[executorch][cuda] Add step_plan for the CUDA batching executor

[Gasoonjia]

Stack from ghstack (oldest at bottom):

• [executorch][muse] Move off-graph KV to the batching path and test batching in OSS CI #23375
• [executorch][muse] Add run_solo_batching #23374
• [executorch][muse] Add export_solo_batching.py #23373
• [executorch][cuda] Let CudaExecutor honor a program's prefill lower bound #23372
• [executorch][llm] Share the batching executors' session logic under extension/llm/batching/util #23371
• [executorch][cuda] Add CudaExecutor #23370
• -> [executorch][cuda] Add step_plan for the CUDA batching executor #23369
• [executorch][llm] Add Buck targets for extension/llm/batching #23368
• [executorch][cuda] Add CudaCellCache #23367
• [executorch][cuda] Lower off-graph KV in a cell layout #23366
• [executorch][cuda] Extract CudaKVPool from CudaSequenceKVCache #23365

Step slicing for the CUDA batching executor, in backends/cuda/batching:
plan_slices (step_plan.h, header-only) cuts a batch of packed tokens into
forwards the exported decode/prefill methods can run, in order, so a slice
attends every cell its predecessors wrote.

Slices take up to max_step_tokens each. A one-token slice runs Decode (the
static, one-token method a CUDA graph is captured for). A slice shorter than
min_prefill_tokens -- the lower bound the prefill method was exported
with -- runs as that many Decode forwards; anything else runs Prefill (dynamic
over [min_prefill_tokens, max_step_tokens]).

Builds with Buck (step_plan + test_step_plan) and CMake (header-only
install, test_step_plan under BUILD_TESTING; the batching dir is added from
the OSS root after extension/llm/batching). The OSS unittest-cuda-runtime job
builds and runs test_step_plan. The CudaExecutor itself, which consumes this
header, follows in the child diff.

Differential Revision: D123066805

[pytorch-bot]

🔗 Helpful Links

🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/23369

• 📄 Preview Python docs built from this PR

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

❌ 1 New Failure, 4 Unclassified Failures

As of commit 80bfc20 with merge base c16dd0a ():

NEW FAILURE - The following job has failed:

• Lint / lintrunner (gh)
  >>> Lint for backends/cuda/batching/CMakeLists.txt:

UNCLASSIFIED FAILURES - DrCI could not classify the following jobs because the workflow did not run on the merge base. The failures may be pre-existing on trunk or introduced by this PR:

• Test CUDA Builds / test-model-cuda-e2e-diar_streaming_sortformer_4spk-v2-non-quantized / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  RuntimeError: Command bash /__w/_temp/exec_script failed with exit code 134
• Test CUDA Builds / unittest-cuda / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  backends/cuda/tests/test_offgraph_kv.py::OffGraphKVCellCompileTest::test_decode_and_prefill_from_two_tokens_lower
• Test CUDA Builds / unittest-cuda-runtime / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  [ FAILED ] CudaKVCacheTest.CacheOutlivesTheStreamItLastSteppedOn
• Test ROCm AOTI / unittest-rocm-gfx950-rocm7.2 / linux-job (gh) (this job did not run on the merge base, so DrCI cannot tell whether the failure is pre-existing)
  backends/cuda/tests/test_offgraph_kv.py::OffGraphKVCellCompileTest::test_decode_and_prefill_from_two_tokens_lower

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.