Adopt runtime::FunctionRef in thread_parallel_interface.h and thread_parallel.h

[swolchok]

This saves a ton of size by no longer requiring callers to create std::function.

Test Plan: bash test/build_optimized_size_test.sh && size test/size_test_all_optimized_ops

before:

ExecuTorch with no ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  153928 Apr 24 13:11 cmake-out/test/size_test
ExecuTorch with portable ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  2150960 Apr 24 13:11 cmake-out/test/size_test_all_ops
ExecuTorch with optimized ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  11182136 Apr 24 13:11 cmake-out/test/size_test_all_optimized_ops

__TEXT	__DATA	__OBJC	others	dec	hex
5341184	98304	0	4300800000	4306239488	100ac0000

after:

ExecuTorch with no ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  153928 Apr 24 13:07 cmake-out/test/size_test
ExecuTorch with portable ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  2150960 Apr 24 13:07 cmake-out/test/size_test_all_ops
ExecuTorch with optimized ops binary size, unstripped:
-rwxr-xr-x  1 swolchok  staff  5927336 Apr 24 13:07 cmake-out/test/size_test_all_optimized_ops

__TEXT	__DATA	__OBJC	others	dec	hex
4505600	98304	0	4296376320	4300980224	1005bc000

__TEXT size improvement is 835584 bytes, or 15%.

[swolchok]

Stack from ghstack (oldest at bottom):

• Reapply #9841: Migrate elementwise_util callers to the variants with out_dtypes in template arguments #10491
• Reapply #9842: Save some size in dtype_util when dtype selective build is not in use #10490
• Save some size in pattern/{bitwise,comparison}_op.h #10489
• -> Adopt runtime::FunctionRef in thread_parallel_interface.h and thread_parallel.h #10442
• Move fully-featured FunctionRef from extension/pytree to ExecuTorch core #10441

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10442

• 📄 Preview Python docs built from this PR

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

⏳ No Failures, 4 Pending

As of commit 5bc4b0c with merge base 6feb623 ():
💚 Looks good so far! There are no failures yet. 💚

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

[digantdesai]

curious how did you find this optimization?

[swolchok]

curious how did you find this optimization?

@digantdesai I was working on reapplying #9842 and #9841 (reverted due to size regression) and iteratively using bloaty on the result of bash test/build_optimized_size_test.sh. There was a clear huge size increase from std::function, and after struggling with how to avoid it, I found my way to function_ref.

[swolchok]

Unit tests look good despite two apparent flakes. will merge after rebase.