AI/rocm-systems
2
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

Replace hipLaunchKernel -> hipLaunchKernelGGL

Browse Source 
Change-Id: I4d99009e1199811d417becf1e1b934ec4d4e30be
This commit is contained in:
Maneesh Gupta
2018-10-17 12:01:44 +05:30
parent 9e167ab02e
commit bfceb14751
23 changed files with 56 additions and 69 deletions
Download Patch File Download Diff File Expand all files Collapse all files
samples/2_Cookbook/9_unroll/unroll.cpp
+2 -3
View File
@@ -35,8 +35,7 @@ THE SOFTWARE.
#define THREADS_PER_BLOCK_Z 1
// Device (Kernel) function, it must be void
// hipLaunchParm provides the execution configuration
__global__ void matrixTranspose(hipLaunchParm lp, float* out, float* in, const int width) {
__global__ void matrixTranspose(float* out, float* in, const int width) {
int x = hipBlockDim_x * hipBlockIdx_x + hipThreadIdx_x;
float val = in[x];
@@ -88,7 +87,7 @@ int main() {
hipMemcpy(gpuMatrix, Matrix, NUM * sizeof(float), hipMemcpyHostToDevice);
// Lauching kernel from host
hipLaunchKernel(matrixTranspose, dim3(1), dim3(THREADS_PER_BLOCK_X * THREADS_PER_BLOCK_Y), 0, 0,
hipLaunchKernelGGL(matrixTranspose, dim3(1), dim3(THREADS_PER_BLOCK_X * THREADS_PER_BLOCK_Y), 0, 0,
gpuTransposeMatrix, gpuMatrix, WIDTH);
// Memory transfer from device to host