Avoid one unnecessary memory allocation in XNNPACK integration.

[AshkanAliabadi]

Stack from ghstack:

• Add XNNPACK Clamp. #35896 [WIP] Add XNNPACK Clamp.
• Enable stateless XNNPACK linear. #35791 Enable stateless XNNPACK linear.
• Enable stateless XNNPACK convolutions. #35790 Enable stateless XNNPACK convolutions.
• Support for XNNPACK max pooling operator. #35354 Support for XNNPACK max pooling operator.
• Propagate input tensor names in XNNPACK backend. #35351 Propagate input tensor names in XNNPACK backend.
• Avoid one unnecessary memory allocation in XNNPACK integration. #35350 Avoid one unnecessary memory allocation in XNNPACK integration.

Currently we call input.contiguous() on the input tensor resulting in an
unecessary allocation and copy in cases where the input is not contiguous
with regards to the requested memory format. The reason is that in such
scenarios, this call re-allocates and copies the input tensor into
contiguous storage, only for this newly allocated tensor to be used as
the source of another copy to the final destination. Instead, if we copy
into the destination directly in such circumstances, we will save an
allocation and a copy.

Differential Revision: D20656798

[dr-ci]

💊 CircleCI build failures summary and remediations

As of commit ac58495 (more details on the Dr. CI page):

---

✅ None of the build failures appear to be your fault 💚

---

• 1/1 broken upstream at merge base 2a4ca70 since Apr 02

  Please rebase on the viable/strict branch (expand for instructions)

  If your commit is newer than viable/strict, you can try basing on an older, stable commit:

  git fetch https://github.com/pytorch/pytorch viable/strict
  git rebase --onto FETCH_HEAD $(git merge-base origin/master HEAD)
  

  If your commit is older than viable/strict:

  git fetch https://github.com/pytorch/pytorch viable/strict
  git rebase FETCH_HEAD
  

  Check out the recency history of this "viable master" tracking branch.

---

🚧 1 upstream failure:

These were probably caused by upstream breakages:

• binary_linux_libtorch_3_6m_cpu_gcc5_4_cxx11-abi_shared-with-deps_build since Apr 02

---

This comment was automatically generated by Dr. CI (expand for details).

Follow this link to opt-out of these comments for your Pull Requests.

Please report bugs/suggestions on the GitHub issue tracker.

See how this bot performed.

This comment has been revised 30 times.

[AshkanAliabadi]

Do we have any tests that I can use to validate that the results are correct? I'm not sure this code is exercised through CI.

[kimishpatel]

Do we have any tests that I can use to validate that the results are correct? I'm not sure this code is exercised through CI.

You can try python test/test_xnnpack_integration.py. I suggest add, in one or two tests, trying different memory formats for inputs, including channels first and channels last.

[kimishpatel]

LG for the most part. Can you add the test I mentioned in the comment or try the existing tests?

[facebook-github-bot]

@AshkanAliabadi merged this pull request in d0ce94d.