echo/CMake
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Tests: Simplify CUDA kernels avoid crash with nvidia driver 550

Browse Source 
Avoid having two identical kernels in PTX.  We don't need a second
kernel anyway.
This commit is contained in:
Robert Maynard 2024-10-17 14:45:57 -04:00 committed by Brad King
parent cb981c3b00
commit ee9e2216ec
1 changed files with 0 additions and 7 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

7
Tests/CudaOnly/SeparateCompilationPTX/kernels.cu
View File

@ -5,10 +5,3 @@ __global__ void kernelA(float* r, float* x, float* y, float* z, int size)
r[i] = x[i] * y[i] + z[i];
}
}
__global__ void kernelB(float* r, float* x, float* y, float* z, int size)
{
for (int i = threadIdx.x; i < size; i += blockDim.x) {
r[i] = x[i] * y[i] + z[i];
}
}