[ExecuTorch][WebGPU] Add Llama K16 online causal attention

[JCNTH]

Stack from ghstack (oldest at bottom):

• [ExecuTorch][WebGPU] Lift FastVLM dispatch limits and harden graph construction #21602
• [ExecuTorch][WebGPU] Cover Conv1d routes and dynamic manifests #21601
• [ExecuTorch][WebGPU] Support general groups=1 Conv1d #21600
• [ExecuTorch][WebGPU] Test byte-packed BOOL tails #21599
• [ExecuTorch][WebGPU] Support byte-packed BOOL storage #21598
• [ExecuTorch][Vulkan] Preserve persistent buffer mutations #21597
• [ExecuTorch][WebGPU] Preserve Linear and Q4 embedding params across dynamic resize #21483
• [ExecuTorch][WebGPU] Generate logical and arithmetic binary shader variants #21451
• [ExecuTorch][WebGPU] Generate extrema and unary shader variants #21450
• [ExecuTorch][WebGPU] Generate typed to_copy shader variants #21449
• [ExecuTorch][WebGPU] Make WGSL generation rollback-safe #21448
• [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

Materializing the full attention-weight matrix for Llama prefill is memory- and
bandwidth-heavy and does not scale to longer sequences. This adds a single-pass
online-softmax K16 causal-attention kernel for the exact Llama Hq32/Hkv8/G4/D64
geometry with fp16 KV, guarded by explicit adapter limits, so accepted prefill
shapes through S512 never materialize attention weights. S1 keeps the existing
FlashDecoding path, and unsupported geometry, storage, or capabilities fall back
to the materialized implementation. No Vulkan analogue (WebGPU-specific): the
Vulkan backend has only a materialized attention (compute the weights, then a
separate multi-pass softmax), with no online-softmax kernel.

Key changes:

• runtime/ops/sdpa/streaming_attention_k16_causal_bound.wgsl (+ generated
  header): the online-softmax single-pass causal kernel.
• Sdpa.cpp: exact Llama geometry and capability guards, S512 transition
  handling, and fallback to the materialized route.
  @exported-using-ghexport

Differential Revision: D113171738

Differential Revision: D113171738

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/21132

• 📄 Preview Python docs built from this PR

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

❌ 1 Unclassified Failure

As of commit 29614aa with merge base ad3a71f ():

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:

• periodic / test-models-linux (buck2, mv3, portable, linux.2xlarge, 90) / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  ##[error]fatal: unable to access 'https://github.com/zeux/volk/': Error in the HTTP2 framing layer

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.

[psiddh]

Can you check linter