gaspersic/triton
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

[TESTS] Updated the test to be compatible with the new runtime API

Browse Source
This commit is contained in:
Philippe Tillet
2020-08-12 19:37:08 -04:00
committed by Philippe Tillet
parent acff1b5e05
commit 150ba0c70b
6 changed files with 88 additions and 61 deletions
Download Patch File Download Diff File Expand all files Collapse all files

40
tests/common/conv.h
View File

@@ -12,6 +12,32 @@
#include "util.h" #include "util.h"
struct conv_arg_t{
CUdeviceptr a;
CUdeviceptr b;
CUdeviceptr c;
float alpha;
int M;
int N;
int K;
int pad_h;
int pad_w;
int stride_h;
int stride_w;
CUdeviceptr adelta;
int lda_z;
int lda_ci;
int lda_h;
int lda_w;
int ldb_ci;
int ldb_r;
int ldb_s;
int ldb_co;
int ldc_z;
int ldc_co;
int ldc_p;
int ldc_q;
};
enum run_mode_t { enum run_mode_t {
BENCH, BENCH,
@@ -93,19 +119,19 @@ void triton_conv(drv::stream* stream,
// kernels // kernels
rt::function function(src::conv, opt); rt::function function(src::conv, opt);
std::vector<rt::arg> args = {&*da, &*db, &*dc, (float)1, Z*P*Q, CO, CI*R*S, conv_arg_t args{*da->cu(), *db->cu(), *dc->cu(), 1, Z*P*Q, CO, CI*R*S,
pad_h, pad_w, stride_h, stride_w, pad_h, pad_w, stride_h, stride_w,
&*ddelta, *ddelta->cu(),
W*H*CI, W*H, W, 1, W*H*CI, W*H, W, 1,
CO*S*R , CO*S, CO, 1, CO*S*R , CO*S, CO, 1,
Q*P*CO, Q*P, Q, 1}; Q*P*CO, Q*P, Q, 1};
auto grid = [Z,P,Q,CO](const rt::function::options_t& x) { auto grid = [Z,P,Q,CO](const rt::function::options_t& x) {
return rt::grid_t{ceil(Z*P*Q, x.D<int>("TM")), return rt::grid_t{ceil(Z*P*Q, x.D<int>("TM")),
ceil(CO , x.D<int>("TN")), ceil(CO , x.D<int>("TN")),
(size_t)x.D<int>("TZ")}; (size_t)x.D<int>("TZ")};
}; };
auto tflops = [&](double nanosec) { return 2.*Z*P*Q*CI*CO*R*S / nanosec * 1e-3; }; auto tflops = [&](double nanosec) { return 2.*Z*P*Q*CI*CO*R*S / nanosec * 1e-3; };
double triton_ns = triton::tools::bench([&]() { function(args, grid, stream);}, stream); double triton_ns = triton::tools::bench([&]() { function((void**)&args, sizeof(args), grid, stream);}, stream);
bench.push_back(tflops(triton_ns)); bench.push_back(tflops(triton_ns));
} }

