replace onlyOnCPUAndCUDA with onlyNativeDeviceTypes by kshitij12345 · Pull Request #65201 · pytorch/pytorch

kshitij12345 · 2021-09-17T07:19:23Z

Reference #53849

Replace onlyOnCPUandCUDA with onlyNativeDeviceTypes which includes cpu, cuda and meta.

facebook-github-bot · 2021-09-17T07:19:29Z

🔗 Helpful links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/65201
📄 Preview docs built from this PR
📄 Preview C++ docs built from this PR
🔧 Opt-in to CIFlow to control what jobs run on your PRs

💊 CI failures summary and remediations

As of commit d0ee5b1 (more details on the Dr. CI page):

💚 💚 Looks good so far! There are no failures yet. 💚 💚

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

Please report bugs/suggestions to the (internal) Dr. CI Users group.

Click here to manually regenerate this comment.

codecov · 2021-09-17T10:33:21Z

Codecov Report

Merging #65201 (c00beb9) into master (feefc94) will increase coverage by 0.02%.
The diff coverage is 66.66%.

❗ Current head c00beb9 differs from pull request most recent head d0ee5b1. Consider uploading reports for the commit d0ee5b1 to get more accurate results

@@            Coverage Diff             @@
##           master   #65201      +/-   ##
==========================================
+ Coverage   66.37%   66.40%   +0.02%     
==========================================
  Files         739      735       -4     
  Lines       94299    94051     -248     
==========================================
- Hits        62595    62451     -144     
+ Misses      31704    31600     -104

…eTypes

kshitij12345 · 2021-09-24T08:04:06Z

test/test_binary_ufuncs.py


    # Verifies that the inplace dunders (like idiv) actually are in place
-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # UserWarning not triggered


Note the expected failure

kshitij12345 · 2021-09-24T08:04:38Z

test/test_nn.py



-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # Runtime Error not raised for meta


Note the expected failure

kshitij12345 · 2021-09-24T08:04:50Z

test/test_nn.py

            self.assertEqual(sample, torch.zeros([1, 1, 1, 2], device=device, dtype=dtype))

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # RuntimeError: Unrecognized tensor type ID: Meta


Note the expected failure

kshitij12345 · 2021-09-24T08:04:56Z

test/test_nn.py

                F.fractional_max_pool2d(x, (2, 2), output_size=output_size, _random_samples=samples)

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # RuntimeError: Unrecognized tensor type ID: Meta


Note the expected failure

kshitij12345 · 2021-09-24T08:05:44Z

test/test_ops.py


    # Validates that each OpInfo specifies its forward and backward dtypes
    #   correctly for CPU and CUDA devices
+    @skipMeta


Note the skip (some tests pass and some tests fail). Hence the skip

kshitij12345 · 2021-09-24T08:08:43Z

test/test_torch.py

            torch.bernoulli(torch.rand_like(x), out=x)

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # RuntimeError not raised


Note expectedFailureMeta

kshitij12345 · 2021-09-24T08:08:49Z

test/test_torch.py

            ind.put_(ind.clone(), ind)

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # UserWarning not triggered


Note expectedFailureMeta

kshitij12345 · 2021-09-24T08:08:53Z

test/test_torch.py

            ind.index_put_((ind.clone(),), ind)

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # UserWarning not triggered


Note expectedFailureMeta

kshitij12345 · 2021-09-24T08:08:58Z

test/test_torch.py

            torch.masked_select(mask.clone(), mask, out=mask)

-    @onlyOnCPUAndCUDA
+    @expectedFailureMeta  # RuntimeError not raised


Note expectedFailureMeta

kshitij12345 · 2021-09-24T08:09:38Z

torch/testing/_internal/common_device_type.py


-# Only runs the test on the CPU and CUDA (the native device types)
-def onlyOnCPUAndCUDA(fn):
+# Only runs the test on the native device type (currently CPU, CUDA, Meta)


Main update

Thanks for the review pointers ;)

mruberry

Thanks for being so thorough and adding the review comments, @kshitij12345!

This looks good to me (and it makes this decorator consistent with onlyCPU and onlyCUDA by removing the "On"). I'll make an internal FB announcement, too.

cc @ezyang -- is this what you were expecting?

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

The ASAN timeout is unrelated but this will need a rebease.

facebook-github-bot · 2021-09-30T06:34:04Z

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…eTypes

ezyang · 2021-09-30T17:04:46Z

cc @ezyang -- is this what you were expecting?

Err, I don't know what I was expecting! Seems... reasonable, I guess?

zou3519 · 2021-09-30T17:14:40Z

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

Is there a replacement we could use?

mruberry · 2021-09-30T22:52:56Z

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

Is there a replacement we could use?

The onlyNativeDeviceTypes decorator introduced with this PR.

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

Is there a replacement we could use?

