Use safe numel() in ET (retake) by lucylq · Pull Request #19130 · pytorch/executorch

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

lucylq · 2026-04-25T01:08:15Z

#19075 failed to cp to main

Differential Revision: D102082911 Pull Request resolved: #19075

pytorch-bot · 2026-04-25T01:08:20Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19130

📄 Preview Python docs built from this PR

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

❗ 1 Active SEVs

There are 1 currently active SEVs. If your PR is affected, please view them below:

Rolling out OSDC (ARC) runners on pull & trunk workflows in PyTorch main

⏳ No Failures, 140 Pending

As of commit 955a561 with merge base 222711e ():
💚 Looks good so far! There are no failures yet. 💚

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

github-actions · 2026-04-25T01:09:02Z

This PR needs a `release notes:` label

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"

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

Copilot

Pull request overview

This PR hardens tensor size/byte-count computations across portable runtime, tensor extension helpers, and WASM bindings by switching to overflow-aware numel calculation and adding explicit overflow checks before allocating or validating buffers.

Changes:

Replace compute_numel with safe_numel in portable tensor resizing and WASM tensor input validation.
Add overflow-checked (numel * elementSize) byte-count computation in make_tensor_ptr, empty_strided, and clone_tensor_ptr.
Improve error handling paths where numel/nbytes computations can fail.

Reviewed changes

Copilot reviewed 5 out of 5 changed files in this pull request and generated 3 comments.

Show a summary per file

File	Description
runtime/core/portable_type/tensor_impl.cpp	Uses `safe_numel` when resizing dynamic tensors to avoid silent overflow.
extension/wasm/wasm_bindings.cpp	Switches expected-numel computation to `safe_numel` and propagates errors to JS.
extension/tensor/tensor_ptr_maker.cpp	Computes `nbytes` with overflow detection when allocating empty tensors.
extension/tensor/tensor_ptr.h	Uses `safe_numel` to validate data length and view element counts safely.
extension/tensor/tensor_ptr.cpp	Adds overflow checks for `nbytes` in tensor creation and cloning.

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

+ auto new_numel_result = safe_numel(new_sizes.data(), dim_);
+ if (!new_numel_result.ok()) {
+ return new_numel_result.error();
+ }
+ const auto new_numel = new_numel_result.get();


 THROW_IF_FALSE(
- data.size() >= computed_numel,
+ data.size() >= static_cast<size_t>(computed_numel.get()),
 "Required %ld elements, given %ld",


+ !c10::mul_overflows(
+ static_cast<size_t>(numel),
+ executorch::aten::elementSize(type),
+ &nbytes),
+ "Overflow computing nbytes: numel=%zd element_size=%zu",
+ numel,
+ executorch::aten::elementSize(type));
+ ET_CHECK_MSG(
+ data.size() == nbytes,
+ "Data size (%zu) does not match tensor size (%zu).",


pytorch#19075 failed to cp to main

Use safe_numel() in et
955a561
Differential Revision: D102082911 Pull Request resolved: #19075

lucylq requested a review from JacobSzwejbka as a code owner April 25, 2026 01:08

CopilotAI review requested due to automatic review settings April 25, 2026 01:08

lucylq requested a review from shoumikhin as a code owner April 25, 2026 01:08

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 Apr 25, 2026

lucylq changed the title ~~...~~Use safe numel() in ET (retake)Apr 25, 2026

Copilotstarted reviewing on behalf of lucylq April 25, 2026 01:08 View session

kirklandsign approved these changes Apr 25, 2026
View reviewed changes

CopilotAI reviewed Apr 25, 2026
View reviewed changes

lucylq merged commit e1cd352 into mainApr 25, 2026
168 of 173 checks passed

lucylq deleted the gh/lucylq/155/base branch April 25, 2026 01:12

zeel2104 pushed a commit to zeel2104/executorch that referenced this pull request May 5, 2026
Use safe numel() in ET (retake) (pytorch#19130)
fda275a
pytorch#19075 failed to cp to main