Remove fbjni dependencies from jni_layer_training #16556
Draft
+882
−280
Conversation
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
🔗 Helpful Links🧪 See artifacts and rendered test results at hud.pytorch.org/pr/pytorch/executorch/16556
Note: Links to docs will display an error until the docs builds have been completed. This comment was automatically generated by Dr. CI and updates every 15 minutes. |
meta-cla
bot
added
the
CLA Signed
This PR needs a
|
- Convert jni_layer_training.cpp from fbjni to pure JNI - Replace HybridClass with native handle pattern - Replace fbjni method registration with RegisterNatives - Update TrainingModule.java to use long mNativeHandle - Update SGD.java to use long mNativeHandle - Add finalize() and destroy() methods for cleanup - Update jni_layer.cpp to pass JNIEnv* to training registration Co-authored-by: kirklandsign <[email protected]>
…Warnings for finalize() Co-authored-by: kirklandsign <[email protected]>
…ormance Co-authored-by: kirklandsign <[email protected]>
Copilot
AI
changed the title
[WIP] Remove fbjni dependencies from jni_layer_training
Remove fbjni dependencies from jni_layer_training
Jan 13, 2026
Contributor
|
@copilot error: use of undeclared identifier 'jni_helper'; did you mean 'executorch::jni_helper'? |
…aration Co-authored-by: kirklandsign <[email protected]>
Author
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
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Outdated suggestions cannot be applied.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
Converts
jni_layer_training.cppfrom fbjni to pure JNI, eliminating the fbjni dependency for the Android training module.C++ (
jni_layer_training.cpp):ExecuTorchTrainingJniandSGDHybridHybridClasses with native handle classesTrainingJniCachefor cached class/method ID lookups (Tensor, EValue, HashMap, ByteBuffer, ByteOrder, Long/Double/Boolean wrappers)RegisterNativesinstead of fbjni registrationJava (
TrainingModule.java,SGD.java):HybridDatawithlong mNativeHandlecom.facebook.jni.*importsdestroy()andfinalize()for native resource cleanupIntegration (
jni_layer.cpp):register_natives_for_training(JNIEnv*)signature to receive env from callerusing namespace executorch::jni_helper;to fix undeclared identifier build errorTest plan
Code review and CodeQL security scan passed. The JNI layer changes follow the same patterns as PR #16553 which converts the other JNI layers. Full Android build/test requires NDK toolchain.
Original prompt
💬 We'd love your input! Share your thoughts on Copilot coding agent in our 2 minute survey.