[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. #65549

ZolotukhinM · 2021-09-23T18:00:43Z

Stack from ghstack:

Previously it had a special handling, with this change it follows the
same mechanism as other ops.

Differential Revision: D31148924

Previously it had a special handling, with this change it follows the same mechanism as other ops. [ghstack-poisoned]

facebook-github-bot · 2021-09-23T18:00:47Z

🔗 Helpful links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/65549
📄 Preview docs built from this PR
📄 Preview C++ docs built from this PR
↩️ [fb-only] Re-run with SSH instructions
🔧 Opt-in to CIFlow to control what jobs run on your PRs

💊 CI failures summary and remediations

As of commit 599606c (more details on the Dr. CI page):

2/4 failures introduced in this PR
2/4 tentatively recognized as flaky ❄️
- Click here to rerun these jobs

🕵️ 2 new failures recognized by patterns

The following CI failures do not appear to be due to upstream breakages:

linux-xenial-py3.6-gcc5.4 / test (default, 2, 2, linux.2xlarge) (1/2)

Step: "Unknown" (full log | diagnosis details | 🔁 rerun)

2021-09-30T18:44:50.5135713Z test_add_done_ca...arg() takes 0 positional arguments but 1 was given

2021-09-30T18:44:50.5117643Z   /opt/conda/lib/python3.6/unittest/suite.py(122): run
2021-09-30T18:44:50.5118137Z   /opt/conda/lib/python3.6/unittest/suite.py(84): __call__
2021-09-30T18:44:50.5118839Z   /opt/conda/lib/python3.6/site-packages/xmlrunner/runner.py(66): run
2021-09-30T18:44:50.5119426Z   /opt/conda/lib/python3.6/unittest/main.py(256): runTests
2021-09-30T18:44:50.5119957Z   /opt/conda/lib/python3.6/unittest/main.py(95): __init__
2021-09-30T18:44:50.5120882Z   /opt/conda/lib/python3.6/site-packages/torch/testing/_internal/common_utils.py(605): run_tests
2021-09-30T18:44:50.5121477Z   test_futures.py(329): <module>
2021-09-30T18:44:50.5121712Z 
2021-09-30T18:44:50.5121967Z ok (0.002s)
2021-09-30T18:44:50.5130670Z   test_add_done_callback_maintains_callback_order (__main__.TestFuture) ... ok (0.002s)
2021-09-30T18:44:50.5135713Z   test_add_done_callback_no_arg_error_is_ignored (__main__.TestFuture) ... [E pybind_utils.h:201] Got the following error when running the callback: TypeError: no_arg() takes 0 positional arguments but 1 was given
2021-09-30T18:44:50.5136545Z ok (0.001s)
2021-09-30T18:44:50.5144782Z   test_add_done_callback_simple (__main__.TestFuture) ... ok (0.001s)
2021-09-30T18:44:50.5171198Z   test_chained_then (__main__.TestFuture) ... ok (0.003s)
2021-09-30T18:44:50.6190849Z   test_collect_all (__main__.TestFuture) ... ok (0.102s)
2021-09-30T18:44:50.6197654Z   test_done (__main__.TestFuture) ... ok (0.001s)
2021-09-30T18:44:50.6208527Z   test_done_exception (__main__.TestFuture) ... ok (0.001s)
2021-09-30T18:44:50.6222440Z   test_interleaving_then_and_add_done_callback_maintains_callback_order (__main__.TestFuture) ... ok (0.001s)
2021-09-30T18:44:50.6230910Z   test_interleaving_then_and_add_done_callback_propagates_error (__main__.TestFuture) ... [E pybind_utils.h:201] Got the following error when running the callback: ValueError: Expected error
2021-09-30T18:44:50.6232161Z 
2021-09-30T18:44:50.6232561Z At:

Lint / clang-tidy (2/2)

Step: "Check for warnings" (full log | diagnosis details | 🔁 rerun)

2021-09-30T17:52:45.1883762Z /__w/pytorch/pytor...necessary-copy-initialization,-warnings-as-errors]

