Save some size in dtype_util when dtype selective build is not in use

[swolchok]

We duplicate a lot of functions depending on the operator name so that dtype selective build will work. We can just detect if dtype selective build is in use and, if not, stop duplicating.

Test Plan: Saves 28288 bytes of text in size_test_all_optimized_ops compared to previous PR on my Mac.
Pull-Request-resolved: #9742

[swolchok]

Stack from ghstack (oldest at bottom):

• Add vectorization in elementwise_util #9432
• BroadcastIndexesRange: leading 1s don't require true broadcasting #9431
• -> Save some size in dtype_util when dtype selective build is not in use #9829
• Migrate elementwise_util callers to the variants with out_dtypes in template arguments #9828
• RFC: Specialize for non-mixed-dtype in elementwise_util #9388
• Refactor elementwise_util: create variants with out_dtypes in template argument list #9387
• Add build_optimized_size_test.sh #9827
• set INTERFACE_LINK_LIBARIES for extension_threadpool #9826
• Fix build gating on optimized_portable_kernels #9825
• Make PyTorch headers available in optimized_portable_kernels, define ET_USE_PYTORCH_HEADERS #9384

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/9829

• 📄 Preview Python docs built from this PR

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

✅ No Failures

As of commit ef74fe1 with merge base 5fcb6b8 ():
💚 Looks good so far! There are no failures yet. 💚

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

[swolchok]

ugh, ghstack seems to have duplicated a ton of PRs in this stack

[swolchok]

duplicate PR