Will Haltom

Faster next / finish by using `rb_profile_frames`

3

## Description - By using [rb_profile_frames](https://github.com/ruby/ruby/blob/d6f21b308bcff03e82f8b3dbf11a852ce111b3b3/include/ruby/debug.h#L51) instead of `rb_make_backtrace` in DEBUGGER__.frame_depth, we can improve the performance of frame_depth substantially for common cases. This leads to much faster `next` and `finish`...

Faster `next` by reducing number of frame_depth calls

5

## Description ### Background - `next` can be quite slow when stepping over code. It appears this is a known issue due to expensive calls to [DEBUGGER__.frame_depth](https://github.com/ruby/debug/blob/2cb44483c681f6806bf6843a8f996d3a10d0e42c/ext/debug/debug.c#L95) that are made...

Skip check for pending breakpoints if no breakpoints are present

4

## Description Describe your changes: - Calls to `ThreadClient.current.on_load` spend time pushing and popping the :load operation on the event queue, and waiting for the :load event to be handled....

I noticed some interesting behavior based on the following two things: - Calls to `Prelude.wrap` for an array of records store a reference to the preloader on each record: https://github.com/seejohnrun/prelude/blob/main/lib/prelude.rb#L9....