bevy/docs/profiling.md
ira ef032040dd Cargo --timings option has been stabilized. Update profiling.md. (#4850)
As of https://github.com/rust-lang/cargo/pull/10245 `--timings` has been stabilized.
Update profiling.md to reflect this.

Co-authored-by: devil-ira <justthecooldude@gmail.com>
2022-05-30 21:16:48 +00:00

3.5 KiB

Profiling

Runtime Flame Graph: tracing spans

Bevy has built-in tracing spans to make it cheap and easy to profile Bevy ECS systems, render logic, engine iternals, and user app code. Enable the trace cargo feature to enable Bevy's built-in spans. You also need to select a tracing backend using the following cargo features:

Backend: trace_chrome

cargo run --release --features bevy/trace_chrome

After running your app a json file in the "chrome tracing format" will be produced. You can open this file in your browser using https://ui.perfetto.dev. It will look something like this (make sure you expand Process 1):

image

Adding your own spans

Add spans to your app like this (these are in bevy::prelude::* and bevy::log::*, just like the normal logging macros).

{
  // creates a span and starts the timer 
  let my_span = info_span!("span_name", name = "span_name").entered();
  do_something_here();
} // my_span is dropped here ... this stops the timer


// You can also "manually" enter the span if you need more control over when the timer starts
// Prefer the previous, simpler syntax unless you need the extra control.
let my_span = info_span!("span_name", name = "span_name");
{
  // starts the span's timer 
  let guard = my_span.enter();
  do_something_here();
} // guard is dropped here ... this stops the timer

Search for info_span! in this repo for some real-world examples.

For more details, check out the tracing span docs.

perf Runtime Flame Graph

This approach requires no extra instrumentation and shows finer-grained flame graphs of actual code call trees. This is useful when you want to identify the specific function of a "hot spot". The downside is that it has higher overhead, so your app will run slower than it normally does.

Install cargo-flamegraph, enable debug symbols in your release build, then run your app using one of the following commands. Note that cargo-flamegraph forwards arguments to cargo. You should treat the cargo-flamegraph command as a replacement for cargo run --release. The commands below include --example EXAMPLE_NAME to illustrate, but you can remove those arguments in favor of whatever you use to run your app:

  • Graph-Like Flame Graph: RUSTFLAGS='-C force-frame-pointers=y' cargo flamegraph -c "record -g" --example EXAMPLE_NAME
  • Flat-ish Flame Graph: RUSTFLAGS='-C force-frame-pointers=y' cargo flamegraph --example EXAMPLE_NAME

After closing your app, an interactive svg file will be produced: image

Project Compile Times

Append --timings to your app's cargo command (ex: cargo build --timings). If you want a "full" profile, make sure you run cargo clean first (note: this will clear previously generated reports). The command will tell you where it saved the report, which will be in your target directory under cargo-timings/. The report is a .html file and can be opened and viewed in your browser. This will show how much time each crate in your app's dependency tree took to build.

image