Add riscv64 (linux_riscv64) wheel to PyPI releases

Add linux_riscv64 wheel to PyPI releases

Summary

pip install uvloop on riscv64 Linux currently requires building from source (~5 min on a 1.6 GHz RISC-V SoC). Adding linux_riscv64 to the wheel build matrix would give riscv64 users a prebuilt wheel.

Evidence

The wheel imports and passes basic smoke tests on riscv64 Linux.

Suggested CI change

Workflow: .github/workflows/release.yml

Add riscv64 to the matrix:

# In strategy.matrix.cibw_arch, add riscv64:
cibw_arch: [x86_64, aarch64, universal2, riscv64]

# Add exclude for riscv64 on non-linux:
exclude:
  - os: macos-latest
    cibw_arch: riscv64

Add QEMU step and test skip:

- name: Set up QEMU
  if: matrix.cibw_arch == 'riscv64'
  uses: docker/setup-qemu-action@v3
  with:
    platforms: riscv64

# Add to env:
CIBW_TEST_SKIP: '*-linux_riscv64'

libuv has riscv64 support since v1.44. Event-loop tests may hang under QEMU, hence the test skip recommendation.

Ecosystem context

• manylinux_2_28_riscv64 images are available on quay.io/pypa/ (landed in pypa/manylinux, 2025)
• cibuildwheel 3.x supports riscv64 via QEMU emulation
• auditwheel supports riscv64 platform tags
• Packages already shipping riscv64 wheels on PyPI: aiohttp, yarl, regex, markupsafe, charset-normalizer, rpds-py, multidict, propcache, watchfiles, rignore, setproctitle
• RISC-V hardware is shipping: SiFive HiFive, SpacemiT K1/K3, Sophgo SG2044 (64-core)

Our temporary index

While upstream support is pending, we maintain a PEP 503 index with 50+ riscv64 wheels for the Python ML/AI stack, built natively on RISC-V hardware.

Tracking repo: https://github.com/gounthar/riscv64-python-wheels

Comienza con .github/workflows/release.yml e inspecciona la matriz cibw_arch, las exclusiones de plataforma y las pruebas de wheel existentes. Añade la configuración de QEMU específica para riscv64 y la entrada CIBW_TEST_SKIP descritas en el issue, y verifica después que el workflow de release produzca un wheel linux_riscv64 sin ejecutar pruebas del bucle de eventos bajo emulación.