dt class="hdlist1">0</dt>
<dd>
<p>Cargo succeeded.</p>
</dd>
<dt class="hdlist1">101</dt>
<dd>
<p>Cargo failed to complete.</p>
</dd>
</dl>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_examples">EXAMPLES</h2>
<div class="sectionbody">
<div class="olist arabic">
<ol class="arabic">
<li>
<p>Build and execute all the benchmarks of the current package:</p>
<div class="literalblock">
<div class="content">
<pre>cargo bench</pre>
</div>
</div>
</li>
<li>
<p>Run only a specific benchmark within a specific benchmark target:</p>
<div class="literalblock">
<div class="content">
<pre>cargo bench --bench bench_name -- modname::some_benchmark</pre>
</div>
</div>
</li>
</ol>
</div>
</div>
</div>
<div class="sect1">
<h2 id="cargo_bench_see_also">SEE ALSO</h2>
<div class="sectionbody">
<div class="paragraph">
<p><a href="commands/index.html">cargo(1)</a>, <a href="commands/cargo-test.html">cargo-test(1)</a></p>
</div>
</div>
</div>