19
tests/common/copy.h
View File

@@ -12,6 +12,15 @@ int32_t off(const std::vector<int32_t>& idx, const std::vector<int32_t>& strides
return res; return res;
} }
struct copy_arg_t{
CUdeviceptr X;
CUdeviceptr Y;
int S0;
int S1;
int S2;
};
enum run_mode_t { enum run_mode_t {
BENCH, BENCH,
TEST TEST
@@ -115,16 +124,16 @@ void triton_copy_nd(drv::stream* stream, const std::vector<int32_t>& shape,
// kernel // kernel
rt::function function(src::copy_nd[rank - 1], opt); rt::function function(src::copy_nd[rank - 1], opt);
std::vector<rt::arg> args = {&*dx, &*dy}; copy_arg_t args = {*dx->cu(), *dy->cu(), shape[0]};
for(int32_t d: shape) if(shape.size() > 1) args.S1 = shape[1];
args.push_back(d); if(shape.size() > 2) args.S2 = shape[2];
std::vector<std::string> ts = {"TS0", "TS1", "TS2"}; std::vector<std::string> ts = {"TS0", "TS1", "TS2"};
auto grid = grid_nd(shape, ts); auto grid = grid_nd(shape, ts);
// metrics // metrics
if(mode == BENCH){ if(mode == BENCH){
auto gbps = [&](double ns) { return 2 * size * dtsize / (ns * 1e-9) * 1e-9; }; auto gbps = [&](double ns) { return 2 * size * dtsize / (ns * 1e-9) * 1e-9; };
double triton_ns = triton::tools::bench([&]() { function(args, grid, stream);}, stream); double triton_ns = triton::tools::bench([&]() { function((void**)&args, sizeof(args), grid, stream);}, stream);
bench.push_back(gbps(triton_ns)); bench.push_back(gbps(triton_ns));
} }
@@ -136,7 +145,7 @@ void triton_copy_nd(drv::stream* stream, const std::vector<int32_t>& shape,
for(size_t i = 0; i < hx.size(); i++) for(size_t i = 0; i < hx.size(); i++)
hx[i] = static_cast<T>((float)rand()/RAND_MAX); hx[i] = static_cast<T>((float)rand()/RAND_MAX);
stream->write(&*dx, true, 0, hx); stream->write(&*dx, true, 0, hx);
function(args, grid, stream); function((void**)&args, sizeof(args), grid, stream);
stream->synchronize(); stream->synchronize();
stream->read(&*dy, true, 0, hy); stream->read(&*dy, true, 0, hy);
cc_copy_nd(hx, ry, shape, x_order, y_order); cc_copy_nd(hx, ry, shape, x_order, y_order);

53
tests/common/dot.h
View File

@@ -13,23 +13,19 @@
#include "util.h" #include "util.h"
//struct dot_arg_t{ struct dot_arg_t{
// CUdeviceptr a; CUdeviceptr a;
// CUdeviceptr b; CUdeviceptr b;
// CUdeviceptr c; CUdeviceptr c;
// float alpha; float alpha;
// int M; int M;
// int N; int N;
// int K; int K;
// int lda; int lda;
// int ldb; int ldb;
// int ldc; int ldc;
// CUdeviceptr locks; CUdeviceptr locks;
//}; };
//typedef std::tuple<CUdeviceptr, CUdeviceptr, CUdeviceptr,
// float, int, int, int, int, int, int,
// CUdeviceptr> dot_arg_t;
template<class T, bool AT, bool BT> template<class T, bool AT, bool BT>
static void cc_dot(std::vector<T> &c, const std::vector<T> &a, const std::vector<T> &b, static void cc_dot(std::vector<T> &c, const std::vector<T> &a, const std::vector<T> &b,
@@ -140,24 +136,9 @@ void triton_dot(drv::stream* stream, bool AT, bool BT,
// kernels // kernels
rt::function function(src::dot, opt); rt::function function(src::dot, opt);
float alpha = 1; dot_arg_t args = {*da->cu(), *db->cu(), *dc->cu(),
char args[60]; 1, M, N, K, lda, ldb, ldc, *dlocks->cu()};
memcpy(args + 0, &*da->cu(), 8);
memcpy(args + 8, &*db->cu(), 8);
memcpy(args + 16, &*dc->cu(), 8);
memcpy(args + 24, &alpha, 4);
memcpy(args + 28, &M, 4);
memcpy(args + 32, &N, 4);
memcpy(args + 36, &K, 4);
memcpy(args + 40, &lda, 4);
memcpy(args + 44, &ldb, 4);
memcpy(args + 48, &ldc, 4);
memcpy(args + 52, &*dlocks->cu(), 8);
// dot_arg_t args = {*da->cu(), *db->cu(), *dc->cu(),
// 1, M, N, K, lda, ldb, ldc, *dlocks->cu()};
// std::cout << sizeof(dot_arg_t) << std::endl;
auto grid = [M, N](const rt::function::options_t& x) { auto grid = [M, N](const rt::function::options_t& x) {
return rt::grid_t{ceil(M, x.D<int>("TM")), return rt::grid_t{ceil(M, x.D<int>("TM")),
ceil(N, x.D<int>("TN")), ceil(N, x.D<int>("TN")),
@@ -167,7 +148,7 @@ void triton_dot(drv::stream* stream, bool AT, bool BT,
// metrics // metrics
if(mode == BENCH){ if(mode == BENCH){
auto tflops = [&](double nanosec) { return 2.*M*N*K / nanosec * 1e-3; }; auto tflops = [&](double nanosec) { return 2.*M*N*K / nanosec * 1e-3; };
double triton_ns = triton::tools::bench([&]() { function((void**)&args, grid, stream);}, stream); double triton_ns = triton::tools::bench([&]() { function((void**)&args, sizeof(args), grid, stream);}, stream);
bench.push_back(tflops(triton_ns)); bench.push_back(tflops(triton_ns));
// cublas // cublas
@@ -198,7 +179,7 @@ void triton_dot(drv::stream* stream, bool AT, bool BT,
stream->write(&*da, true, 0, ha); stream->write(&*da, true, 0, ha);
stream->write(&*db, true, 0, hb); stream->write(&*db, true, 0, hb);
// run kernel // run kernel
function((void**)&args, grid, stream); function((void**)&args, sizeof(args), grid, stream);
// write back // write back
stream->synchronize(); stream->synchronize();
// compare with CPU // compare with CPU

