Skip to content

GitLab

Commit 6ec90d65 authored by Shucai Xiao's avatar Shucai Xiao
Browse files

code backup after testing a few scenarios.

parent b9d45e76
Hide whitespace changes
Inline Side-by-side
Showing with 101 additions and 48 deletions
src/include/migraphx/operators.hpp
View file @ 6ec90d65
......@@ -894,7 +894,7 @@ struct dot
shape compute_shape(std::vector<shape> inputs) const
{
// If there are 3 inputs, then A and B must be matrices and
// C is broadcastable to A * B
// C should be broadcastable to A * B
if(inputs.size() == 3)
{
check_shapes{inputs, *this}.has(3).same_type();
......@@ -911,13 +911,14 @@ struct dot
}
auto out_lens = a_lens;
out_lens[0] = b_lens[0];
out_lens[1] = b_lens[1];
// check whether C is broadcastable to A * B
auto c_lens = inputs[2].lens();
if(c_lens.size() > 2 ||
(c_lens.size() >= 1 && (c_lens[0] != 1 && c_lens[0] != b_lens[0])) ||
(c_lens.size() == 2 && (c_lens[1] != 1 && c_lens[1] != a_lens[1])))
(c_lens.size() == 1 && (c_lens[0] != 1 && c_lens[0] != b_lens[1])) ||
(c_lens.size() == 2 && (c_lens[0] != 1 && c_lens[0] != a_lens[0])) ||
(c_lens.size() == 2 && (c_lens[1] != 1 && c_lens[1] != b_lens[1])))
{
MIGRAPHX_THROW("DOT: C {" + to_string_range(c_lens) +
"} is not broadcastable to A * B {" + to_string_range(out_lens) +
......@@ -955,9 +956,9 @@ struct dot
is_b_appended = true;
}
std::size_t dim_0 = a_lens.size() - 1;
std::size_t dim_1 = b_lens.size() - 2;
if(a_lens[dim_0] != b_lens[dim_1])
std::size_t dim_0 = a_lens.size() - 2;
std::size_t dim_1 = b_lens.size() - 1;
if(a_lens[dim_1] != b_lens[dim_0])
{
MIGRAPHX_THROW("DOT : dimension mismatch, operand A: {" + to_string_range(a.lens()) +
"}, cannot multiply operand B: {" + to_string_range(b.lens()) + "}");
......
src/targets/gpu/gemm.cpp
View file @ 6ec90d65
......@@ -168,7 +168,8 @@ rocblas_half to_rocblas_type(half x) { return reinterpret_cast<const rocblas_hal
shape miopen_gemm::compute_shape(const std::vector<shape>& inputs) const
{
return op.compute_shape(inputs);
std::vector<shape> orig_inputs(inputs.begin(), inputs.begin() + inputs.size() - 1);
return op.compute_shape(orig_inputs);
}
void miopen_gemm::fill_result(const shape& output_shape,
......@@ -177,6 +178,7 @@ void miopen_gemm::fill_result(const shape& output_shape,
{
auto out_lens = output_shape.lens();
auto c_lens = c.get_shape().lens();
auto type_size = output_shape.type_size();
if(output_shape == c.get_shape())
{
output_shape.visit_type([&](auto as) {
......@@ -188,13 +190,13 @@ void miopen_gemm::fill_result(const shape& output_shape,
else if(c.single())
{
output_shape.visit_type([&](auto as) {
auto to_pointer = [&](auto&& arg, std::size_t offset = 0) {
return to_rocblas_type(as.from(arg.data() + offset));
auto to_pointer = [&](auto&& arg, std::size_t offset_byte = 0) {
return to_rocblas_type(as.from(arg.data() + offset_byte));
};
for(std::size_t i = 0; i < output_shape.elements(); ++i)
{
hipMemcpy(to_pointer(result, i),
hipMemcpy(to_pointer(result, i * type_size),
to_pointer(c),
c.get_shape().bytes(),
hipMemcpyDeviceToDevice);
......@@ -212,7 +214,7 @@ void miopen_gemm::fill_result(const shape& output_shape,
for(std::size_t i = 0; i < m; ++i)
{
hipMemcpy(to_pointer(result, i * n),
hipMemcpy(to_pointer(result, i * n * type_size),
to_pointer(c),
c.get_shape().bytes(),
hipMemcpyDeviceToDevice);
