cargo bench

NAME

cargo-bench - Execute benchmarks of a package

SYNOPSIS

cargo bench [OPTIONS] [BENCHNAME] [— BENCH-OPTIONS]

DESCRIPTION

Compile and execute benchmarks.

The benchmark filtering argument BENCHNAME and all the arguments followingthe two dashes (—) are passed to the benchmark binaries and thus tolibtest (rustc’s built in unit-test and micro-benchmarking framework). Ifyou’re passing arguments to both Cargo and the binary, the ones after — goto the binary, the ones before go to Cargo. For details about libtest’sarguments see the output of cargo bench — —help. As an example, this willrun only the benchmark named foo (and skip other similarly named benchmarkslike foobar):

  1. cargo bench -- foo --exact

Benchmarks are built with the —test option to rustc which creates anexecutable with a main function that automatically runs all functionsannotated with the #[bench] attribute. Cargo passes the —bench flag tothe test harness to tell it to run only benchmarks.

The libtest harness may be disabled by setting harness = false in the targetmanifest settings, in which case your code will need to provide its own mainfunction to handle running benchmarks.

OPTIONS

Benchmark Options

  • —no-run

  • Compile, but don’t run benchmarks.

  • —no-fail-fast

  • Run all benchmarks regardless of failure. Without this flag, Cargo will exitafter the first executable fails. The Rust test harness will run allbenchmarks within the executable to completion, this flag only applies tothe executable as a whole.

Package Selection

By default, when no package selection options are given, the packages selecteddepend on the selected manifest file (based on the current working directory if—manifest-path is not given). If the manifest is the root of a workspace thenthe workspaces default members are selected, otherwise only the package definedby the manifest will be selected.

The default members of a workspace can be set explicitly with theworkspace.default-members key in the root manifest. If this is not set, avirtual workspace will include all workspace members (equivalent to passing—workspace), and a non-virtual workspace will include only the root crate itself.

  • -pSPEC…​
  • —packageSPEC…​

  • Benchmark only the specified packages. See cargo-pkgid(1) for theSPEC format. This flag may be specified multiple times.

  • —workspace

  • Benchmark all members in the workspace.

  • —all

  • Deprecated alias for —workspace.

  • —excludeSPEC…​

  • Exclude the specified packages. Must be used in conjunction with the—workspace flag. This flag may be specified multiple times.

Target Selection

When no target selection options are given, cargo bench will build thefollowing targets of the selected packages:

  • lib — used to link with binaries and benchmarks

  • bins (only if benchmark targets are built and required features areavailable)

  • lib as a benchmark

  • bins as benchmarks

  • benchmark targets

The default behavior can be changed by setting the bench flag for the targetin the manifest settings. Setting examples to bench = true will build andrun the example as a benchmark. Setting targets to bench = false will stopthem from being benchmarked by default. Target selection options that take atarget by name ignore the bench flag and will always benchmark the giventarget.

Passing target selection flags will benchmark only thespecified targets.

  • —lib

  • Benchmark the package’s library.

  • —binNAME…​

  • Benchmark the specified binary. This flag may be specified multiple times.

  • —bins

  • Benchmark all binary targets.

  • —exampleNAME…​

  • Benchmark the specified example. This flag may be specified multiple times.

  • —examples

  • Benchmark all example targets.

  • —testNAME…​

  • Benchmark the specified integration test. This flag may be specified multipletimes.

  • —tests

  • Benchmark all targets in test mode that have the test = true manifestflag set. By default this includes the library and binaries built asunittests, and integration tests. Be aware that this will also build anyrequired dependencies, so the lib target may be built twice (once as aunittest, and once as a dependency for binaries, integration tests, etc.).Targets may be enabled or disabled by setting the test flag in themanifest settings for the target.

  • —benchNAME…​

  • Benchmark the specified benchmark. This flag may be specified multiple times.

  • —benches

  • Benchmark all targets in benchmark mode that have the bench = truemanifest flag set. By default this includes the library and binaries builtas benchmarks, and bench targets. Be aware that this will also build anyrequired dependencies, so the lib target may be built twice (once as abenchmark, and once as a dependency for binaries, benchmarks, etc.).Targets may be enabled or disabled by setting the bench flag in themanifest settings for the target.

  • —all-targets

  • Benchmark all targets. This is equivalent to specifying —lib —bins—tests —benches —examples.

Feature Selection

When no feature options are given, the default feature is activated forevery selected package.

  • —featuresFEATURES

  • Space or comma separated list of features to activate. These features onlyapply to the current directory’s package. Features of direct dependenciesmay be enabled with <dep-name>/<feature-name> syntax.

