You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Updates NIGHTLY_VERSION in torch_pin.py to dev20260809
Updates .ci/docker/ci_commit_pins/pytorch.txt to the corresponding nightly commit hash
Syncs c10 headers from PyTorch into runtime/core/portable_type/c10/
This PR was created automatically. If CI fails, Claude will attempt to fix issues (up to 3 attempts). If CI still fails, human review will be requested.
pull / test-qnn-delegate-linux / linux-job (gh) /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/torch/csrc/api/include/torch/all.h:5:2: error: C++20 or later compatible compiler is required to use PyTorch.
pull / test-qnn-passes-linux / linux-job (gh) /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/ATen/core/TensorBase.h:1034:40: error: expected ';' at end of declaration
pull / unittest-editable / linux / linux-job (gh) /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/ATen/ATen.h:5:2: error: C++20 or later compatible compiler is required to use ATen.
pull / unittest-nxp-neutron / linux-job (gh) /opt/conda/envs/py_3.10/lib/python3.10/site-packages/torch/include/ATen/ATen.h:5:2: error: C++20 or later compatible compiler is required to use ATen.
UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:
trunk / test-custom-ops-macos (cmake) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing) RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
trunk / test-models-macos-cpu (llama, portable) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing) RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
trunk / test-models-macos-cpu (mv3, portable) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing) RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
trunk / test-selective-build-macos (cmake) / macos-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing) RuntimeError: Command bash /Users/ec2-user/runner/_work/_temp/exec_script failed with exit code 1
This comment was automatically generated by Dr. CI and updates every 15 minutes.
meta-claBot
added
the
CLA Signed
This label is managed by the Facebook bot. Authors need to sign the CLA before a PR can be reviewed.
label
Aug 10, 2026
If your change should be included in the release notes (i.e. would users of this library care about this change?), please use a label starting with release notes:. This helps us keep track and include your important work in the next release notes.
To add a label, you can comment to pytorchbot, for example @pytorchbot label "release notes: none"
Read the Dr. CI comment on this PR for a summary of which jobs failed and whether they are flaky. Ignore failures marked as FLAKY.
Use your CI tools to download the failure logs for the non-flaky failing jobs
Identify the root cause of the failure
If this is a build or test failure caused by PyTorch API changes, fix the ExecuTorch code to be compatible with the new PyTorch version
If this is a c10 header sync issue, the headers have already been synced by the pin bump script — the issue is likely in ExecuTorch code that uses those headers
Run lintrunner -a on any files you change
Push your fix as a new commit to this PR branch
Important constraints:
Do NOT modify torch_pin.py or .ci/docker/ci_commit_pins/pytorch.txt — the pin itself is correct
Do NOT modify files under runtime/core/portable_type/c10/ unless the sync introduced a new API that ExecuTorch code needs to adapt to
Focus on fixing ExecuTorch code to be compatible with the new PyTorch APIs
If this is a major BC-breaking change that requires architectural discussion, say so clearly and stop — do not attempt a fix
CI Fix — PyTorch pin bump dev20260809 [attempt 1/3]
Read Dr. CI summary & identify non-flaky failures
Download failure logs & find root cause
Identify fix scope
Fix ExecuTorch code — stopping: requires architectural decision (see below)
Root cause
The new nightly (dev20260809) raises PyTorch's minimum C++ standard from C++17 to C++20. Every target that includes ATen/torch headers now fails at preprocessing:
torch/include/ATen/ATen.h:5:2: error: C++20 or later compatible compiler is required to use ATen.
torch/include/torch/csrc/api/include/torch/all.h:5:2: error: C++20 or later compatible compiler is required to use PyTorch.
The scattered TensorBase.h:1034: expected ';' errors are the same cause — a C++20-only construct in a header being compiled as C++17.
This is not a c10-header-sync issue and not a narrow API change. It's a compiler-standard floor bump on PyTorch's side.
Why I'm not auto-fixing this
The failure spans essentially the entire matrix (pybind/editable installs, QNN, Arm, Vulkan, WebGPU, XNNPACK, Cortex-M, buck2, macOS). ExecuTorch still pins C++17 as the default in many independent build surfaces:
CMakeLists.txt:116-118 (set(CMAKE_CXX_STANDARD 17)) and ~20 other CMakeLists.txt files that default to 17
.buckconfig:26 (-std=c++17) and third-party/gtest_defs.bzl:5
Bumping ExecuTorch's minimum C++ standard to 20 is a BC-breaking, cross-cutting decision. It affects downstream consumers who set CMAKE_CXX_STANDARD=17, and — more importantly — the embedded/bare-metal targets (Cortex-M, Zephyr, Ethos-U), where moving to C++20 has real toolchain and code-size implications that shouldn't be decided by an automated pin-bump. A migration was already started (portable_lib, extension/llm/runner, extension/training, wheel config, and Windows CI already use 20), but completing it project-wide — especially for embedded — needs human/architectural sign-off.
Per the constraint "If this is a major BC-breaking change that requires architectural discussion, say so clearly and stop", I'm stopping here rather than pushing a sweeping standard bump.
Recommended path (for a human)
Decide whether to raise ExecuTorch's global minimum to C++20 in lockstep with PyTorch. If yes, the change is mechanical but broad: flip the default 17→20 in the CMake files, .buckconfig, gtest_defs.bzl, .lintrunner.toml, and evaluate each embedded toolchain individually. cc @jakeszwe
· branch automated/pytorch-pin-bump-dev20260809
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
GitHub Actions
Inactive
meta-cla
Bot
added
the
CLA Signed
Summary
Automated weekly PyTorch pin bump.
NIGHTLY_VERSIONintorch_pin.pytodev20260809.ci/docker/ci_commit_pins/pytorch.txtto the corresponding nightly commit hashruntime/core/portable_type/c10/This PR was created automatically. If CI fails, Claude will attempt to fix issues (up to 3 attempts). If CI still fails, human review will be requested.
cc @jakeszwe