2021-09-30T17:52:44.9752696Z �[36;1mcd "${GITHUB_WORKSPACE}"�[0m
2021-09-30T17:52:44.9753028Z �[36;1mset -eu�[0m
2021-09-30T17:52:44.9753483Z �[36;1mcat "${GITHUB_WORKSPACE}"/clang-tidy-output.txt�[0m
2021-09-30T17:52:44.9754152Z �[36;1mif grep -Fq "Warnings detected!" "${GITHUB_WORKSPACE}"/clang-tidy-output.txt; then�[0m
2021-09-30T17:52:44.9754800Z �[36;1m  echo 'Please fix the above clang-tidy warnings.'�[0m
2021-09-30T17:52:44.9755185Z �[36;1m  false�[0m
2021-09-30T17:52:44.9755452Z �[36;1mfi�[0m
2021-09-30T17:52:44.9755862Z shell: sh -e {0}
2021-09-30T17:52:44.9756149Z ##[endgroup]
2021-09-30T17:52:45.1881110Z Processing 1 clang-tidy jobs
2021-09-30T17:52:45.1883762Z /__w/pytorch/pytorch/torch/csrc/jit/tensorexpr/kernel.cpp:1970:18: error: the variable 'b' is copy-constructed from a const reference but is only used as const reference; consider making it a const reference [performance-unnecessary-copy-initialization,-warnings-as-errors]
2021-09-30T17:52:45.1885660Z             auto b = c10::get<BufHandle>(inputs[0]);
2021-09-30T17:52:45.1886104Z                  ^
2021-09-30T17:52:45.1886394Z             const  &
2021-09-30T17:52:45.1887046Z Warnings detected!
2021-09-30T17:52:45.1887375Z Summary:
2021-09-30T17:52:45.1888649Z [performance-unnecessary-copy-initialization] occurred 1 times
2021-09-30T17:52:45.1889777Z     /__w/pytorch/pytorch/torch/csrc/jit/tensorexpr/kernel.cpp:1970
2021-09-30T17:52:45.1890176Z 
2021-09-30T17:52:45.1890886Z Please fix the above clang-tidy warnings.
2021-09-30T17:52:45.1922951Z ##[error]Process completed with exit code 1.

❄️ 2 failures tentatively classified as flaky

but reruns have not yet been triggered to confirm:

linux-xenial-cuda11.3-py3.6-gcc7 / test (default, 2, 2, linux.8xlarge.nvidia.gpu) (1/2)

