Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Use GemmStridedBatched in batch matmul kernels #38290

Merged
merged 1 commit into from
Apr 10, 2020
Merged

Use GemmStridedBatched in batch matmul kernels #38290

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Jump to
Jump to file
Failed to load files.
Diff view
Diff view
77 changes: 74 additions & 3 deletions tensorflow/core/kernels/batch_matmul_op_impl.h
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -330,8 +330,26 @@ struct LaunchBatchMatMul<GPUDevice, Scalar> {
auto* a_base_ptr = in_x.template flat<Scalar>().data();
auto* b_base_ptr = in_y.template flat<Scalar>().data();
auto* c_base_ptr = out->template flat<Scalar>().data();

if (!bcast.IsBroadcastingRequired()) {
uint64 a_stride;
uint64 b_stride;
uint64 c_stride;

bool is_full_broadcast =
std::min(bcast.x_batch_size(), bcast.y_batch_size()) == 1;
bool use_strided_batched =
(!bcast.IsBroadcastingRequired() || is_full_broadcast) &&
batch_size > 1;
if (use_strided_batched) {
a_stride = bcast.x_batch_size() != 1 ? m * k : 0;
b_stride = bcast.y_batch_size() != 1 ? k * n : 0;
c_stride = m * n;
a_device_memory.push_back(AsDeviceMemory(a_base_ptr));
b_device_memory.push_back(AsDeviceMemory(b_base_ptr));
c_device_memory.push_back(AsDeviceMemory(c_base_ptr));
a_ptrs.push_back(&a_device_memory.back());
b_ptrs.push_back(&b_device_memory.back());
c_ptrs.push_back(&c_device_memory.back());
} else if (!bcast.IsBroadcastingRequired()) {
for (int64 i = 0; i < batch_size; ++i) {
a_device_memory.push_back(AsDeviceMemory(a_base_ptr + i * m * k));
b_device_memory.push_back(AsDeviceMemory(b_base_ptr + i * k * n));
Expand Down Expand Up @@ -407,6 +425,23 @@ struct LaunchBatchMatMul<GPUDevice, Scalar> {
", k=", k));
}
}
} else if (use_strided_batched) {
bool blas_launch_status =
stream
->ThenBlasGemmStridedBatched(
blas_transpose_b, blas_transpose_a, n, m, k,
static_cast<Coefficient>(1.0), *b_ptrs[0], adj_y ? k : n,
b_stride, *a_ptrs[0], adj_x ? m : k, a_stride,
static_cast<Coefficient>(0.0), c_ptrs[0], n, c_stride,
batch_size)
.ok();
if (!blas_launch_status) {
context->SetStatus(errors::Internal(
"Blas xGEMMStridedBatched launch failed : a.shape=",
in_x.shape().DebugString(),
", b.shape=", in_y.shape().DebugString(), ", m=", m, ", n=", n,
", k=", k, ", batch_size=", batch_size));
}
} else {
BlasScratchAllocator scratch_allocator(context);
bool blas_launch_status =
Expand Down Expand Up @@ -467,7 +502,26 @@ struct LaunchBatchMatMul<GPUDevice, Eigen::half> {
auto* b_base_ptr = in_y.template flat<Scalar>().data();
auto* c_base_ptr = out->template flat<Scalar>().data();

if (!bcast.IsBroadcastingRequired()) {
uint64 a_stride;
uint64 b_stride;
uint64 c_stride;

bool is_full_broadcast =
std::min(bcast.x_batch_size(), bcast.y_batch_size()) == 1;
bool use_strided_batched =
(!bcast.IsBroadcastingRequired() || is_full_broadcast) &&
batch_size > 1;
if (use_strided_batched) {
a_stride = bcast.x_batch_size() != 1 ? m * k : 0;
b_stride = bcast.y_batch_size() != 1 ? k * n : 0;
c_stride = m * n;
a_device_memory.push_back(AsDeviceMemory(a_base_ptr));
b_device_memory.push_back(AsDeviceMemory(b_base_ptr));
c_device_memory.push_back(AsDeviceMemory(c_base_ptr));
a_ptrs.push_back(&a_device_memory.back());
b_ptrs.push_back(&b_device_memory.back());
c_ptrs.push_back(&c_device_memory.back());
} else if (!bcast.IsBroadcastingRequired()) {
for (int64 i = 0; i < batch_size; ++i) {
a_device_memory.push_back(AsDeviceMemory(a_base_ptr + i * m * k));
b_device_memory.push_back(AsDeviceMemory(b_base_ptr + i * k * n));
Expand Down Expand Up @@ -518,6 +572,23 @@ struct LaunchBatchMatMul<GPUDevice, Eigen::half> {
", b.shape=", in_y.shape().DebugString(), ", m=", m, ", n=", n,
", k=", k));
}
} else if (use_strided_batched) {
bool blas_launch_status =
stream
->ThenBlasGemmStridedBatched(
blas_transpose_b, blas_transpose_a, n, m, k,
static_cast<Coefficient>(1.0), *b_ptrs[0], adj_y ? k : n,
b_stride, *a_ptrs[0], adj_x ? m : k, a_stride,
static_cast<Coefficient>(0.0), c_ptrs[0], n, c_stride,
batch_size)
.ok();
if (!blas_launch_status) {
context->SetStatus(errors::Internal(
"Blas xGEMMStridedBatched launch failed : a.shape=",
in_x.shape().DebugString(),
", b.shape=", in_y.shape().DebugString(), ", m=", m, ", n=", n,
", k=", k, ", batch_size=", batch_size));
}
} else {
BlasScratchAllocator scratch_allocator(context);
bool blas_launch_status =
Expand Down