Fix build gating on optimized_portable_kernels

[swolchok]

If we don't need optimized kernels, we don't need optimized portable kernels.

[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 #9842
• Migrate elementwise_util callers to the variants with out_dtypes in template arguments #9841
• 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 #9840
• set INTERFACE_LINK_LIBARIES for extension_threadpool #9839
• -> Fix build gating on optimized_portable_kernels #9838
• 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/9665

• 📄 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 34eb5d4 with merge base 65ebabb ():
💚 Looks good so far! There are no failures yet. 💚

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

[swolchok]

replaced by #9838 . not sure what's going on with ghstack :\