perf(build): parallelize cython extension compilation

bdraco · bdraco · commit dbba4cfa3194 · 2026-05-16T14:11:30.000-07:00
Default `build_ext`'s `parallel` to `os.cpu_count()` so the per-module
`gcc` compile step runs in parallel. Wheel builds for emulated archs
(riscv64 + armv7l via QEMU) currently spend ~16 of their ~20 minutes
serializing 18 `gcc -O3` invocations; on 4-core runners this should drop
the compile phase to ~4-5 minutes.
diff --git a/build_ext.py b/build_ext.py
@@ -46,6 +46,8 @@
 
 class BuildExt(build_ext):
     def build_extensions(self) -> None:
+        if self.parallel is None:  # type: ignore[has-type]
+            self.parallel = os.cpu_count() or 1
         try:
             super().build_extensions()
         except Exception:
