Refactor: drop _prepared_ prefix from internal DeviceRunner callable methods#911
Merged
ChaoWao merged 1 commit intoMay 30, 2026
Conversation
… methods
Internal DeviceRunner methods carried a `_prepared_` infix that historically
distinguished "register-via-the-prepare_callable-path" from a now-removed
eager-register path. With only one path remaining, the prefix is dead
naming and creates an asymmetry with the public C API
(`prepare_callable` / `unregister_callable` / `run_prepared`).
Renames (pure mechanical, no behavior change):
- `DeviceRunner::register_prepared_callable[_host_orch]` → `register_callable[_host_orch]`
- `DeviceRunner::has_prepared_callable` → `has_callable`
- `DeviceRunner::unregister_prepared_callable` → `unregister_callable`
- `DeviceRunner::bind_prepared_callable_to_runtime` → `bind_callable_to_runtime`
- `prepared_callables_` → `callables_`
- `PreparedCallableState` → `CallableState`
- `BindPreparedCallableResult` → `BindCallableResult`
- `PreparedCallableArtifacts` → `CallableArtifacts`
- `bind_prepared_to_runtime_impl` → `bind_callable_to_runtime_impl` (runtime_maker)
Public Python / c_api entry points (`prepare_callable`, `run_prepared`,
`aicpu_dlopen_count`, `host_dlopen_count`) and the public
`unregister_callable` free function are unchanged — those names have real
semantic meaning ("prepare the callable", "run something previously
prepared").
Both arches built clean (onboard + sim, both runtimes). a2a3 local smoke
(dummy_task, prepared_callable suite) 7/7 passed in 11s.
📝 WalkthroughWalkthroughThis PR systematically renames the callable-registration and binding APIs from "prepared_callable" to "callable" terminology across all DeviceRunner platform implementations (a2a3 onboard, a2a3 simulation, a5 onboard, a5 simulation), updating type names, method signatures, internal state storage, runtime-binding implementations, and documentation. ChangesCallable API Rename
🎯 4 (Complex) | ⏱️ ~60 minutes
🚥 Pre-merge checks | ✅ 4 | ❌ 1❌ Failed checks (1 warning)
✅ Passed checks (4 passed)
✏️ Tip: You can configure your own custom pre-merge checks in the settings. Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out. Comment |
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request performs a comprehensive refactoring across the codebase to simplify terminology by renaming 'prepared callable' references to simply 'callable'. This includes renaming classes (e.g., PreparedCallableState to CallableState, PreparedCallableArtifacts to CallableArtifacts), functions (e.g., register_prepared_callable to register_callable, bind_prepared_callable_to_runtime to bind_callable_to_runtime), member variables, and associated comments across Python, C++, and test files. No review comments were provided, so there is no additional feedback to address.
![coderabbitai[bot]](https://avatars.githubusercontent.com/in/347564?s=60&v=4){kind=small}
There was a problem hiding this comment.
Caution
Some comments are outside the diff and can’t be posted inline due to platform limitations.
⚠️ Outside diff range comments (2)
src/a2a3/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp (1)
97-109:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winRemove the stale
@param runtimeentry from this docblock.
prepare_callable_implno longer takes aRuntime *, so this comment now documents a nonexistent parameter.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/a2a3/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp` around lines 97 - 109, Remove the stale "`@param` runtime" line from the docblock above prepare_callable_impl since the function signature no longer accepts a Runtime pointer; update the comment to only document the current parameters (callable, upload_fn, out) and return value so the docblock matches the prepare_callable_impl declaration and related symbols like CallableArtifacts and ChipCallable.src/a5/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp (1)
97-109:⚠️ Potential issue | 🟡 Minor | ⚡ Quick winDrop the stale
@param runtimeline here too.The comment still documents a
Runtime *parameter that is no longer part ofprepare_callable_impl.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/a5/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp` around lines 97 - 109, The function comment above prepare_callable_impl still documents a removed Runtime* parameter; update the docblock for prepare_callable_impl by removing the stale "`@param` runtime" line (and any accompanying description) so the comment accurately reflects the current signature (prepare_callable_impl(const ChipCallable*, uint64_t (*upload_fn)(const void *), CallableArtifacts*)). Ensure remaining `@param` entries match the actual parameters (callable, upload_fn, out) and keep the rest of the explanatory text unchanged.
🧹 Nitpick comments (1)
src/a5/runtime/tensormap_and_ringbuffer/runtime/runtime.h (1)
300-303: ⚡ Quick winFinish the terminology rename in this comment.
After switching the API reference to
bind_callable_to_runtime, "prepared kernel binaries" is now the odd term out in the same sentence. Consider renaming it to "callable kernel binaries" for consistency.🤖 Prompt for AI Agents
Verify each finding against current code. Fix only still-valid issues, skip the rest with a brief reason, keep changes minimal, and validate. In `@src/a5/runtime/tensormap_and_ringbuffer/runtime/runtime.h` around lines 300 - 303, Update the comment in runtime.h to use consistent terminology: replace "prepared kernel binaries" with "callable kernel binaries" in the comment that documents replaying a previously-uploaded kernel address onto a fresh Runtime (the block referencing bind_callable_to_runtime, registered_kernel_func_ids_, and validate_runtime_impl) so the phrasing matches the API rename to bind_callable_to_runtime.