......@@ -229,9 +231,9 @@ void miopen_gemm::fill_result(const shape& output_shape,
for(std::size_t i = 0; i < output_shape.elements(); ++i)
{
hipMemcpy(to_pointer(result, i),
to_pointer(c, i / out_lens[0]),
c.get_shape().type_size(),
hipMemcpy(to_pointer(result, i * type_size),
to_pointer(c, i / out_lens[1] * type_size),
type_size,
hipMemcpyDeviceToDevice);
}
});
......@@ -254,12 +256,28 @@ argument miopen_gemm::compute(context& ctx,
bool transb = args[1].get_shape().transposed();
rocblas_int lda = args[0].get_shape().strides()[transa ? 1 : 0];
rocblas_int ldb = args[1].get_shape().strides()[transb ? 1 : 0];
rocblas_int ldc = args[2].get_shape().strides()[0];
rocblas_int ldc = args[3].get_shape().strides()[0];
auto out_lens = output_shape.lens();
rocblas_int m = out_lens[0];
rocblas_int n = out_lens[1];
rocblas_int k = args[0].get_shape().lens()[1];
auto to_pointer = [&](auto&& arg) { return to_rocblas_type(as.from(arg.data())); };
auto cpu_a = migraphx::gpu::from_gpu(args[0]);
auto cpu_b = migraphx::gpu::from_gpu(args[1]);
auto cpu_res = migraphx::gpu::from_gpu(args[3]);
std::cout << "gpu::gemm, cpu_a = " << cpu_a << std::endl;
std::cout << "gpu::gemm, cpu_b = " << cpu_b << std::endl;
std::cout << "gpu::gemm, cpu_res = " << cpu_res << std::endl;
std::cout << "gpu::gemm, transb = " << transb << std::endl;
std::cout << "gpu::gemm, transa = " << transb << std::endl;
std::cout << "gpu::gemm, m = " << m << std::endl;
std::cout << "gpu::gemm, n = " << n << std::endl;
std::cout << "gpu::gemm, k = " << k << std::endl;
std::cout << "gpu::gemm, lda = " << lda << std::endl;
std::cout << "gpu::gemm, ldb = " << ldb << std::endl;
std::cout << "gpu::gemm, ldc = " << ldc << std::endl;
generic_rocblas_gemm(as,
ctx.get_stream().get_rocblas(),
transb ? rocblas_operation_transpose : rocblas_operation_none,
......@@ -273,7 +291,7 @@ argument miopen_gemm::compute(context& ctx,
to_pointer(args[0]),
lda,
&beta_r,
to_pointer(args[2]),
to_pointer(args[3]),
ldc);
});
......@@ -283,6 +301,7 @@ argument miopen_gemm::compute(context& ctx,
// 2 input arguments cases
// vector inner product
std::size_t type_size = output_shape.type_size();
if(output_shape.elements() == 1)
{
assert(args[0].get_shape().elements() == args[1].get_shape().elements());
......@@ -331,12 +350,12 @@ argument miopen_gemm::compute(context& ctx,
m,
n,
&alpha_r,
to_pointer(args[0], batch_no * m * n),
to_pointer(args[0], batch_no * m * n * type_size),
lda,
to_pointer(args[1]),
1,
&beta_r,
to_pointer(args[2], batch_no * n),
to_pointer(args[2], batch_no * n * type_size),
1);
}
});
......@@ -344,41 +363,74 @@ argument miopen_gemm::compute(context& ctx,
// vector * matrix
else if(args[0].get_shape().lens().size() == 1)
{
auto a_lens = args[0].get_shape().lens();
auto b_lens = args[1].get_shape().lens();
std::size_t dim_0 = b_lens.size() - 2;
std::size_t dim_1 = b_lens.size() - 1;
bool trans = !args[1].get_shape().transposed();
rocblas_int m = b_lens[trans ? dim_1 : dim_0];
rocblas_int n = b_lens[trans ? dim_0 : dim_1];
bool transb = args[1].get_shape().transposed();
bool transa = false;
rocblas_int lda = a_lens[0];
rocblas_int ldb = args[1].get_shape().strides()[(transb ? dim_1 : dim_0)];
rocblas_int ldc = b_lens[dim_1];
rocblas_int m = 1;
rocblas_int n = args[1].get_shape().lens()[dim_1];
rocblas_int k = a_lens[0];
float beta = 0.0f;
