Internal Operation Error from cublasDgelsBatched

Accelerated Computing GPU-Accelerated Libraries
1

I’m getting internal operation failed from a call to cublasDgelsBatched. I’ve been able to determine the error is coming from a call to geqrSolve_batch_kernel, but haven’t been able to get any further. Any suggestions on how to figure out what’s going on?

2

It’s helpful to provide reproducers. You can also try using the logger functionality to retrieve more information.

3

Here’s the code and logger output.

Build:
nvcc -o gelscu gels_cublas.cu -lcublas -lcurand -lcudart
Code:
gels_cublas.cu (2.5 KB)

Output:
gelscu.output.txt (1.9 KB)