fixing overloaded functions as macro
이 커밋은 다음에 포함됨:
1 부모
7b779f8409
커밋
9e2f37bf77
2 파일이 변경됨
+6
-6
검색과 일치하는 파일 없음
@@ -111,21 +111,21 @@ int main() {
|
||||
// correlation reagion33
|
||||
roctracer_activity_push_external_correlation_id(33);
|
||||
|
||||
roctxMarkA("before hipLaunchKernel");
|
||||
roctxRangePushA("hipLaunchKernel");
|
||||
roctxMark("before hipLaunchKernel");
|
||||
roctxRangePush("hipLaunchKernel");
|
||||
|
||||
// Lauching kernel from host
|
||||
hipLaunchKernel(matrixTranspose, dim3(WIDTH / THREADS_PER_BLOCK_X, WIDTH / THREADS_PER_BLOCK_Y),
|
||||
dim3(THREADS_PER_BLOCK_X, THREADS_PER_BLOCK_Y), 0, 0, gpuTransposeMatrix,
|
||||
gpuMatrix, WIDTH);
|
||||
|
||||
roctxMarkA("after hipLaunchKernel");
|
||||
roctxMark("after hipLaunchKernel");
|
||||
|
||||
// correlation reagion end
|
||||
roctracer_activity_pop_external_correlation_id(NULL);
|
||||
|
||||
// Memory transfer from device to host
|
||||
roctxRangePushA("hipMemcpy");
|
||||
roctxRangePush("hipMemcpy");
|
||||
|
||||
hipMemcpy(TransposeMatrix, gpuTransposeMatrix, NUM * sizeof(float), hipMemcpyDeviceToHost);
|
||||
|
||||
|
||||
새 이슈에서 참조
사용자 차단