Run freeverb FFI benchmark in macOS CI

Add a step to the build-macos job that runs benchmark/run-freeverb.sh
(3 runs, 10000 iterations) after the LLVM build+test, and posts the
results to the GitHub job summary. This exercises lyte_compiler_compile
— the xcframework codepath the CLI never touches — on every push.

Also: set LIBRARY_PATH in run-freeverb.sh for CI link robustness, and
gitignore the built freeverb_bench binary.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>

Author: wtholliday

.github/workflows/rust.yml

@@ -78,6 +78,19 @@ jobs:
         LLVM_SYS_180_PREFIX: /opt/homebrew/opt/llvm@18
         LIBRARY_PATH: /opt/homebrew/opt/llvm@18/lib
       run: cargo test -p lyte -p lyte-cli --features llvm --verbose
+    - name: Run freeverb FFI benchmark
+      env:
+        LLVM_SYS_180_PREFIX: /opt/homebrew/opt/llvm@18
+        LIBRARY_PATH: /opt/homebrew/opt/llvm@18/lib
+      run: |
+        ./benchmark/run-freeverb.sh 3 10000 | tee freeverb_output.txt
+    - name: Post freeverb results to job summary
+      if: always()
+      run: |
+        echo '## Freeverb FFI Benchmark (macOS ARM64, LLVM)' >> $GITHUB_STEP_SUMMARY
+        echo '```' >> $GITHUB_STEP_SUMMARY
+        cat freeverb_output.txt >> $GITHUB_STEP_SUMMARY
+        echo '```' >> $GITHUB_STEP_SUMMARY
     - name: Install x86_64 Homebrew
       # The xcframework's macOS x86_64 slice is built with cross-compiled
       # LLVM, which requires a native x86_64 llvm-config + libs. Install

.gitignore

@@ -30,6 +30,7 @@ benchmark/fft_c_o2
 benchmark/fft_c_o3
 benchmark/sort_c_o2
 benchmark/sort_c_o3
+benchmark/freeverb_bench
 
 vscode-lyte/node_modules/
 

benchmark/run-freeverb.sh

@@ -32,6 +32,7 @@ if [ -z "$LLVM_PREFIX" ] || [ ! -d "$LLVM_PREFIX" ]; then
     exit 1
 fi
 export LLVM_SYS_180_PREFIX="$LLVM_PREFIX"
+export LIBRARY_PATH="${LIBRARY_PATH:+$LIBRARY_PATH:}$LLVM_PREFIX/lib"
 
 echo "Building liblyte.dylib (release, --features llvm)..."
 cargo build --release --features llvm --lib --quiet