Links

Content Skeleton

This Page

Previous topic

CUDA Tests

Next topic

CUDA envvar

CUDA Error Handling including timeouts

CUDA Driver API Errors

CUDA_ERROR_LAUNCH_FAILED = 700
An exception occurred on the device while executing a kernel. Common causes include dereferencing an invalid device pointer and accessing out of bounds shared memory. The context cannot be used, so it must be destroyed (and a new one should be created). All existing device memory allocations from this context are invalid and must be reconstructed if the program is to continue using CUDA.
CUDA_ERROR_LAUNCH_TIMEOUT = 702
This indicates that the device kernel took too long to execute. This can only occur if timeouts are enabled - see the device attribute CU_DEVICE_ATTRIBUTE_KERNEL_EXEC_TIMEOUT for more information. The context cannot be used (and must be destroyed similar to CUDA_ERROR_LAUNCH_FAILED). All existing device memory allocations from this context are invalid and must be reconstructed if the program is to continue using CUDA.
CU_DEVICE_ATTRIBUTE_KERNEL_EXEC_TIMEOUT = 17
Specifies whether there is a run time limit on kernels

deviceQuery

delta:w blyth$ cuda-samples-bin-deviceQuery | grep limit
  Run time limit on kernels:                     Yes