move generated-autograd-headers to the shared build structure

Pull Request resolved: https://github.com/pytorch/pytorch/pull/76183

This is a relatively simple target but we have to fix our header
expansion to understand generated files. Next step will be to use this
in Bazel.

Differential Revision: [D35820541](https://our.internmc.facebook.com/intern/diff/D35820541/)

**NOTE FOR REVIEWERS**: This PR has internal Facebook specific changes or comments, please review them on [Phabricator](https://our.internmc.facebook.com/intern/diff/D35820541/)!

Approved by: https://github.com/dreiss, https://github.com/malfet
This commit is contained in:
Michael Andreas Dagitses
2022-05-18 11:28:43 -07:00
committed by PyTorch MergeBot
parent 0c91efb64e
commit c2ff413622

View File

@ -44,6 +44,12 @@ def define_targets(rules):
"--gen_lazy_ts_backend",
)
rules.cc_library(
name = "generated-autograd-headers",
hdrs = [":{}".format(h) for h in _GENERATED_AUTOGRAD_CPP_HEADERS + _GENERATED_AUTOGRAD_PYTHON_HEADERS],
visibility = ["//visibility:public"],
)
rules.genrule(
name = "version_h",
srcs = [
@ -76,8 +82,6 @@ _GENERATED_AUTOGRAD_CPP_HEADERS = [
"torch/csrc/autograd/generated/variable_factories.h",
]
GENERATED_AUTOGRAD_H = _GENERATED_AUTOGRAD_CPP_HEADERS + _GENERATED_AUTOGRAD_PYTHON_HEADERS
GENERATED_TESTING_PY = [
"torch/testing/_internal/generated/annotated_fn_args.py",
]