dkrottap
4e1a3ff850
Migrated all module related files to CATCH2 framework and optimized to have single module kernel file Change-Id: I39aa28ef22c1b2f4d0014ca32b59b9c645b725dc
122 baris
4.0 KiB
C++
122 baris
4.0 KiB
C++
/*
|
|
Copyright (c) 2021 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:
|
|
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.
|
|
*/
|
|
|
|
/*
|
|
This testcase verifies hipModuleLoad API in multithreaded scenario
|
|
*/
|
|
#include <stdio.h>
|
|
#include "hip/hip_runtime.h"
|
|
#if HT_AMD
|
|
#include "hip/hip_ext.h"
|
|
#endif
|
|
#include <fstream>
|
|
#include <algorithm>
|
|
#include <atomic>
|
|
#include <functional>
|
|
#include <vector>
|
|
#include <future>
|
|
#define THREADS 8
|
|
#define MAX_NUM_THREADS 128
|
|
|
|
#include "hip_test_common.hh"
|
|
#include "hip_test_checkers.hh"
|
|
|
|
#define NUM_GROUPS 1
|
|
#define GROUP_SIZE 1
|
|
#define WARMUP_RUN_COUNT 10
|
|
#define TIMING_RUN_COUNT 100
|
|
#define TOTAL_RUN_COUNT WARMUP_RUN_COUNT + TIMING_RUN_COUNT
|
|
#define FILENAME "module_kernels.code"
|
|
#define kernel_name "EmptyKernel"
|
|
|
|
/*
|
|
This thread function loads the kernel file , synchronizes the threads
|
|
and Launches the kernel .
|
|
*/
|
|
void hipModuleLaunchKernel_enqueue(std::atomic_int* shared, int max_threads) {
|
|
// resources necessary for this thread
|
|
hipStream_t stream;
|
|
HIPCHECK(hipStreamCreate(&stream));
|
|
hipModule_t module;
|
|
hipFunction_t function;
|
|
|
|
HIPCHECK(hipModuleLoad(&module, FILENAME));
|
|
HIPCHECK(hipModuleGetFunction(&function, module, kernel_name));
|
|
|
|
void* kernel_params = nullptr;
|
|
|
|
// synchronize all threads, before running
|
|
shared->fetch_add(1, std::memory_order_release);
|
|
while (max_threads != shared->load(std::memory_order_acquire)) {}
|
|
|
|
for (auto i = 0; i < TOTAL_RUN_COUNT; ++i) {
|
|
HIPCHECK(hipModuleLaunchKernel(function, 1, 1,
|
|
1, 1, 1, 1, 0, stream,
|
|
&kernel_params, nullptr));
|
|
}
|
|
HIPCHECK(hipModuleUnload(module));
|
|
HIPCHECK(hipStreamDestroy(stream));
|
|
}
|
|
|
|
/*
|
|
thread pool class contains launching the threads using std::async API
|
|
with future variable "threads".
|
|
The start API Launches the threads and finish API waits for the
|
|
thread execution to end.
|
|
*/
|
|
struct thread_pool {
|
|
explicit thread_pool(int total_threads) : max_threads(total_threads) {
|
|
}
|
|
void start(std::function<void(std::atomic_int*, int)> f) {
|
|
for (int i = 0; i < max_threads; ++i) {
|
|
threads.push_back(std::async(std::launch::async, f,
|
|
&shared, max_threads));
|
|
}
|
|
}
|
|
void finish() {
|
|
for (auto&&thread : threads) {
|
|
thread.get();
|
|
}
|
|
threads.clear();
|
|
shared = 0;
|
|
}
|
|
~thread_pool() {
|
|
finish();
|
|
}
|
|
private:
|
|
std::atomic_int shared {0};
|
|
std::vector<char> buffer;
|
|
std::vector<std::future<void>> threads;
|
|
int max_threads = 1;
|
|
};
|
|
|
|
/*
|
|
This testcase verifies the Multithreaded scenario of hipModule API
|
|
where in threadpool object is created and the object invokes start API
|
|
which launches multiple threads where each thread loads the kernel object
|
|
using hipModuleLoad API and launches the kernel in parallel.
|
|
*/
|
|
TEST_CASE("Unit_hipModuleLoad_MultiThread") {
|
|
int max_threads = min(THREADS * std::thread::hardware_concurrency(),
|
|
MAX_NUM_THREADS);
|
|
thread_pool task(max_threads);
|
|
task.start(hipModuleLaunchKernel_enqueue);
|
|
task.finish();
|
|
}
|