【UnitTestFix No.9】: Fix test_memcpy_op unittest error and support PIR API compatibility #75395
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
PR Category Operator Mechanism PR Types Bug fixes Description Fix test_memcpy_op unittest error and support PIR API compatibility. The Paddle framework has migrated from old IR to new PIR (Program IR), causing API compatibility issues in static graph tests. test_memcpy_op unit test now supports PIR API and uses dynamic graph mode for better compatibility. Fixed data type checking issues where paddle.float32 corresponds to different objects in different contexts.
|
|
|
你的PR提交成功，感谢你对开源项目的贡献! |
…py_op.py - Replace all Chinese comments with English equivalents - Improve code readability for international developers - Maintain consistent comment language throughout the codebase
- Replace static graph API with dynamic graph mode to avoid PIR compatibility issues - Fix create_var and append_op calls that are not available in PIR API - Add CUDA fallback handling for CPU-only environments - Ensure all tests pass in both CUDA and CPU environments - Maintain original test logic and assertions PR Category: Operator Mechanism PR Types: Bug fixes Description: Fix memcpy unittest error and support PIR API compatibility. The test now uses dynamic graph mode to avoid PIR API compatibility issues while maintaining the same test coverage.
- Remove dependency on op_test module to fix import errors - Add local get_device_place function implementation - Fix static graph mode issues in API test - Ensure all tests pass in GPU environment with CUDA support All 5 tests now pass successfully: - test_gpu_copy_to_pinned: GPU to GPU Pinned memory copy - test_pinned_copy_gpu: GPU Pinned memory to GPU copy - test_hip_copy_bool_value: HIP environment bool value copy - test_SELECTED_ROWS: SELECTED_ROWS type error handling - test_api: API functionality test
Contributor
|
请修复 CodeStyle 流水线错误 |
Contributor
Author
|
@luotao1 修复好了，请求再次批准检查 |
YqGe585
suggested changes
Sep 23, 2025
Contributor
Author
|
@luotao1 修复好了，等待检查 |
Contributor
|
PR名称麻烦按照规范修改。 |
Dayuxiaoshui
changed the title
Contributor
Author
|
修改完成 |
Contributor
Author
|
@luotao1 请求审查 |
luotao1
approved these changes
Sep 24, 2025
Contributor
|
hi, @Dayuxiaoshui
|
wanglezz
pushed a commit
to wanglezz/Paddle
that referenced
this pull request
Sep 25, 2025
… API compatibility (PaddlePaddle#75395) * fix: Fix test_memcpy_op unittest error and support PIR API compatibility PR Category Operator Mechanism PR Types Bug fixes Description Fix test_memcpy_op unittest error and support PIR API compatibility. The Paddle framework has migrated from old IR to new PIR (Program IR), causing API compatibility issues in static graph tests. test_memcpy_op unit test now supports PIR API and uses dynamic graph mode for better compatibility. Fixed data type checking issues where paddle.float32 corresponds to different objects in different contexts. * refactor: Replace Chinese comments with English comments in test_memcpy_op.py - Replace all Chinese comments with English equivalents - Improve code readability for international developers - Maintain consistent comment language throughout the codebase * Fix test_memcpy_op unittest compatibility with PIR API - Replace static graph API with dynamic graph mode to avoid PIR compatibility issues - Fix create_var and append_op calls that are not available in PIR API - Add CUDA fallback handling for CPU-only environments - Ensure all tests pass in both CUDA and CPU environments - Maintain original test logic and assertions PR Category: Operator Mechanism PR Types: Bug fixes Description: Fix memcpy unittest error and support PIR API compatibility. The test now uses dynamic graph mode to avoid PIR API compatibility issues while maintaining the same test coverage. * Fix test_memcpy_op.py for GPU environment compatibility - Remove dependency on op_test module to fix import errors - Add local get_device_place function implementation - Fix static graph mode issues in API test - Ensure all tests pass in GPU environment with CUDA support All 5 tests now pass successfully: - test_gpu_copy_to_pinned: GPU to GPU Pinned memory copy - test_pinned_copy_gpu: GPU Pinned memory to GPU copy - test_hip_copy_bool_value: HIP environment bool value copy - test_SELECTED_ROWS: SELECTED_ROWS type error handling - test_api: API functionality test * chore: fix EOF newline for test_memcpy_op.py via pre-commit end-of-file-fixer * test(legacy): replace get_device_place with base.CUDAPlace(0); fix EOF newline
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
4 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.
PR Category
Operator Mechanism
PR Types
Bug fixes
Description