Jan 7, 2025

Yeah, pref shorter. The tool is called tracy, maybe --tracer. I'm not sure what works yet, but might be nice to pull and build it and just have it integrated into vscode ux.

maybe --with-tracy just to follow our dependency pattern? wouldn't really matter but then in the code the flag would be pretty clear and consistent

yeah i could add a build option, that would work

in fact, like the optimizations, adding at system would propagate everywhere

It's not a dependency though, it's a tool that's applied at the test or console project (I think). The VS profiling tools don't require any source instrumentation, but I think Tracy does require a symbol defined in the sources.

We should prob validate the tool before integrating.

involves conditional includes

requires zones to be defined

"Add the macro ZoneScoped as the first line of your function definitions to include them in the profile"

doesn't give anything for free as far as i can tell

Oooo, that's nuts.

I read on it and I thought that they had said the instrumentation was one include and that's it.

woulda been nice if that were the case

Adding a macro to every function is a ridiculous way to discover performance hotspots.

In the VS profiler it doesn't require a debug build, we can attach to the running process remotely, and there is no source instrumentation.

Not sure how we'd do this.

sounds like its just gdb then - haven't used that in ages

lolz

I don't think that's capable of doing what we need.

There's a good tool integrated with xcode. Otherwise this one is the only other I found outside of VS.

Might need a tool to generate the macro insertions.

It's hard to imagine however that this won't affect performance.

sed

The tool to build the tool.

So I was just gonna go with this quickstart

Then refine from there

I watched a bit of the video linked on their github and the guy does a whole walkthrough on it

It looked like if you did it this "quick" way it would give you info about everything

which would probably be a firehose

https://libbitcoin.slack.com/archives/C6LH74ACQ/p1736282118292719?thread_ts=1736278617.939899&cid=C6LH74ACQ it definitely affects performance, but i think the idea is we wouldn't build it in unless the --with-tracy option were included it claims to be "low overhead" fwiw, but yeah this is a debugging tool

As long as the effect on performance is proportional it's fine.

Above still shows explicitly instrumenting each individual function.

Seems like that's where the work is going to be.

Presumably I can just throw it in something super high-level like main() or something? Haha

I'm guessing that will give main: 100%

I may be mistaken. Hopefully

haha guess we'll see