19
tests/common/reduce.h
View File

@@ -13,6 +13,15 @@
namespace drv = triton::driver; namespace drv = triton::driver;
namespace rt = triton::runtime; namespace rt = triton::runtime;
struct reduce_arg_t{
CUdeviceptr X;
CUdeviceptr Y;
int S0;
int S1;
int S2;
};
template<class T> template<class T>
void cc_reduce_nd(std::vector<T> &y, const std::vector<T> &x, reduce_op_t op, size_t axis, const std::vector<int>& shapes) { void cc_reduce_nd(std::vector<T> &y, const std::vector<T> &x, reduce_op_t op, size_t axis, const std::vector<int>& shapes) {
assert(axis <= shapes.size() - 1); assert(axis <= shapes.size() - 1);
@@ -123,16 +132,16 @@ void triton_reduce_nd(drv::stream* stream, const std::vector<int32_t>& shape_x,
auto dy = std::unique_ptr<drv::buffer>(drv::buffer::create(context, size_y*dtsize)); auto dy = std::unique_ptr<drv::buffer>(drv::buffer::create(context, size_y*dtsize));
// grid // grid
std::vector<rt::arg> args = {&*dx, &*dy}; reduce_arg_t args = {*dx->cu(), *dy->cu(), shape_x[0]};
for(int32_t d: shape_x) if(shape_x.size() > 1) args.S1 = shape_x[1];
args.push_back(d); if(shape_x.size() > 2) args.S2 = shape_x[2];
std::vector<std::string> ts = {"TS0", "TS1", "TS2"}; std::vector<std::string> ts = {"TS0", "TS1", "TS2"};
auto grid = grid_nd(shape_x, ts); auto grid = grid_nd(shape_x, ts);
// metrics // metrics
if(mode == BENCH){ if(mode == BENCH){
auto gbps = [&](double ns) { return 2 * size_x * dtsize / (ns * 1e-9) * 1e-9; }; auto gbps = [&](double ns) { return 2 * size_x * dtsize / (ns * 1e-9) * 1e-9; };
double triton_ns = triton::tools::bench([&]() { function(args, grid, stream);}, stream); double triton_ns = triton::tools::bench([&]() { function((void**)&args, sizeof(args), grid, stream);}, stream);
bench.push_back(gbps(triton_ns)); bench.push_back(gbps(triton_ns));
} }
@@ -144,7 +153,7 @@ void triton_reduce_nd(drv::stream* stream, const std::vector<int32_t>& shape_x,
init_zeros(hy); init_zeros(hy);
init_rand(hx); init_rand(hx);
stream->write(&*dx, true, 0, hx); stream->write(&*dx, true, 0, hx);
function(args, grid, stream); function((void**)&args, sizeof(args), grid, stream);
stream->synchronize(); stream->synchronize();
stream->read(&*dy, true, 0, hy); stream->read(&*dy, true, 0, hy);
cc_reduce_nd(ry, hx, op, axis, shape_x); cc_reduce_nd(ry, hx, op, axis, shape_x);