Step: "Test" (full log | diagnosis details | 🔁 rerun) ❄️

2021-09-30T21:16:45.2015507Z unknown file: Failure

2021-09-30T21:16:45.1993010Z �[0;32m[       OK ] �[mNVFuserTest.FusionTVSplit_CUDA (0 ms)
2021-09-30T21:16:45.1994009Z �[0;32m[ RUN      ] �[mNVFuserTest.FusionTVMerge_CUDA
2021-09-30T21:16:45.1995009Z �[0;32m[       OK ] �[mNVFuserTest.FusionTVMerge_CUDA (0 ms)
2021-09-30T21:16:45.1996039Z �[0;32m[ RUN      ] �[mNVFuserTest.FusionTVReorder_CUDA
2021-09-30T21:16:45.1997092Z �[0;32m[       OK ] �[mNVFuserTest.FusionTVReorder_CUDA (0 ms)
2021-09-30T21:16:45.1998134Z �[0;32m[ RUN      ] �[mNVFuserTest.FusionEquality_CUDA
2021-09-30T21:16:45.1999165Z �[0;32m[       OK ] �[mNVFuserTest.FusionEquality_CUDA (0 ms)
2021-09-30T21:16:45.2000223Z �[0;32m[ RUN      ] �[mNVFuserTest.FusionDependency_CUDA
2021-09-30T21:16:45.2001327Z �[0;32m[       OK ] �[mNVFuserTest.FusionDependency_CUDA (0 ms)
2021-09-30T21:16:45.2002342Z �[0;32m[ RUN      ] �[mNVFuserTest.FusionParser_CUDA
2021-09-30T21:16:45.2015507Z unknown file: Failure
2021-09-30T21:16:45.2017330Z C++ exception with description "Couldn't find an operator for aten::_softmax_backward_data(Tensor grad_output, Tensor output, int dim, Tensor self) -> Tensor. Do you have to update a set of hardcoded JIT ops?
2021-09-30T21:16:45.2018909Z Exception raised from lookupByLiteral at /var/lib/jenkins/workspace/torch/csrc/jit/runtime/operator.cpp:141 (most recent call first):
2021-09-30T21:16:45.2020883Z frame #0: c10::Error::Error(c10::SourceLocation, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> >) + 0x6b (0x7fecacb2acbb in /opt/conda/lib/python3.6/site-packages/torch/bin/libc10.so)
2021-09-30T21:16:45.2023118Z frame #1: c10::detail::torchCheckFail(char const*, char const*, unsigned int, std::__cxx11::basic_string<char, std::char_traits<char>, std::allocator<char> > const&) + 0xce (0x7fecacb2690e in /opt/conda/lib/python3.6/site-packages/torch/bin/libc10.so)
2021-09-30T21:16:45.2025315Z frame #2: torch::jit::getOperatorForLiteral(char const*) + 0x1a0c (0x7fecb0dbe34c in /opt/conda/lib/python3.6/site-packages/torch/bin/libtorch_cpu.so)
2021-09-30T21:16:45.2026928Z frame #3: <unknown function> + 0xd8156e (0x7fec9e35456e in /opt/conda/lib/python3.6/site-packages/torch/bin/libtorch_cuda_cu.so)
2021-09-30T21:16:45.2028983Z frame #4: torch::jit::fuser::cuda::parseJitIR(std::shared_ptr<torch::jit::Graph> const&) + 0x7a5 (0x7fec9e3577b5 in /opt/conda/lib/python3.6/site-packages/torch/bin/libtorch_cuda_cu.so)
2021-09-30T21:16:45.2030761Z frame #5: torch::jit::NVFuserTest_FusionParser_CUDA_Test::TestBody() + 0x531 (0x6f7ef1 in /opt/conda/lib/python3.6/site-packages/torch/bin/test_jit)
2021-09-30T21:16:45.2032997Z frame #6: void testing::internal::HandleExceptionsInMethodIfSupported<testing::Test, void>(testing::Test*, void (testing::Test::*)(), char const*) + 0x4a (0x805f5a in /opt/conda/lib/python3.6/site-packages/torch/bin/test_jit)
2021-09-30T21:16:45.2034725Z frame #7: /opt/conda/lib/python3.6/site-packages/torch/bin/test_jit() [0x7f5a60]

pytorch_linux_xenial_py3_6_gcc5_4_test (2/2)

Step: "Check for no AVX instruction by default" (full log | diagnosis details | 🔁 rerun) ❄️

E: Failed to fetch https://deb.nodesource.com/n...: /etc/ssl/certs/ca-certificates.crt CRLfile: none

Ign:14 https://deb.nodesource.com/node_12.x xenial/main amd64 Packages
Ign:13 https://deb.nodesource.com/node_12.x xenial/main all Packages
Err:10 https://deb.nodesource.com/node_12.x xenial/main Sources
  server certificate verification failed. CAfile: /etc/ssl/certs/ca-certificates.crt CRLfile: none
Ign:14 https://deb.nodesource.com/node_12.x xenial/main amd64 Packages
Ign:13 https://deb.nodesource.com/node_12.x xenial/main all Packages
Get:16 http://archive.ubuntu.com/ubuntu xenial-updates/universe amd64 Packages [1544 kB]
Fetched 4466 kB in 29s (152 kB/s)
Reading package lists...
W: The repository 'https://deb.nodesource.com/node_12.x xenial Release' does not have a Release file.
E: Failed to fetch https://deb.nodesource.com/node_12.x/dists/xenial/main/source/Sources  server certificate verification failed. CAfile: /etc/ssl/certs/ca-certificates.crt CRLfile: none
E: Some index files failed to download. They have been ignored, or old ones used instead.


Exited with code exit status 100

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.

Click here to manually regenerate this comment.

ZolotukhinM · 2021-09-23T18:02:27Z

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

navahgar

LGTM

navahgar · 2021-09-24T22:55:50Z

torch/csrc/jit/tensorexpr/kernel.cpp

+  if (op == prim::ConstantChunk) {
+    auto const& n = v->node();
+    argInputs.push_back(toArg(inputs[0]));
+    argInputs.push_back((int64_t)v->offset());


"static_cast<int64_t>" instead of the traditional cast.

