Help Request: Perf issue when using puffin

[siedentop]

Thank you for this awesome project!

I am trying to do some tracing measurements for [binocle][1]. However, when I use tracing with puffin (+puffin_http + viewer), there is a significant slowdown and most scopes do not show.

I prepared a branch which you can run: https://github.com/siedentop/binocle/tree/feature/eframe

cargo run --example create_test_file
cargo run --features trace -- binocle_test_file

Could you please help me and tell me what I might have done wrong?

I know that the Binocle::draw method is using most of the time. (And yes, some work should move to a different thread, but I'd like some measurements first.) However, it only gets called if there is a RedrawRequest event, so it doesn't show in the trace viewer.

With puffin::GlobalProfiler::lock().new_frame(); in the main loop:

And now with puffin::GlobalProfiler::lock().new_frame(); just inside the if let Event::RedrawRequested(_) = event:

Any ideas? Many thanks (also feel free to close, if this isn't the right venue).

[emilk]

If you call puffin::GlobalProfiler::lock().new_frame(); right inside event_loop.run you will be calling it for every new event that is received, which will mean each event will be treated as a new "frame", which is likely not what you want.

Calling it on RedrawRequested seems far more sensible.

---

As noted in the readme, each puffin profile scope costs 100-200 ns, so creating one for each pixel is a bad idea. Puffin is meant for more coarse-grained profiling.

[siedentop]

Thank you @emilk! This fixes it.

Originally, I had a trace point inside color_at_index method, which was called for every pixel! But I didn't realize it.

Great framework, and I already found one more performance issue 😊