6
tests/common/src/conv.h
View File

@@ -7,7 +7,9 @@ R"(
TYPE *C __noalias __aligned(16), TYPE *C __noalias __aligned(16),
float alpha, float alpha,
// equivalent matmul // equivalent matmul
int M, int N, int K, int M __retune,
int N __retune,
int K __retune,
// convolution properties // convolution properties
int pad_h, int pad_w, int stride_h, int stride_w, int pad_h, int pad_w, int stride_h, int stride_w,
// pointer increment // pointer increment
@@ -16,7 +18,7 @@ R"(
int lda_z __multipleof(8), int lda_ci __multipleof(8), int lda_h __multipleof(8), int lda_w __multipleof(8), int lda_z __multipleof(8), int lda_ci __multipleof(8), int lda_h __multipleof(8), int lda_w __multipleof(8),
int ldb_ci __multipleof(8), int ldb_r __multipleof(8), int ldb_s __multipleof(8), int ldb_co __multipleof(8), int ldb_ci __multipleof(8), int ldb_r __multipleof(8), int ldb_s __multipleof(8), int ldb_co __multipleof(8),
int ldc_z __multipleof(8), int ldc_co __multipleof(8), int ldc_p __multipleof(8), int ldc_q __multipleof(8)) { int ldc_z __multipleof(8), int ldc_co __multipleof(8), int ldc_p __multipleof(8), int ldc_q __multipleof(8)) {
// prologue // prologue
int ridx = get_program_id(0); int ridx = get_program_id(0);
int ridy = get_program_id(1); int ridy = get_program_id(1);
int ridz = get_program_id(2); int ridz = get_program_id(2);

12
tests/common/src/copy.h
View File

@@ -7,7 +7,7 @@ namespace src {
R"( R"(
void copy1d(TYPE * X __noalias __readonly __aligned(16), void copy1d(TYPE * X __noalias __readonly __aligned(16),
TYPE * Y __noalias __readonly __aligned(16), TYPE * Y __noalias __readonly __aligned(16),
int S0) { int S0 __retune) {
int pid0 = get_program_id(0); int pid0 = get_program_id(0);
int rs0[TS0] = pid0 * TS0 + 0 ... TS0; int rs0[TS0] = pid0 * TS0 + 0 ... TS0;
TYPE* px[TS0] = X + rs0; TYPE* px[TS0] = X + rs0;
@@ -20,8 +20,8 @@ void copy1d(TYPE * X __noalias __readonly __aligned(16),
R"( R"(
void copy2d(TYPE * X __noalias __readonly __aligned(16), void copy2d(TYPE * X __noalias __readonly __aligned(16),
TYPE * Y __noalias __writeonly __aligned(16), TYPE * Y __noalias __writeonly __aligned(16),
int S0 __multipleof(8), int S0 __multipleof(8) __retune,
int S1 __multipleof(8)) { int S1 __multipleof(8) __retune) {
int pid0 = get_program_id(0); int pid0 = get_program_id(0);
int pid1 = get_program_id(1); int pid1 = get_program_id(1);
int rs0[TS0] = pid0 * TS0 + 0 ... TS0; int rs0[TS0] = pid0 * TS0 + 0 ... TS0;
@@ -37,9 +37,9 @@ void copy2d(TYPE * X __noalias __readonly __aligned(16),
R"( R"(
void copy3d(TYPE * X __noalias __readonly __aligned(16), void copy3d(TYPE * X __noalias __readonly __aligned(16),
TYPE * Y __noalias __writeonly __aligned(16), TYPE * Y __noalias __writeonly __aligned(16),
int S0 __multipleof(8), int S0 __multipleof(8) __retune,
int S1 __multipleof(8), int S1 __multipleof(8) __retune,
int S2 __multipleof(8)) { int S2 __multipleof(8) __retune) {
// program id // program id
int pid0 = get_program_id(0); int pid0 = get_program_id(0);
int pid1 = get_program_id(1); int pid1 = get_program_id(1);