Files
composable_kernel/include/ck/device_utility/kernel_launch.hpp
Chao Liu 4a27f120ea Absolute include path (#281)
* ad gelu and fast_gelu

* added GeLU and fast GeLU

* clean up

* add gemm+fastgelu example

* add gemm+gelu instances

* update profiler

* clean up

* clean up

* adding gemm+bias+activation

* clean

* adding bias

* clean

* adding gemm multiple d

* debugging

* add gemm bias add fastgelu

* rename, clean

* refactoring; add readme

* refactor

* refactor

* refactor

* refactor

* refactor

* refactor

* fix

* fix

* update example

* update example

* rename

* update example

* add ckProfiler

* clean

* clean

* clean

* clean

* add client app example

* update readme

* delete obselete files

* remove old client app

* delete old file

* cleaning

* clean

* remove half

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path for all examples

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* fix header path

* revert client app example

* clean build

* fix build

* temporary disable client test on Jenkins

* clean

* clean

* clean

[ROCm/composable_kernel commit: d1db6a0c3e]
2022-06-24 20:51:04 -05:00

72 lines
1.9 KiB
C++

#pragma once
#include <hip/hip_runtime.h>
#include "ck/ck.hpp"
#include "ck/stream_config.hpp"
#include "ck/device_utility/hip_check_error.hpp"
template <typename... Args, typename F>
float launch_and_time_kernel(const StreamConfig& stream_config,
F kernel,
dim3 grid_dim,
dim3 block_dim,
std::size_t lds_byte,
Args... args)
{
#if CK_TIME_KERNEL
if(stream_config.time_kernel_)
{
printf("%s: grid_dim {%d, %d, %d}, block_dim {%d, %d, %d} \n",
__func__,
grid_dim.x,
grid_dim.y,
grid_dim.z,
block_dim.x,
block_dim.y,
block_dim.z);
const int nrepeat = 10;
printf("Warm up 1 time\n");
// warm up
kernel<<<grid_dim, block_dim, lds_byte, stream_config.stream_id_>>>(args...);
printf("Start running %d times...\n", nrepeat);
hipEvent_t start, stop;
hip_check_error(hipEventCreate(&start));
hip_check_error(hipEventCreate(&stop));
hip_check_error(hipDeviceSynchronize());
hip_check_error(hipEventRecord(start, stream_config.stream_id_));
for(int i = 0; i < nrepeat; ++i)
{
kernel<<<grid_dim, block_dim, lds_byte, stream_config.stream_id_>>>(args...);
}
hip_check_error(hipEventRecord(stop, stream_config.stream_id_));
hip_check_error(hipEventSynchronize(stop));
float total_time = 0;
hip_check_error(hipEventElapsedTime(&total_time, start, stop));
return total_time / nrepeat;
}
else
{
kernel<<<grid_dim, block_dim, lds_byte, stream_config.stream_id_>>>(args...);
return 0;
}
#else
kernel<<<grid_dim, block_dim, lds_byte, stream_config.stream_id_>>>(args...);
return 0;
#endif
}