🤖 Prompt for all review comments with AI agents
Verify each finding against current code. Fix only still-valid issues, skip the
rest with a brief reason, keep changes minimal, and validate.
Outside diff comments:
In `@src/a2a3/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp`:
- Around line 97-109: Remove the stale "`@param` runtime" line from the docblock
above prepare_callable_impl since the function signature no longer accepts a
Runtime pointer; update the comment to only document the current parameters
(callable, upload_fn, out) and return value so the docblock matches the
prepare_callable_impl declaration and related symbols like CallableArtifacts and
ChipCallable.
In `@src/a5/runtime/tensormap_and_ringbuffer/host/runtime_maker.cpp`:
- Around line 97-109: The function comment above prepare_callable_impl still
documents a removed Runtime* parameter; update the docblock for
prepare_callable_impl by removing the stale "`@param` runtime" line (and any
accompanying description) so the comment accurately reflects the current
signature (prepare_callable_impl(const ChipCallable*, uint64_t
(*upload_fn)(const void *), CallableArtifacts*)). Ensure remaining `@param`
entries match the actual parameters (callable, upload_fn, out) and keep the rest
of the explanatory text unchanged.
---
Nitpick comments:
In `@src/a5/runtime/tensormap_and_ringbuffer/runtime/runtime.h`:
- Around line 300-303: Update the comment in runtime.h to use consistent
terminology: replace "prepared kernel binaries" with "callable kernel binaries"
in the comment that documents replaying a previously-uploaded kernel address
onto a fresh Runtime (the block referencing bind_callable_to_runtime,
registered_kernel_func_ids_, and validate_runtime_impl) so the phrasing matches
the API rename to bind_callable_to_runtime.
ℹ️ Review info
⚙️ Run configuration
Configuration used: Organization UI
Review profile: CHILL
Plan: Pro
Run ID: 0e2c7584-1e2e-4643-9112-34c944aef3d1
📒 Files selected for processing (26)
python/simpler/worker.pysrc/a2a3/platform/onboard/host/device_runner.cppsrc/a2a3/platform/onboard/host/device_runner.hsrc/a2a3/platform/onboard/host/pto_runtime_c_api.cppsrc/a2a3/platform/sim/host/device_runner.cppsrc/a2a3/platform/sim/host/device_runner.hsrc/a2a3/platform/sim/host/pto_runtime_c_api.cppsrc/a2a3/runtime/host_build_graph/host/runtime_maker.cppsrc/a2a3/runtime/host_build_graph/runtime/runtime.hsrc/a2a3/runtime/tensormap_and_ringbuffer/aicpu/aicpu_executor.cppsrc/a2a3/runtime/tensormap_and_ringbuffer/host/runtime_maker.cppsrc/a2a3/runtime/tensormap_and_ringbuffer/runtime/runtime.hsrc/a5/platform/onboard/host/device_runner.cppsrc/a5/platform/onboard/host/device_runner.hsrc/a5/platform/onboard/host/pto_runtime_c_api.cppsrc/a5/platform/sim/host/device_runner.cppsrc/a5/platform/sim/host/device_runner.hsrc/a5/platform/sim/host/pto_runtime_c_api.cppsrc/a5/runtime/host_build_graph/host/runtime_maker.cppsrc/a5/runtime/tensormap_and_ringbuffer/aicpu/aicpu_executor.cppsrc/a5/runtime/tensormap_and_ringbuffer/host/runtime_maker.cppsrc/a5/runtime/tensormap_and_ringbuffer/runtime/runtime.hsrc/common/task_interface/callable_protocol.hsrc/common/task_interface/prepare_callable_common.htests/st/a2a3/host_build_graph/prepared_callable/test_prepared_callable.pytests/ut/py/test_worker/test_host_worker.py
ChaoWao
approved these changes
May 30, 2026
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
2 participants
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.
Summary
Pure rename PR. Drops the meaningless
_prepared_infix from internalDeviceRunnermethods and supporting types. With only one registration path remaining today, the prefix is dead naming and creates an awkward asymmetry with the public C API (prepare_callable/unregister_callable/run_prepared).Renames
DeviceRunner::register_prepared_callable[_host_orch]register_callable[_host_orch]DeviceRunner::has_prepared_callablehas_callableDeviceRunner::unregister_prepared_callableunregister_callable(matches c_api)DeviceRunner::bind_prepared_callable_to_runtimebind_callable_to_runtimeprepared_callables_(map)callables_PreparedCallableState(struct)CallableStateBindPreparedCallableResult(struct)BindCallableResultPreparedCallableArtifacts(struct)CallableArtifactsbind_prepared_to_runtime_impl(free fn)bind_callable_to_runtime_implWhat's unchanged
Public API surface keeps its original names because the verb "prepared" carries real meaning there:
prepare_callable(public Python + c_api — "prepare this callable")run_prepared(public Python + c_api — "run something previously prepared")aicpu_dlopen_count/host_dlopen_count(public diagnostics)unregister_callable(public c_api — already had this name, now method matches)Verification
Test plan