feat(mlx): add handler for aten.roll

[IshanG97]

Summary

Adds an MLX delegate handler for aten.roll, mapping torch.roll onto mlx::core::roll via a new RollNode in the schema. Replaces the default decomposition (index_select + arange + cat) with a single native kernel — needed by Swin Transformer's shift-window attention.

Flat roll (dims=[]) raises NotImplementedError for now; no known consumer needs it yet.

Generated files (MLXLoader.*, schema_generated.h, mlx_graph_schema.py, _generated_serializers.py, _generated_inspector.py, _generated/) are regenerated from schema.fbs by backends/mlx/CMakeLists.txt at build time and are deliberately not committed.

Fixes #18919.

Test plan

• python backends/mlx/serialization/generate.py — regenerates cleanly with RollNode in all expected outputs.
• lintrunner --skip MYPY --paths-cmd 'git diff --name-only upstream/main' — no issues.
• End-to-end run_all_tests -k roll not run locally (no executorch build on this machine); relying on CI. Happy to push fixes if it finds anything.

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/19038

• 📄 Preview Python docs built from this PR

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

⚠️ 12 Awaiting Approval

As of commit eb9cc01 with merge base 3be4546 ():

AWAITING APPROVAL - The following workflows need approval before CI can run:

• Build documentation (gh)
• Build Presets (gh)
• Cadence Build & Test (gh)
• Lint (gh)
• MLX (gh)
• periodic (gh)
• pull (gh)
• Test ARM Backend (gh)
• Test CoreML Backend (gh)
• Test QNN Backend (gh)
• Test Vulkan Backend (gh)
• Test XNNPACK Backend (gh)

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

[github-actions]

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.

[metascroy]

Looks fantastic @IshanG97!

Let's see what CI says, but is there a reason you didn't run the test locally? Ideally, I'd like to see the output of the test in the PR summary :)

[metascroy]

Ignore all of the test-mlx-llm result failures in CI, they have to do with an HF token issue on external contributions.

[IshanG97]

Looks fantastic @IshanG97!

Let's see what CI says, but is there a reason you didn't run the test locally? Ideally, I'd like to see the output of the test in the PR summary :)

Hi @metascroy! Sorry for the late response, I didn't initially because I couldn't get the latest version of Xcode on my machine, I had to use a proxy (to be honest, I was relying on the CI to catch anything haha!). I've now downloaded Xcode and run the tests locally, see my trimmed output:

  $ python -m executorch.backends.mlx.test.run_all_tests roll -v

  ============================================================
  Running test: roll_shift(2)_dim(0)
  ============================================================

  Step 1: Generating test files...
  Exporting model to .../op_tests/roll_shift(2)_dim(0)/model.pte

  ============================================================
  EXPORTED PROGRAM (torch.export)
  ============================================================
  ExportedProgram:
      class GraphModule(torch.nn.Module):
          def forward(self, x: "f32[8]"):
              roll: "f32[8]" = torch.ops.aten.roll.default(x, [2], [0]);  x = None
              return (roll,)

  ============================================================
  EDGE PROGRAM (after decomposition)
  ============================================================
  ExportedProgram:
      class GraphModule(torch.nn.Module):
          def forward(self, x: "f32[8]"):
              lowered_module_0 = self.lowered_module_0
              executorch_call_delegate = torch.ops.higher_order.executorch_call_delegate(lowered_module_0, x);  lowered_module_0 = x = None
              getitem: "f32[8]" = executorch_call_delegate[0];  executorch_call_delegate = None
              return (getitem,)

  ======================================================================
  MLX Graph Summary
  ======================================================================
  Version: 1
  Constant tensors: 0
  Input tensors: 1
  Output tensors: 1
  Instruction chains: 1
  Main chain idx: 0

  Chain 0 (main) (1 instructions):
    [0] RollNode
        x: tid 0
        out: tid 1
        shift: [{'literal': 2}]
        axes: [0]

  Inputs:  tid 0
  Outputs: tid 1
  ======================================================================

  Step 2: Verifying MLX delegation...
    Expected MLX segments: 1
    Actual MLX segments:   1
  ✓ MLX delegation verified

  Step 3: Running C++ binary...
  C++ binary output: Loading model...
  Model loaded successfully
    Input 0: dtype=0, shape=[8]
  Executing forward...
  Execution succeeded, 1 outputs
    Output 0: dtype=0, shape=[8]
  OK

  Step 4: Comparing outputs...
  ✓ PASSED: All outputs match

  ============================================================
  Running test: roll_shift(1)_dim(0)        ... ✓ PASSED
  Running test: roll_shift(-2)_dim(1)       ... ✓ PASSED
  Running test: roll_shift(3)_dim(2)        ... ✓ PASSED
  Running test: roll_shift(1,2)_dim(0,2)    ... ✓ PASSED
  Running test: roll_shift(-1,-2,-3)_dim(0,1,2) ... ✓ PASSED
  Running test: roll_shift(2)_dim(-1)       ... ✓ PASSED
  ============================================================

  ============================================================
  TEST SUMMARY
  ============================================================
  Passed: 7
  Failed: 0
  ============================================================