Fixed issue with memory format inconsistency for interpolate op

[vfdev-5]

Related to #68430

Description:

• Output's memory format is not respected in interpolate op when input is squeezed/unsqueezed channels last

For example:

import torch
print(torch.__version__)


def t1(x, mode, aa):
    do_squeeze = False
    if x.dim() == 3:
        do_squeeze = True
        x = x[None, ...]
    out = torch.nn.functional.interpolate(
        x, size=224, mode=mode, antialias=aa
    )
    if do_squeeze:
        out = out[0, ...]
    return out


aa = True
mode = "bilinear"
dtype = torch.uint8
mf = torch.channels_last
device = "cpu"

print("-", aa, mode, mf, device, dtype)
x = torch.randint(0, 256, size=(1, 3, 256, 256), dtype=dtype, device=device).contiguous(memory_format=mf)

x = x[0, ...]
x = x[None, ...]

y = t1(x, mode, aa)

input_mem_format = "CL" if x.is_contiguous(memory_format=torch.channels_last) else "CF"
if input_mem_format == "CF":
    assert x.is_contiguous(memory_format=torch.contiguous_format)

output_mem_format = "CL" if y.is_contiguous(memory_format=torch.channels_last) else "CF"
if output_mem_format == "CF":
    assert y.is_contiguous(memory_format=torch.contiguous_format)

if input_mem_format != output_mem_format:
    print(f"2 {mf}, {device}, {dtype}: {output_mem_format} != {input_mem_format}\n")

We can see have the following output on pytorch nightly:

2 torch.channels_last, cpu, torch.uint8: CF != CL

In details, memory format inconsistency comes from the strides:

x = torch.ones(1, 3, 256, 256).contiguous(memory_format=torch.channels_last)
x = x[0, ...].unsqueeze(0)
assert x.stride() == (3, 1, 768, 3)  # instead of (196608, 1, 768, 3)

The problem essentially comes from x.unsqueeze(0) where new stride (i.e. stride(0)) is computed as for a tensor with channels first memory format:

pytorch/aten/src/ATen/native/TensorShape.cpp

Line 3136 in 4582ceb

int64_t new_stride = dim >= tensor.dim() ? 1 : result.sizes[dim] * result.strides[dim];

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/100373

• 📄 Preview Python docs built from this PR
• 📄 Preview C++ docs built from this PR
• ❓ Need help or want to give feedback on the CI? Visit the bot commands wiki or our office hours

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

❌ 11 New Failures, 5 Pending

As of commit c469c14:

NEW FAILURES - The following jobs have failed:

• linux-bionic-py3.11-clang9 / test (crossref, 2, 2, linux.2xlarge) (gh)
• linux-bionic-py3.11-clang9 / test (default, 3, 3, linux.2xlarge) (gh)
• linux-bionic-py3.11-clang9 / test (dynamo, 1, 2, linux.2xlarge) (gh)
• linux-bionic-py3.8-clang9 / test (crossref, 1, 2, linux.2xlarge) (gh)
• linux-bionic-py3.8-clang9 / test (default, 1, 3, linux.2xlarge) (gh)
• linux-bionic-py3.8-clang9 / test (dynamo, 2, 2, linux.2xlarge) (gh)
• linux-focal-py3.8-gcc7 / test (default, 1, 3, linux.2xlarge) (gh)
• linux-focal-py3.9-clang7-asan / test (default, 1, 6, linux.4xlarge) (gh)
• linux-focal-py3.9-clang7-asan / test (default, 6, 6, linux.4xlarge) (gh)
• macos-12-py3-arm64 / test (default, 3, 3, macos-m1-12) (gh)
• win-vs2019-cpu-py3 / test (default, 3, 3, windows.4xlarge.nonephemeral) (gh)

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

[vfdev-5]

Removed this heuristics due to failing test_fake_tensor.py tests and also https://github.com/pytorch/pytorch/pull/91260/files#r1183144329
cc @jbschlosser

[ezyang]

@jbschlosser holler if you are not the right person

[github-actions]

Looks like this PR hasn't been updated in a while so we're going to go ahead and mark this as Stale.
Feel free to remove the Stale label if you feel this was a mistake.
If you are unable to remove the Stale label please contact a maintainer in order to do so.
If you want the bot to never mark this PR stale again, add the no-stale label.
Stale pull requests will automatically be closed after 30 days of inactivity.

[ezyang]

@pytorchbot merge -r

[github-actions]

This PR needs a release notes: label

If your changes are user facing and intended to be a part of release notes, please use a label starting with release notes:.

If not, please add the topic: not user facing label.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "topic: not user facing"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

[pytorchmergebot]

@pytorchbot started a rebase job onto refs/remotes/origin/viable/strict. Check the current status here

[pytorchmergebot]

Successfully rebased fix-interp-output-memformat-1CHW-CL onto refs/remotes/origin/viable/strict, please pull locally before adding more changes (for example, via git checkout fix-interp-output-memformat-1CHW-CL && git pull --rebase)

[pytorchmergebot]

Merge failed

Reason: This PR needs a release notes: label
If your changes are user facing and intended to be a part of release notes, please use a label starting with release notes:.

If not, please add the topic: not user facing label.

To add a label, you can comment to pytorchbot, for example
@pytorchbot label "topic: not user facing"

For more information, see
https://github.com/pytorch/pytorch/wiki/PyTorch-AutoLabel-Bot#why-categorize-for-release-notes-and-how-does-it-work.

Details for Dev Infra team

Raised by workflow job

[vfdev-5]

@ezyang thanks for the review and the approval. Right now I set this PR to draft as there are failing tests and we also plan to make more tests from torchvision-side