Arm backend: Add validation for same dtype to operators

[Sebastian-Larsson]

When applicable, check that the data types of inputs to a given operator are the same.

cc @digantdesai @freddan80 @per @zingo @oscarandersson8218

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/10872

• 📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

• Large queue time for macos-m2-15 instances

❌ 2 New Failures, 2 Pending, 1 Unrelated Failure

As of commit 00d8784 with merge base c913634 ():

NEW FAILURES - The following jobs have failed:

• Lint / lintrunner / linux-job (gh)
  >>> Lint for examples/models/llama/export_llama_lib.py:
• pull / unittest-editable / linux / linux-job (gh)
  extension/llm/modules/test/test_position_embeddings.py::TiledTokenPositionalEmbeddingTest::test_tiled_token_positional_embedding_smoke

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / test-llava-runner-linux / linux-job (gh) (trunk failure)
  test_llava_export

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

[zingo]

lintrunner llama and llava fails unrelated,