… other ops." Previously it had a special handling, with this change it follows the same mechanism as other ops. Differential Revision: [D31148924](https://our.internmc.facebook.com/intern/diff/D31148924) [ghstack-poisoned]

ZolotukhinM · 2021-09-28T23:03:03Z

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

… other ops." Previously it had a special handling, with this change it follows the same mechanism as other ops. Differential Revision: [D31148924](https://our.internmc.facebook.com/intern/diff/D31148924) [ghstack-poisoned]

ZolotukhinM · 2021-09-29T19:48:03Z

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

ZolotukhinM · 2021-09-29T21:13:36Z

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

… other ops." Previously it had a special handling, with this change it follows the same mechanism as other ops. Differential Revision: [D31148924](https://our.internmc.facebook.com/intern/diff/D31148924) [ghstack-poisoned]

ZolotukhinM · 2021-09-29T22:40:02Z

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

… other ops." Previously it had a special handling, with this change it follows the same mechanism as other ops. Differential Revision: [D31148924](https://our.internmc.facebook.com/intern/diff/D31148924) [ghstack-poisoned]

ZolotukhinM · 2021-09-30T04:40:37Z

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

… other ops." Previously it had a special handling, with this change it follows the same mechanism as other ops. Differential Revision: [D31148924](https://our.internmc.facebook.com/intern/diff/D31148924) [ghstack-poisoned]

ZolotukhinM · 2021-09-30T17:31:15Z

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

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops.

670cd5f

Previously it had a special handling, with this change it follows the same mechanism as other ops. [ghstack-poisoned]

facebook-github-bot added the oncall: jit Add this issue/PR to JIT oncall triage queue label Sep 23, 2021

facebook-github-bot added the cla signed label Sep 23, 2021

navahgar approved these changes Sep 24, 2021

View reviewed changes

ZolotukhinM mentioned this pull request Sep 29, 2021

[TensorExpr] Use schema instead of op name in NNC lowerings. #65843

Closed

ZolotukhinM mentioned this pull request Sep 29, 2021

[TensorExpr] Fix lowerings for aten::view and aten::reshape. #65852

Closed

ZolotukhinM mentioned this pull request Sep 29, 2021

[TensorExpr] Fix lowering for aten::t. #65859

Closed

facebook-github-bot closed this in d84191f Oct 1, 2021

facebook-github-bot deleted the gh/ZolotukhinM/458/head branch October 4, 2021 14:24

Provide feedback

Saved searches

Use saved searches to filter your results more quickly

Uh oh!

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. #65549

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. #65549

Uh oh!

ZolotukhinM commented Sep 23, 2021 •

edited

Loading

Uh oh!

facebook-github-bot commented Sep 23, 2021 •

edited

Loading

Uh oh!

ZolotukhinM commented Sep 23, 2021

Uh oh!

navahgar left a comment

Uh oh!

navahgar Sep 24, 2021

Uh oh!

ZolotukhinM commented Sep 28, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 30, 2021

Uh oh!

ZolotukhinM commented Sep 30, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. #65549

[TensorExpr] Kernel: make prim::ConstantChunk handled like other ops. #65549

Uh oh!

Conversation

ZolotukhinM commented Sep 23, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

Uh oh!

facebook-github-bot commented Sep 23, 2021 • edited Loading Uh oh! There was an error while loading. Please reload this page.

Uh oh!

🔗 Helpful links

💊 CI failures summary and remediations

🕵️ 2 new failures recognized by patterns

linux-xenial-py3.6-gcc5.4 / test (default, 2, 2, linux.2xlarge) (1/2)

Lint / clang-tidy (2/2)

❄️ 2 failures tentatively classified as flaky

linux-xenial-cuda11.3-py3.6-gcc7 / test (default, 2, 2, linux.8xlarge.nvidia.gpu) (1/2)

pytorch_linux_xenial_py3_6_gcc5_4_test (2/2)

Uh oh!

ZolotukhinM commented Sep 23, 2021

Uh oh!

navahgar left a comment

Choose a reason for hiding this comment

Uh oh!

navahgar Sep 24, 2021

Choose a reason for hiding this comment

Uh oh!

ZolotukhinM commented Sep 28, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 29, 2021

Uh oh!

ZolotukhinM commented Sep 30, 2021

Uh oh!

ZolotukhinM commented Sep 30, 2021

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

3 participants

ZolotukhinM commented Sep 23, 2021 •

edited

Loading

facebook-github-bot commented Sep 23, 2021 •

edited

Loading