[ExecuTorch][WebGPU] Replace broad QKV fusion with BK64 kernel

[JCNTH]

Stack from ghstack (oldest at bottom):

• [ExecuTorch][WebGPU] Add Qwen3 K16 attention and fp16 KV cache #21136
• [ExecuTorch][WebGPU] Add HuggingFace rotate-half RoPE #21135
• [ExecuTorch][WebGPU] Repair dynamic SDPA routing, add attestation #21134
• [ExecuTorch][WebGPU] Select emdawnwebgpu for Emscripten builds #21133
• [ExecuTorch][WebGPU] Add Llama K16 online causal attention #21132
• -> [ExecuTorch][WebGPU] Replace broad QKV fusion with BK64 kernel #21131
• [ExecuTorch][WebGPU] Route exact prefill shapes to the BK64 GEMM #21130
• [ExecuTorch][WebGPU] Fuse the SwiGLU elementwise chain #21129
• [ExecuTorch][WebGPU] Coordinate dynamic dispatch routes #21128
• [ExecuTorch][WebGPU] Add certified scoped output suppression #21127

The previously landed broad QKV fusion applied too widely and did not match the
BK64 schedule now used for the ordinary projections. This corrective diff
replaces it with a capability- and geometry-qualified BK64 kernel that fuses the
exact Llama Q/K/V projection triple at live M128, M508, and M512, packing the
constant weights and scales once and scattering the result into three distinct
planner-safe outputs. Outside the accepted shapes it switches atomically back
to the ordinary Steel and bicol routes, and it deletes the obsolete broad
shader and header so only one QKV path remains. Mirrors Vulkan
xplat/executorch/backends/vulkan/runtime/graph/ops/glsl/q4gsw_linear_gemm__w_4x8.glsl
for the per-projection GEMM; the three-output fusion itself is WebGPU-specific.

Key changes:

• Adds runtime/ops/quantized_linear/q4gsw_qkv_bk64.wgsl (+ generated header);
  removes the obsolete q4gsw_linear_gemm_qkv_fused shader and header.
• WebGPUGraph.{cpp,h}: geometry and capability predicate, one-time constant
  packing, distinct Q/K/V outputs, and atomic fallback to Steel/bicol.
  @exported-using-ghexport

Differential Revision: D113171749

Differential Revision: D113171749

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21131

• 📄 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:

• Unrelated MacOS MPS test jobs time out (30 min) due to stale cpp_extension lock file poisoning self-hosted runners

❌ 2 New Failures, 1 Unrelated Failure

As of commit 9d1f0c1 with merge base 86c3470 ():

NEW FAILURES - The following jobs have failed:

• pull / test-models-linux (ic4, portable, linux.4xlarge.memory) / linux-job (gh)
  RuntimeError: Command docker exec -t 4ed5382a0e06f051659349d894e37f9dda20e442252fb1dbbaf004273250608e /exec failed with exit code 1
• pull / test-qnn-wheel-packages-linux (3.12) / linux-job (gh)
  RuntimeError: Command docker exec -t 3825281256d4b1c2cc82111e89998ceab1c1db49b297e6c3f695aa4da1f3fe64 /exec failed with exit code 1

BROKEN TRUNK - The following job failed but were present on the merge base:

👉 Rebase onto the `viable/strict` branch to avoid these failures

• pull / unittest-nxp-neutron / linux-job (gh) (trunk failure)
  ##[error]The operation was canceled.

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.