Summary:
Tests in TestNumbaIntegration are failing during PyTorch unit test execution on CPU platform with segmentation faults and CUDA errors.

Test Class: test/test_numba_integration.py::TestNumbaIntegration
Number of Failing Tests: 3
Platform: CPU
Test Type: Unit Test

Version Information:

• PyTorch Commit: 6bdd8c9
• Branch: main
• Test Date: 2026-01-14
• Sprint: Sprint 24

Failure Pattern:
Tests failing with 2 related error patterns - segmentation faults and CUDA context errors

Error Patterns:
1.

Fatal Python error: Segmentation fault

2.

numba.cuda.cudadrv.driver.CudaAPIError: [201] Call to cuCtxGetDevice results in CUDA_ERROR_INVALID_CONTEXT

Failing Tests:
1. test_active_device
2. test_array_adaptor
3. test_from_cuda_array_interface_inferred_strides

Steps to Reproduce:
1. Run test command:

   TEST_CONFIG=cpu python3 test/run_test.py -i test_numba_integration
   TEST_CONFIG=cuda python3 test/run_test.py -i test_numba_integration
   

2. Observe segmentation faults and CUDA context errors

Expected Result:
Tests should pass

Actual Result:
Tests fail with segmentation faults and CUDA API errors on CPU platform

Additional Context:

• Note: sGPU ticket AIPCC-8267 exists for the same test class
• This is the CPU-specific failure
• Tests are failing consistently after reruns

Logs:
Test execution logs: /home/ktanmay/Downloads/Run 1-20260120T060019Z-1-001/Run 1/20260114_024940_commit_6bdd8c9/cpu_tests.log

Priority: P3

Labels: pytorch, unittest, cpu, numba, segfault