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

Apply .clangformat to all repo source files

Browse Source 
Change-Id: I7e79c6058f0303f9a98911e3b7dd2e8596079344


[ROCm/clr commit: 9e47fccc89]
This commit is contained in:
Maneesh Gupta
2018-03-12 11:29:03 +05:30
parent ecbb701440
commit 46ddefedee
293 changed files with 43980 additions and 45830 deletions
Download Patch File Download Diff File Expand all files Collapse all files
projects/clr/hipamd/tests/src/kernel/inline_asm_vadd.cpp
+33 -45
View File
@@ -1,19 +1,19 @@
/* Copyright (c) 2015-2016 Advanced Micro Devices, Inc. All rights reserved.
Permission is hereby granted, free of charge, to any person obtaining a copy of this software and
associated documentation files (the "Software"), to deal in the Software without restriction, including
without limitation the rights to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
copies of the Software, and to permit persons to whom the Software is furnished to do so, subject to the
following conditions:
associated documentation files (the "Software"), to deal in the Software without restriction,
including without limitation the rights to use, copy, modify, merge, publish, distribute,
sublicense, and/or sell copies of the Software, and to permit persons to whom the Software is
furnished to do so, subject to the following conditions:
The above copyright notice and this permission notice shall be included in all copies or substantial
portions of the Software.
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT NOT
LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO
EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY, WHETHER
IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR
THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR IMPLIED, INCLUDING BUT
NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY, FITNESS FOR A PARTICULAR PURPOSE AND
NONINFRINGEMENT. IN NO EVENT SHALL THE AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM,
DAMAGES OR OTHER LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM, OUT
OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
/* HIT_START
* BUILD: %t %s
@@ -22,46 +22,37 @@ THE USE OR OTHER DEALINGS IN THE SOFTWARE. */
*/
#include<iostream>
#include <iostream>
// hip header file
#include "hip/hip_runtime.h"
#define NUM 1024
#define NUM 1024
#define THREADS_PER_BLOCK_X 4
#define THREADS_PER_BLOCK_X 4
// Device (Kernel) function, it must be void
// hipLaunchParm provides the execution configuration
__global__ void vadd_asm(hipLaunchParm lp,
float *out,
float *in)
{
__global__ void vadd_asm(hipLaunchParm lp, float* out, float* in) {
int i = blockDim.x * blockIdx.x + threadIdx.x;
#ifdef __HIP_PLATFORM_NVCC__
asm volatile("add.f32 %0,%1,%2;":"=f"(out[i]):"f"(in[i]),"f"(out[i]));
asm volatile("add.f32 %0,%1,%2;" : "=f"(out[i]) : "f"(in[i]), "f"(out[i]));
#endif
#ifdef __HIP_PLATFORM_HCC__
asm volatile ("v_add_f32_e32 %0, %1, %2" : "=v" (out[i]) : "v"(in[i]),"v" (out[i]));
asm volatile("v_add_f32_e32 %0, %1, %2" : "=v"(out[i]) : "v"(in[i]), "v"(out[i]));
#endif
}
// CPU implementation of Vector Result
void addCPUReference(
float * output,
float * input)
{
for(unsigned int j=0; j < NUM; j++)
{
output[j]= input[j] + output[j];
void addCPUReference(float* output, float* input) {
for (unsigned int j = 0; j < NUM; j++) {
output[j] = input[j] + output[j];
}
}
int main(){
int main() {
float* VectorA;
float* ResultVector;
float* VectorB;
@@ -78,8 +69,8 @@ int main(){
// initialize the input data
for (i = 0; i < NUM; i++) {
VectorA[i] = (float)i*10.0f;
VectorB[i] = (float)i*30.0f;
VectorA[i] = (float)i * 10.0f;
VectorB[i] = (float)i * 30.0f;
}
// allocate the memory on the device side
@@ -87,18 +78,15 @@ int main(){
hipMalloc((void**)&gpuResultVector, NUM * sizeof(float));
// Memory transfer from host to device
hipMemcpy(gpuVector, VectorA, NUM*sizeof(float), hipMemcpyHostToDevice);
hipMemcpy(gpuResultVector, VectorB, NUM*sizeof(float), hipMemcpyHostToDevice);
hipMemcpy(gpuVector, VectorA, NUM * sizeof(float), hipMemcpyHostToDevice);
hipMemcpy(gpuResultVector, VectorB, NUM * sizeof(float), hipMemcpyHostToDevice);
// Lauching kernel from host
hipLaunchKernel(vadd_asm,
dim3(NUM/THREADS_PER_BLOCK_X),
dim3(THREADS_PER_BLOCK_X),
0, 0,
gpuResultVector , gpuVector);
hipLaunchKernel(vadd_asm, dim3(NUM / THREADS_PER_BLOCK_X), dim3(THREADS_PER_BLOCK_X), 0, 0,
gpuResultVector, gpuVector);
// Memory transfer from device to host
hipMemcpy(ResultVector, gpuResultVector, NUM*sizeof(float), hipMemcpyDeviceToHost);
hipMemcpy(ResultVector, gpuResultVector, NUM * sizeof(float), hipMemcpyDeviceToHost);
// CPU Result computation
addCPUReference(VectorB, VectorA);
@@ -107,23 +95,23 @@ int main(){
errors = 0;
double eps = 1.0E-3;
for (i = 0; i < NUM; i++) {
if (std::abs(ResultVector[i] - VectorB[i]) > eps ) {
errors++;
if (std::abs(ResultVector[i] - VectorB[i]) > eps) {
errors++;
}
}
if (errors!=0) {
printf("FAILED: %d errors\n",errors);
if (errors != 0) {
printf("FAILED: %d errors\n", errors);
} else {
printf ("PASSED!\n");
printf("PASSED!\n");
}
//free the resources on device side
// free the resources on device side
hipFree(gpuVector);
hipFree(gpuResultVector);
hipDeviceReset();
//free the resources on host side
// free the resources on host side
free(VectorA);
free(ResultVector);
free(VectorB);