  • —all-features

  • Activate all available features of all selected packages.

  • —no-default-features

  • Do not activate the default feature of the current directory’spackage.

Compilation Options

  • —targetTRIPLE
  • Benchmark for the given architecture. The default is the hostarchitecture. The general format of the triple is<arch><sub>-<vendor>-<sys>-<abi>. Run rustc —print target-list for alist of supported targets.

This may also be specified with the build.targetconfig value.

Output Options

  • —target-dirDIRECTORY
  • Directory for all generated artifacts and intermediate files. May also bespecified with the CARGO_TARGET_DIR environment variable, or thebuild.target-dir config value. Defaultsto target in the root of the workspace.

Display Options

By default the Rust test harness hides output from benchmark execution to keepresults readable. Benchmark output can be recovered (e.g., for debugging) bypassing —nocapture to the benchmark binaries:

  1. cargo bench -- --nocapture
  • -v
  • —verbose

  • Use verbose output. May be specified twice for "very verbose" output whichincludes extra output such as dependency warnings and build script output.May also be specified with the term.verboseconfig value.

  • -q

  • —quiet

  • No output printed to stdout.

  • —colorWHEN

  • Control when colored output is used. Valid values:

  • auto (default): Automatically detect if color support is available on theterminal.

  • always: Always display colors.

  • never: Never display colors.

May also be specified with the term.colorconfig value.

  • —message-formatFMT
  • The output format for diagnostic messages. Can be specified multiple timesand consists of comma-separated values. Valid values:

  • human (default): Display in a human-readable text format.

  • short: Emit shorter, human-readable text messages.

  • json: Emit JSON messages to stdout.

  • json-diagnostic-short: Ensure the rendered field of JSON messages containsthe "short" rendering from rustc.

  • json-diagnostic-rendered-ansi: Ensure the rendered field of JSON messagescontains embedded ANSI color codes for respecting rustc’s default colorscheme.

  • json-render-diagnostics: Instruct Cargo to not include rustc diagnostics inin JSON messages printed, but instead Cargo itself should render theJSON diagnostics coming from rustc. Cargo’s own JSON diagnostics and otherscoming from rustc are still emitted.

Manifest Options

  • —manifest-pathPATH

  • Path to the Cargo.toml file. By default, Cargo searches in the currentdirectory or any parent directory for the Cargo.toml file.

  • —frozen

  • —locked
  • Either of these flags requires that the Cargo.lock file isup-to-date. If the lock file is missing, or it needs to be updated, Cargo willexit with an error. The —frozen flag also prevents Cargo fromattempting to access the network to determine if it is out-of-date.

These may be used in environments where you want to assert that theCargo.lock file is up-to-date (such as a CI build) or want to avoid networkaccess.

  • —offline
  • Prevents Cargo from accessing the network for any reason. Without thisflag, Cargo will stop with an error if it needs to access the network andthe network is not available. With this flag, Cargo will attempt toproceed without the network if possible.

Beware that this may result in different dependency resolution than onlinemode. Cargo will restrict itself to crates that are downloaded locally, evenif there might be a newer version as indicated in the local copy of the index.See the cargo-fetch(1) command to download dependencies before goingoffline.

May also be specified with the net.offline config value.

Common Options

  • -h
  • —help

  • Prints help information.

  • -ZFLAG…​

  • Unstable (nightly-only) flags to Cargo. Run cargo -Z help fordetails.

Miscellaneous Options

The —jobs argument affects the building of the benchmark executable butdoes not affect how many threads are used when running the benchmarks. TheRust test harness runs benchmarks serially in a single thread.

  • -jN
  • —jobsN
  • Number of parallel jobs to run. May also be specified with thebuild.jobs config value. Defaults tothe number of CPUs.

PROFILES

Profiles may be used to configure compiler options such as optimization levelsand debug settings. Seethe referencefor more details.

Benchmarks are always built with the bench profile. Binary and lib targetsare built separately as benchmarks with the bench profile. Library targetsare built with the release profiles when linked to binaries and benchmarks.Dependencies use the release profile.

If you need a debug build of a benchmark, try building it withcargo-build(1) which will use the test profile which is by defaultunoptimized and includes debug information. You can then run the debug-enabledbenchmark manually.

ENVIRONMENT

See the reference fordetails on environment variables that Cargo reads.

Exit Status

  • 0

  • Cargo succeeded.

  • 101

  • Cargo failed to complete.

EXAMPLES

  • Build and execute all the benchmarks of the current package:
  1. cargo bench
  • Run only a specific benchmark within a specific benchmark target:
  1. cargo bench --bench bench_name -- modname::some_benchmark

SEE ALSO

cargo(1), cargo-test(1)