[ExecuTorch][WebGPU] Preserve Linear and Q4 embedding params across dynamic resize by JCNTH · Pull Request #21483 · pytorch/executorch

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

JCNTH · 2026-07-30T03:35:06Z

Stack from ghstack (oldest at bottom):

Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.

Problem
The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update.

Solution

Before: build and resize populated control words independently, allowing immutable shader state to reset.
After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch.

Implementation

Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes.
EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize.
Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch.

Constraints
WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed.

Co-authored-with: Claude Code.
@exported-using-ghexport

Differential Revision: D113992326

[ghstack-poisoned]

pytorch-bot · 2026-07-30T03:35:09Z

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21483

📄 Preview Python docs built from this PR

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

❌ 4 New Failures, 31 Pending

As of commit b63add9 with merge base 28a7fac ():

NEW FAILURES - The following jobs have failed:

Cadence Build & Test / cpu-test / test-aot / test-aot (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID B98E:1BCB3:38320EF:BF0C7FE:6A760B4C and timestamp 2026-08-07 16:43:56 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
Cadence Build & Test / cpu-test / test-ops / test-ops (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID CDC4:35D683:8212F5D:1B50AB6E:6A760B4B and timestamp 2026-08-07 16:43:55 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, models) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 9E96:1780A5:36E7F7A:BAE9AA5:6A760B8D and timestamp 2026-08-07 16:45:01 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting
pull / test-qnn-testsuite-linux / test-backend-linux (qnn, operators) / linux-job (gh)
##[error]API rate limit exceeded for installation. If you reach out to GitHub Support for help, please include the request ID 83C0:303096:847FDF6:1BE44F2F:6A760B93 and timestamp 2026-08-07 16:45:07 UTC. For more on scraping GitHub and how it may affect your rights, please review our Terms of Service (https://docs.github.com/en/site-policy/github-terms/github-terms-of-service) - https://docs.github.com/en/rest/using-the-rest-api/getting-started-with-the-rest-api#rate-limiting

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

github-actions · 2026-07-30T03:35:58Z

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.

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409033795 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

[ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 409300312 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

psiddh

lgtm

fbsource master [ghstack-poisoned]

…ynamic resize Pull Request resolved: #21483 **Preserve immutable Linear and Q4 embedding shader controls across dynamic resize through typed parameter authorities.** **Problem** The Linear resize hook dropped has_bias, and the Q4 embedding resize path could drop is_linear_weight. Both defects produced correct build-time outputs but silently changed semantics after a live shape update. **Solution** - Before: build and resize populated control words independently, allowing immutable shader state to reset. - After: each build/resize pair calls one typed helper while recomputing only live counts and dispatch. **Implementation** - Linear.cpp — make_linear_params preserves the bias flag across vec4 and tiled routes. - EmbeddingQ4gsw.cpp — EmbeddingLayout and make_embedding_params preserve nibble layout across resize. - Mirrors Vulkan runtime/graph/ops/impl/Linear.cpp and EmbeddingQ4gsw.cpp, which carry immutable shader controls through dynamic dispatch. **Constraints** WGSL bytes, shader selection, bindings, uniform sizes, queue-write counts, dispatch formulas, and pipeline topology are unchanged. Production handler code is smaller because duplicated field population and scalar captures are removed. Co-authored-with: Claude Code. ghstack-source-id: 411961489 @exported-using-ghexport Differential Revision: [D113992326](https://our.internmc.facebook.com/intern/diff/D113992326/)

Update
26ffec7
[ghstack-poisoned]

JCNTH temporarily deployed to cadence July 30, 2026 03:35 — with GitHub Actions Inactive

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 Jul 30, 2026

JCNTH temporarily deployed to cadence July 30, 2026 04:18 — with GitHub Actions Inactive

Update
011ba2d
[ghstack-poisoned]

JCNTH temporarily deployed to cadence July 31, 2026 16:46 — with GitHub Actions Inactive

meta-codesyncBot added the meta-exported label Jul 31, 2026

JCNTH temporarily deployed to cadence July 31, 2026 17:11 — with GitHub Actions Inactive

Update
2039e33
[ghstack-poisoned]

JCNTH temporarily deployed to cadence July 31, 2026 21:21 — with GitHub Actions Inactive

JCNTH temporarily deployed to cadence August 5, 2026 20:35 — with GitHub Actions Inactive

psiddh approved these changes Aug 6, 2026
View reviewed changes

Rebase onto latest
b63add9
fbsource master [ghstack-poisoned]

JCNTH temporarily deployed to cadence August 7, 2026 16:36 — with GitHub Actions Inactive

JCNTH temporarily deployed to cadence August 7, 2026 17:01 — with GitHub Actions Inactive

meta-codesyncBot merged commit 533541f into gh/JCNTH/203/baseAug 7, 2026
179 of 184 checks passed

meta-codesyncBot deleted the gh/JCNTH/203/head branch August 7, 2026 17:22

meta-codesyncBottemporarily deployed to cherry-pick-bot August 7, 2026 17:22 Inactive

pytorchbot mentioned this pull request Aug 7, 2026
[ExecuTorch][WebGPU] Preserve Linear and Q4 embedding params across dynamic resize #21661
Merged