Seems thrust::sort bug with device_vector

Accelerated Computing CUDA CUDA Programming and Performance
1

I encountered a tricky bug these days.

The bug comes when pytorch ungrade from 1.7 to 1.10.

The bug can be reproduced when calling

torch.mode(torch.from_numpy(np.array([1,111,1,0,1,0,1444,1,10,4]*600)).to(torch.int64).cuda(0), 0)

then do thrust::sort will raise RuntimeError: radix_sort: failed on 1st step: cudaErrorInvalidDevice: invalid device ordinal.

code used for testing:
int MAX_N = 10000;
thrust::host_vector h(MAX_N);
for (int i = 0; i < MAX_N; i++) {
h[i] = rand() % 998;
}
thrust::device_vector d = h;
thrust::sort(h.begin(), h.end());
thrust::sort(d.begin(), d.end());

sort with host_vector works always, but device_vector does not work.

2

I suggest asking pytorch questions on the pytorch forum. Regarding your thrust code, that is not valid thrust code, so I’m not surprised it doesn’t work. It won’t compile, for starters. If we fix the compilation issues, the thrust code works fine for me.