aot

AOT compiler benchmarks

These benchmarks measure generated Glojure code without including process startup, source parsing, or namespace loading in the timed region. They are intended to expose compiler optimizations that the portable end-to-end benchmarks cannot isolate.

The runner creates a temporary Go module, copies the fixtures, generates fresh AOT loaders with the current compiler, and runs Go benchmarks against the loaded run Vars:

go run ./benchmark/aot -count 5 -benchtime 2s

Use -glojure /path/to/glj to reuse an existing bootstrap compiler executable. By default, the runner builds one from the current checkout with glj_no_aot_stdlib. Use -keep to retain the temporary module for inspection.

The fixtures deliberately cover optimization gaps:

• constant-arithmetic.glj: repeated arithmetic containing a constant subexpression, through a direct function call.
• float-kernel.glj: repeated floating-point arithmetic through a direct function call.
• letgo-map-filter.glj: let-go's exact map/filter/take benchmark, including its anonymous integer-square callback.
• reduce-pipeline.glj: reduction over lazy map and filter producers.

Run a baseline before changing the compiler and compare it with the same command afterward. The reported allocation counts are particularly useful for checking unboxing and sequence fusion.

For startup-inclusive, cross-runtime workloads designed to run for roughly one second or longer in at least one runtime, use the long-running AOT suite. It builds the same Clojure fixtures as native Glojure and let-go applications and compares their median wall times.