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


[ROCm/hip-tests commit: bfceb14751]
This commit is contained in:
Maneesh Gupta
2018-10-17 12:01:44 +05:30
parent 1162ec5c3e
commit 1bc2b62f98
23 changed files with 56 additions and 69 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/hip-tests/samples/2_Cookbook/6_dynamic_shared/Readme.md
+1 -1
View File
@@ -25,7 +25,7 @@ Shared memory is way more faster than that of global and constant memory and acc
here the first parameter is the data type while the second one is the variable name.
The other important change is:
` hipLaunchKernel(matrixTranspose, `
` hipLaunchKernelGGL(matrixTranspose, `
dim3(WIDTH/THREADS_PER_BLOCK_X, WIDTH/THREADS_PER_BLOCK_Y),
dim3(THREADS_PER_BLOCK_X, THREADS_PER_BLOCK_Y),
sizeof(float)*WIDTH*WIDTH, 0,
projects/hip-tests/samples/2_Cookbook/6_dynamic_shared/dynamic_shared.cpp
+2 -3
View File
@@ -33,8 +33,7 @@ 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) {
// declare dynamic shared memory
HIP_DYNAMIC_SHARED(float, sharedMem);
@@ -90,7 +89,7 @@ int main() {
hipMemcpy(gpuMatrix, Matrix, NUM * sizeof(float), hipMemcpyHostToDevice);
// Lauching kernel from host
hipLaunchKernel(matrixTranspose, dim3(WIDTH / THREADS_PER_BLOCK_X, WIDTH / THREADS_PER_BLOCK_Y),
hipLaunchKernelGGL(matrixTranspose, dim3(WIDTH / THREADS_PER_BLOCK_X, WIDTH / THREADS_PER_BLOCK_Y),
dim3(THREADS_PER_BLOCK_X, THREADS_PER_BLOCK_Y), sizeof(float) * WIDTH * WIDTH,
0, gpuTransposeMatrix, gpuMatrix, WIDTH);