ROCm/composable_kernel
1
0
Fork 0
mirror of https://github.com/ROCm/composable_kernel.git synced 2026-04-19 22:39:03 +00:00
Code Issues Packages Projects Releases Wiki Activity

Add host API (#220)

Browse Source 
* Add host API

* manually rebase on develop

* clean

* manually rebase on develop

* exclude tests from all target

* address review comments

* update client app name

* fix missing lib name

* clang-format update

* refactor

* refactor

* refactor

* refactor

* refactor

* fix test issue

* refactor

* refactor

* refactor

* upate cmake and readme

Co-authored-by: Chao Liu <chao.liu2@amd.com>
This commit is contained in:
JD
2022-05-12 09:21:01 -05:00
committed by GitHub
parent 0f912e205e
commit cec69bc3bc
131 changed files with 1666 additions and 1089 deletions
Download Patch File Download Diff File Expand all files Collapse all files

13
example/12_reduce/reduce_blockwise.cpp
View File

@@ -116,10 +116,9 @@ class SimpleAppArgs
std::vector<size_t> inLengths;
std::vector<float> scales;
bool do_verification = false;
int init_method = 1;
int nrepeat = 5;
bool do_verification = true;
int init_method = 1;
bool time_kernel = false;
public:
void show_usage(const char* cmd)
@@ -135,7 +134,7 @@ class SimpleAppArgs
std::cout << "Arg1 -- init method (0=no init, 1=single integer value, 2=scope integer "
"value, 3=decimal value)"
<< std::endl;
std::cout << "Arg2 -- number of repeats to run the kernel" << std::endl;
std::cout << "Arg2 -- time kernel (0=n0, 1=yes)" << std::endl;
};
int processArgs(int argc, char* argv[])
@@ -182,7 +181,7 @@ class SimpleAppArgs
throw std::runtime_error("Invalid cmd-line arguments, more argumetns are needed!");
init_method = std::atoi(argv[optind++]);
nrepeat = std::atoi(argv[optind]);
time_kernel = std::atoi(argv[optind]);
if(scales.empty())
{
@@ -352,7 +351,7 @@ int main(int argc, char* argv[])
auto invoker_ptr = reduce.MakeInvokerPointer();
float avg_time = invoker_ptr->Run(argument_ptr.get(), args.nrepeat);
float avg_time = invoker_ptr->Run(argument_ptr.get(), StreamConfig{nullptr, args.time_kernel});
std::size_t num_bytes = invariant_total_length * reduce_total_length * sizeof(InDataType) +
invariant_total_length * sizeof(OutDataType);