rocblas_int lda = args[1].get_shape().strides()[trans ? dim_1 : dim_0];
assert(b_lens[dim_0] == args[0].get_shape().elements());
assert(b_lens.back() == args[0].get_shape().elements());
std::size_t batch_num = std::accumulate(
b_lens.rbegin() + 2, b_lens.rend(), std::size_t{1}, std::multiplies<std::size_t>());
auto cpu_a = migraphx::gpu::from_gpu(args[0]);
auto cpu_b = migraphx::gpu::from_gpu(args[1]);
auto cpu_res = migraphx::gpu::from_gpu(args[2]);
std::cout << "gpu::gemm, cpu_a = " << cpu_a << std::endl;
std::cout << "gpu::gemm, cpu_b = " << cpu_b << std::endl;
std::cout << "gpu::gemm, cpu_res = " << cpu_res << std::endl;
std::cout << "gpu::gemm, transb = " << transb << std::endl;
std::cout << "gpu::gemm, transa = " << transb << std::endl;
std::cout << "gpu::gemm, m = " << m << std::endl;
std::cout << "gpu::gemm, n = " << n << std::endl;
std::cout << "gpu::gemm, k = " << k << std::endl;
std::cout << "gpu::gemm, lda = " << lda << std::endl;
std::cout << "gpu::gemm, ldb = " << ldb << std::endl;
std::cout << "gpu::gemm, ldc = " << ldc << std::endl;
output_shape.visit_type([&](auto as) {
auto to_pointer = [&](auto&& arg) { return to_rocblas_type(as.from(arg.data())); };
hipMemset(to_pointer(args[2]), 0, output_shape.bytes());
});
output_shape.visit_type([&](auto as) {
auto alpha_r = to_rocblas_type(as(op.alpha));
auto beta_r = to_rocblas_type(as(beta));
auto to_pointer = [&](auto&& arg, std::size_t offset = 0) {
return to_rocblas_type(as.from(arg.data() + offset));
auto to_pointer = [&](auto&& arg) {
return to_rocblas_type(as.from(arg.data()));
};
for(std::size_t batch_no = 0; batch_no < batch_num; ++batch_no)
{
generic_rocblas_gemv(as,
ctx.get_stream().get_rocblas(),
trans ? rocblas_operation_transpose : rocblas_operation_none,
n,
m,
&alpha_r,
to_pointer(args[0]),
lda,
to_pointer(args[1], batch_no * m * n),
1,
&beta_r,
to_pointer(args[2], batch_no * m),
1);
}
generic_rocblas_batched_gemm(
as,
ctx.get_stream().get_rocblas(),
transb ? rocblas_operation_transpose : rocblas_operation_none,
transa ? rocblas_operation_transpose : rocblas_operation_none,
n,
m,
k,
&alpha_r,
to_pointer(args[1]),
ldb,
k * n,
to_pointer(args[0]),
lda,
0,
&beta_r,
to_pointer(args[2]),
ldc,
m * n,
batch_num);
});
return args[2];
}
// (batch) matrix multiplication
else
......@@ -465,14 +517,14 @@ argument miopen_gemm::compute(context& ctx,
m,
k,
&alpha_r,
to_pointer(args[1], k * n * num_matrices * b_ind),
to_pointer(args[1], k * n * num_matrices * b_ind * type_size),
ldb,
k * n,
to_pointer(args[0], m * k * num_matrices * a_ind),
to_pointer(args[0], m * k * num_matrices * a_ind * type_size),
lda,
m * k,
&beta_r,
to_pointer(args[2], m * n * num_matrices * out_ind),
to_pointer(args[2], m * n * num_matrices * out_ind * type_size),
ldc,
m * n,
num_matrices);
......@@ -489,12 +541,12 @@ argument miopen_gemm::compute(context& ctx,
m,
k,
&alpha_r,
to_pointer(args[1], k * n * num_matrices * b_ind),
to_pointer(args[1], k * n * b_ind * type_size),
ldb,
to_pointer(args[0], m * k * num_matrices * a_ind),
to_pointer(args[0], m * k * a_ind * type_size),
lda,
&beta_r,
to_pointer(args[2], m * n * num_matrices * out_ind),
to_pointer(args[2], m * n * out_ind * type_size),
ldc);
}
......
Markdown is supported
0% or .
You are about to add 0 people to the discussion. Proceed with caution.
Finish editing this message first!
Please register or to comment