Run `iree-import-onnx` test(s) on CI

[ScottTodd]

https://github.com/openxla/iree/pull/16622 fixes an issue that should have been caught by presubmit tests.

We have a test but it has a soft failure mode if onnx is not installed: https://github.com/openxla/iree/blob/4c97ab6800820392589ce440f5b6b98f9bd855da/compiler/bindings/python/test/tools/import_onnx_test.py#L51-L56

I don't see a convenient way right now to either

• install that dep in the existing job that runs the test (build_all or a test job like test_gpu, complicated due to Docker)
• run the test in a job that can easily install it (pkgci, which doesn't use ctest)

This ONNX importer test/component should be low complexity and be unlikely to break (famous last words), so just putting this on the backlog to handle with future infrastructure work.

[ScottTodd]

We should also get https://github.com/iree-org/iree/blob/main/compiler/bindings/python/test/extras/fx_importer_test.py actually running by installing torch. That can be in a test workflow similar to https://github.com/iree-org/iree/blob/main/.github/workflows/pkgci_test_tensorflow_cpu.yml.

• pkgci_test_torch_cpu.yml
• merge into something like pkgci_test_frameworks.yml with multiple jobs/steps?
• pkgci_test_onnx.yml (or onnx under pkgci_test_frameworks.yml) could have https://github.com/iree-org/iree-test-suites/tree/main/onnx_ops and the iree-import-onnx tests