Cortex-M: add opt-in explicit layout lowering - #21928
Draft
rascani wants to merge 1 commit into
Draft
Conversation
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21928
Note: Links to docs will display an error until the docs builds have been completed. ❌ 29 New Failures, 1 Unrelated Failure, 1 Unclassified FailureAs of commit 624ccd3 with merge base ed2d9ec ( NEW FAILURES - The following jobs have failed:
UNCLASSIFIED FAILURE - DrCI could not classify the following job because the workflow did not run on the merge base. The failure may be pre-existing on trunk or introduced by this PR:
BROKEN TRUNK - The following job failed but were present on the merge base:👉 Rebase onto the `viable/strict` branch to avoid these failures
This comment was automatically generated by Dr. CI and updates every 15 minutes. |
This was referenced Aug 18, 2026
This PR needs a
|
meta-cla
Bot
added
the
CLA Signed
rascani
marked this pull request as draft
August 18, 2026 23:08
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
Stack from ghstack (oldest at bottom):
Thread an experimental explicit-layout mode through the shared Cortex-M Edge compile configuration, AOT compiler, quantizer support, tester, and pass manager. Legacy dim-order export remains the default. Explicit mode disables dim-order operators, converts Conv1d through the common transform, and uses the common channels-last dialect and region optimizer to form layout regions across supported convolution and pooling operations.
Lower channels-last anchors to fixed NHWC Cortex-M schemas and surviving structural boundaries to cortex_m::transpose. Constant padding inside a region selects pad_nhwc. Channel-broadcast add and multiply use explicit-layout quantizer checks and fixed NHWC runtime schemas, keeping these regions int8 instead of introducing float QDQ islands. Equal-shape add and multiply continue to use the layout-independent legacy schemas. Softmax remains outside automatic propagation so its dimension satisfies the Cortex-M last-dimension kernel contract.
Preserve complete operator metadata through common replacement, while explicitly requiring per-tensor activation and output qparams for tensors whose dimensions change. Per-channel activation or output axes are rejected from explicit regions until a backend remapping contract exists; per-channel weight metadata remains valid because weights are not permuted.
Add graph, planned-memory, and FVP coverage for regular, depthwise, transpose, pooling, padding, multi-op, channel-broadcast add/multiply, softmax-barrier, and Conv1d paths, including the public AOT flag. Planned-arena tests assert the expected targets before comparing memory. The channel-broadcast cases plan 1,600 bytes in both modes, eliminating the previous explicit-mode 2,304-byte float-intermediate plan.
Differential Revision: D116374086
cc @digantdesai @freddan80 @per @zingo @oscarandersson8218 @mansnils @Sebastian-Larsson @robell