Remove usage of environment variables in wheel building

[jathu]

Summary

We seem to be using a combination of CMAKE_ARGS and environment variables when creating wheels. Ultimately, CMake only uses the cmake args, however we redefine some of these flags as env vars to help setup.py determine if a certain feature is turned on. Specifically, it looks for pybinding vars to bundle pybindings.

Let's remove this redundancy and just use the CMAKE_ARGS as the single source of truth. For more details and other considerations, see #9494 (abandoned).

Note that even in the wheel building jobs, we use cmake args instead of environment variables to control features:

executorch/.ci/scripts/wheel/envvar_base.sh

Line 20 in 644b7dd

CMAKE_ARGS="${CMAKE_ARGS} -DEXECUTORCH_BUILD_XNNPACK=ON"

executorch/.ci/scripts/wheel/envvar_macos.sh

Lines 14 to 15 in 644b7dd

	CMAKE_ARGS="${CMAKE_ARGS} -DEXECUTORCH_BUILD_COREML=ON"
	CMAKE_ARGS="${CMAKE_ARGS} -DEXECUTORCH_BUILD_MPS=ON"

Test plan

build + check CMakeCache.txt to ensure flags are set

# Expected: EXECUTORCH_BUILD_PYBIND=OFF EXECUTORCH_BUILD_XNNPACK=OFF EXECUTORCH_BUILD_COREML=OFF
$ rm -rf pip-out dist && ./install_executorch.sh --pybind off

# Expected: EXECUTORCH_BUILD_PYBIND=ON EXECUTORCH_BUILD_XNNPACK=ON EXECUTORCH_BUILD_COREML=OFF
$ rm -rf pip-out dist && ./install_executorch.sh

# Expected: EXECUTORCH_BUILD_PYBIND=ON EXECUTORCH_BUILD_XNNPACK=OFF EXECUTORCH_BUILD_COREML=ON
$ rm -rf pip-out dist && ./install_executorch.sh --pybind coreml

# Expected: EXECUTORCH_BUILD_PYBIND=ON EXECUTORCH_BUILD_XNNPACK=ON EXECUTORCH_BUILD_COREML=ON
$ rm -rf pip-out dist && ./install_executorch.sh --pybind xnnpack coreml

# Throws an error
$ rm -rf pip-out dist && ./install_executorch.sh --pybind coreml off

cc @larryliu0820 @lucylq

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/9583

• 📄 Preview Python docs built from this PR

Note: Links to docs will display an error until the docs builds have been completed.

❌ 2 New Failures, 2 Pending, 2 Unrelated Failures

As of commit 69a5ba4 with merge base 7248b19 ():

NEW FAILURES - The following jobs have failed:

• trunk / test-models-linux-aarch64 (emformer_join, portable) / linux-job (gh)
  RuntimeError: Command docker exec -t 04ff1e7c12ece577269c842d8224174415abd9b732a627fbfc7be329c4a74af6 /exec failed with exit code 1
• trunk / test-models-macos (llama2, xnnpack-quantization-delegation) / macos-job (gh)
  RuntimeError: torch.histc: range of [nan, nan] is not finite

BROKEN TRUNK - The following jobs failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / test-llava-runner-linux / linux-job (gh) (trunk failure)
  test_generated_output
• pull / unittest-editable / macos / macos-job (gh) (trunk failure)
  backends/xnnpack/test/ops/test_check_quant_params.py::TestCheckQuantParams::test_inject_invalid_zp

This comment was automatically generated by Dr. CI and updates every 15 minutes.

[mergennachin]

See inline comment