@zou3519 Yes we can find a replacement -- but would you prefer to run meta, cpu, and cuda variants of your tests or maintain just running the cpu and cuda variants?

…eTypes

zou3519 · 2021-10-18T15:27:03Z

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

Is there a replacement we could use?

The onlyNativeDeviceTypes decorator introduced with this PR.

fyi @zou3519 this will break functorch, which is importing the onlyOnCPUAndCUDA decorator

Is there a replacement we could use?

@zou3519 Yes we can find a replacement -- but would you prefer to run meta, cpu, and cuda variants of your tests or maintain just running the cpu and cuda variants?

Sorry for the delayed reply. I excised all the usages of onlyOnCPUAndCUDA in functorch (it turns out functorch wasn't actually using this decorator).

mruberry · 2021-10-21T06:56:02Z

Thanks for the update, @zou3519. @kshitij12345, this just needs a rebase and then ping me and I'll merge it!

…eTypes

pytorch-probot · 2021-10-21T07:01:42Z

CI Flow Status

⚛️ CI Flow

Ruleset - Version: v1
Ruleset - File: https://github.com/kshitij12345/pytorch/blob/d0ee5b1e2958b997c6257f0f4e775f171444c77b/.github/generated-ciflow-ruleset.json
PR ciflow labels: ciflow/default

Workflows	Labels (bold enabled)	Status
Triggered Workflows
linux-bionic-py3.6-clang9	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/noarch`, `ciflow/xla`	✅ triggered
linux-vulkan-bionic-py3.6-clang9	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/vulkan`	✅ triggered
linux-xenial-cuda11.3-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/default`, `ciflow/linux`	✅ triggered
linux-xenial-py3-clang5-mobile-build	`ciflow/all`, `ciflow/default`, `ciflow/linux`, `ciflow/mobile`	✅ triggered
linux-xenial-py3-clang5-mobile-custom-build-dynamic	`ciflow/all`, `ciflow/default`, `ciflow/linux`, `ciflow/mobile`	✅ triggered
linux-xenial-py3-clang5-mobile-custom-build-static	`ciflow/all`, `ciflow/default`, `ciflow/linux`, `ciflow/mobile`	✅ triggered
linux-xenial-py3.6-clang7-asan	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/sanitizers`	✅ triggered
linux-xenial-py3.6-clang7-onnx	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`, `ciflow/onnx`	✅ triggered
linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`	✅ triggered
linux-xenial-py3.6-gcc7	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`	✅ triggered
linux-xenial-py3.6-gcc7-bazel-test	`ciflow/all`, `ciflow/bazel`, `ciflow/cpu`, `ciflow/default`, `ciflow/linux`	✅ triggered
win-vs2019-cpu-py3	`ciflow/all`, `ciflow/cpu`, `ciflow/default`, `ciflow/win`	✅ triggered
win-vs2019-cuda11.3-py3	`ciflow/all`, `ciflow/cuda`, `ciflow/default`, `ciflow/win`	✅ triggered
Skipped Workflows
caffe2-linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/linux`	🚫 skipped
docker-builds	`ciflow/all`	🚫 skipped
libtorch-linux-xenial-cuda10.2-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`	🚫 skipped
libtorch-linux-xenial-cuda11.3-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`	🚫 skipped
linux-bionic-cuda10.2-py3.9-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/slow`	🚫 skipped
linux-xenial-py3-clang5-mobile-code-analysis	`ciflow/all`, `ciflow/linux`, `ciflow/mobile`	🚫 skipped
parallelnative-linux-xenial-py3.6-gcc5.4	`ciflow/all`, `ciflow/cpu`, `ciflow/linux`	🚫 skipped
periodic-libtorch-linux-xenial-cuda11.1-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/libtorch`, `ciflow/linux`, `ciflow/scheduled`	🚫 skipped
periodic-linux-xenial-cuda10.2-py3-gcc7-slow-gradcheck	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/scheduled`, `ciflow/slow`, `ciflow/slow-gradcheck`	🚫 skipped
periodic-linux-xenial-cuda11.1-py3.6-gcc7	`ciflow/all`, `ciflow/cuda`, `ciflow/linux`, `ciflow/scheduled`	🚫 skipped
periodic-win-vs2019-cuda11.1-py3	`ciflow/all`, `ciflow/cuda`, `ciflow/scheduled`, `ciflow/win`	🚫 skipped

You can add a comment to the PR and tag @pytorchbot with the following commands:

# ciflow rerun, "ciflow/default" will always be added automatically
@pytorchbot ciflow rerun

# ciflow rerun with additional labels "-l <ciflow/label_name>", which is equivalent to adding these labels manually and trigger the rerun
@pytorchbot ciflow rerun -l ciflow/scheduled -l ciflow/slow

For more information, please take a look at the CI Flow Wiki.

