Fix: delay CUDADriverWrapper instantiation to avoid uncaught exception #25117
+1
−5
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
…ns when CUDA is unavailable Moved the static instance of CUDADriverWrapper into GetInstance() as a function-local static. This ensures the constructor is only called on first use, allowing exceptions (e.g., missing libcuda.so.1) to be caught gracefully, rather than causing process termination during static initialization.
|
@microsoft-github-policy-service agree company="Pix4D" |
tianleiwu
approved these changes
Jun 19, 2025
|
/azp run Linux QNN CI Pipeline, Win_TRT_Minimal_CUDA_Test_CI, Windows ARM64 QNN CI Pipeline, Windows GPU Doc Gen CI Pipeline, Windows x64 QNN CI Pipeline |
|
Azure Pipelines successfully started running 5 pipeline(s). |
snnn
reviewed
Jun 19, 2025
| const CUDADriverWrapper* CUDADriverWrapper::GetInstance() { | ||
| static CUDADriverWrapper instance; |
There was a problem hiding this comment.
Please be very careful that this object will be deleted before all global variables. Please check who depends on this.
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Outdated suggestions cannot be applied.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.
Fix: delay CUDADriverWrapper instantiation to avoid uncaught exceptions when CUDA is unavailable
Description
This PR moves the static instantiation of CUDADriverWrapper from a class-level static field to a function-local static inside CUDADriverWrapper::GetInstance(). This change ensures that the CUDA driver is only loaded when the instance is actually needed, rather than at static initialization time. It preserves the singleton behavior while deferring instantiation to runtime.
Motivation and Context
When libcuda.so.1 is not available on the system, the constructor of CUDADriverWrapper throws an exception. Previously, this exception was triggered during static initialization, leading to an uncatchable std::terminate() and process termination. By moving the instance into GetInstance() as a function-local static, the exception can now be caught by client code (e.g., in try/catch), allowing graceful fallback when CUDA is unavailable.