[WIP] Migrate put_() from TH to ATen (CPU)

[leonvol]

Move CPU implementation for put_() from TH to ATen. Fixes #24751

[leonvol]

Does anyone have some advice, if I can take a similar approach to take_out_cpu_template() to use at::parallel_for as the main loop in the migrated function? @VitalyFedyunin

[facebook-github-bot]

💊 CI failures summary and remediations

As of commit e6f0ad5 (more details on the Dr. CI page):

---

• 1/1 failures possibly* introduced in this PR
  • 1/1 non-CircleCI failure(s)

---

ci.pytorch.org: 1 failed

• Failed: pr/pytorch-linux-bionic-rocm3.10-py3.6

---

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 to the (internal) Dr. CI Users group.

[nikitaved]

Hey, @leonvol ! You can restride your input to be efficiently used with TensorIterator. That way you will parallelize over self.numel() elements, which will eliminate the need for the serial for-loop over indices. You could have a look at #50578.

[lezcano]

Hi @leonvol

Is this PR still active? If not, I would not mind taking over it! :)

[leonvol]

Hi @lezcano , feel free to take this PR over. I am currently quite busy :)

[lezcano]

Thanks! I will do so in the next few days :)

[lezcano]

This port was done in #53356