…eTypes

kshitij12345 · 2021-10-22T07:36:56Z

@mruberry this should be ready once the rocm build finishes. (Also maybe we should run it on ci-all workflow?)

kshitij12345 · 2021-10-26T08:50:00Z

@mruberry gentle ping :)

…eTypes

facebook-github-bot · 2021-10-30T11:04:27Z

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

…eTypes

facebook-github-bot · 2021-11-01T07:51:24Z

@mruberry has imported this pull request. If you are a Facebook employee, you can view this diff on Phabricator.

Summary: Reference #53849 Replace `onlyOnCPUandCUDA` with `onlyNativeDeviceTypes` which includes `cpu, cuda and meta`. Pull Request resolved: #65201 Reviewed By: mrshenli Differential Revision: D31299718 Pulled By: mruberry fbshipit-source-id: 2d8356450c035d6a314209ab51b2c237583920fd

replace onlyOnCPUAndCUDA with onlyNativeDeviceTypes

2f90f3c

facebook-github-bot added the cla signed label Sep 17, 2021

pytorchbot added the open source label Sep 17, 2021

kshitij12345 added 20 commits September 17, 2021 16:10

temporarily add skip for meta raising warning

95b4bb2

warning in meta

ecf55f0

add comment

8a96632

remove stray file

5cda140

update test_nn with skips

c00beb9

update expected failure for meta

740eb91

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

7b073f2

…eTypes

make flake8 happy

232053f

undo floor changes

b99239b

skip failing reductions tests

06f37c0

add missing import

14f617a

update test shape ops

5fa7abb

remove unused import

734b50a

update test_tensor_creation

04d6afe

update test_testing

862384c

update test_torch

9c70831

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

40155fe

…eTypes

update test with expectedFailures

fe7006d

update comments

56f955c

update binary_ufuncs with expectedFailure

fd8650a

kshitij12345 commented Sep 24, 2021

View reviewed changes

kshitij12345 marked this pull request as ready for review September 24, 2021 09:45

kshitij12345 requested review from IvanYashchuk, lezcano and nikitaved as code owners September 24, 2021 09:45

mruberry approved these changes Sep 30, 2021

View reviewed changes

kshitij12345 added 2 commits September 30, 2021 06:56

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

c0e1459

…eTypes

update comment

1ff0ff7

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

c38c6d7

…eTypes

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

dafa823

…eTypes

pytorch-probot bot added the ciflow/default label Oct 21, 2021

kshitij12345 added 4 commits October 21, 2021 08:20

update new occurences

767ba7c

update errors

7f79182

add expectedFailureMeta for ctor_device_inference test

adada58

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

577c4cd

…eTypes

mruberry mentioned this pull request Oct 29, 2021

Check if XLA-specific failures marked with onlyOnCPUAndCUDA in TestNNDeviceType can be skipped on the XLA side #67568

Closed

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

5a44508

…eTypes

kshitij12345 added 2 commits October 31, 2021 06:57

Merge branch 'master' into replace-onlyOnCPUCUDA-with-onlyNativeDevic…

9410504

…eTypes

update the decorator usage

d0ee5b1

kshitij12345 mentioned this pull request Nov 1, 2021

[testing] gradient OpInfo tests fail in Python 3.10 #67612

Closed

facebook-github-bot closed this Nov 1, 2021

malfet mentioned this pull request Feb 10, 2025

Open up PT UTs to cover additional devices #145589

Closed



		@onlyOnCPUAndCUDA
		@expectedFailureMeta # Runtime Error not raised for meta

Conversation

kshitij12345 commented Sep 17, 2021

Uh oh!

facebook-github-bot commented Sep 17, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful links

💊 CI failures summary and remediations

Uh oh!

codecov bot commented Sep 17, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Codecov Report

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

Choose a reason for hiding this comment

Uh oh!

mruberry left a comment

Choose a reason for hiding this comment

Uh oh!

facebook-github-bot commented Sep 30, 2021

Uh oh!

ezyang commented Sep 30, 2021

Uh oh!

zou3519 commented Sep 30, 2021

Uh oh!

mruberry commented Sep 30, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

zou3519 commented Oct 18, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

mruberry commented Oct 21, 2021

Uh oh!

pytorch-probot bot commented Oct 21, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

⚛️ CI Flow

Uh oh!

kshitij12345 commented Oct 22, 2021

Uh oh!

kshitij12345 commented Oct 26, 2021

Uh oh!

facebook-github-bot commented Oct 30, 2021

Uh oh!

facebook-github-bot commented Nov 1, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

6 participants

facebook-github-bot commented Sep 17, 2021 •

edited

Loading

codecov bot commented Sep 17, 2021 •

edited

Loading

mruberry commented Sep 30, 2021 •

edited

Loading

zou3519 commented Oct 18, 2021 •

edited

Loading

pytorch-probot bot commented Oct 21, 2021 •

edited

Loading