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

Fix compile problem when cuda_arch < 6000 #29576

Merged
merged 3 commits into from
Dec 14, 2020
Merged

Fix compile problem when cuda_arch < 6000 #29576

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
0b5b963
fix compile problem when cuda_arch < 6000
zhiqiu Dec 14, 2020
b2e1b34
refine code
zhiqiu Dec 14, 2020
f6a92ed
refine code
zhiqiu Dec 14, 2020
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
22 changes: 14 additions & 8 deletions paddle/fluid/operators/layer_norm_op.cu
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -109,18 +109,25 @@ struct PairForLayerNormAddFunctor {

template <typename T>
__inline__ __device__ T rsqrt(const T val) {
return ::rsqrt(val);
return static_cast<T>(1) / sqrt(val);
}

template <>
__inline__ __device__ float rsqrt(const float val) {
return rsqrtf(val);
}

template <>
__inline__ __device__ double rsqrt(const double val) {
return rsqrt(val);
}

#if CUDA_ARCH_FP16_SUPPORTED(__CUDA_ARCH__)
template <>
__inline__ __device__ half rsqrt(const half val) {
return hrsqrt(val);
}
#endif

template <typename T, typename U, int BlockDim>
__global__ void LayerNormForward(const T *x, const U *scale, const U *bias,
Expand Down Expand Up @@ -841,6 +848,7 @@ class LayerNormKernel<platform::CUDADeviceContext, T>
: public framework::OpKernel<T> {
public:
void Compute(const framework::ExecutionContext &ctx) const override {
using U = LayerNormParamType<T>;
const float epsilon = ctx.Attr<float>("epsilon");
auto *scale = ctx.Input<Tensor>("Scale");
auto *bias = ctx.Input<Tensor>("Bias");
Expand All @@ -854,12 +862,10 @@ class LayerNormKernel<platform::CUDADeviceContext, T>
const auto x_dims = x->dims();
auto *x_data = x->data<T>();
auto *y_data = y->mutable_data<T>(ctx.GetPlace());
auto *mean_data = mean->mutable_data<LayerNormParamType<T>>(ctx.GetPlace());
auto *var_data = var->mutable_data<LayerNormParamType<T>>(ctx.GetPlace());
auto *scale_data =
(scale == nullptr ? nullptr : scale->data<LayerNormParamType<T>>());
auto *bias_data =
(bias == nullptr ? nullptr : bias->data<LayerNormParamType<T>>());
auto *mean_data = mean->mutable_data<U>(ctx.GetPlace());
auto *var_data = var->mutable_data<U>(ctx.GetPlace());
auto *scale_data = (scale == nullptr ? nullptr : scale->data<U>());
auto *bias_data = (bias == nullptr ? nullptr : bias->data<U>());

auto matrix_dim = framework::flatten_to_2d(x_dims, begin_norm_axis);
int batch_size = static_cast<int>(matrix_dim[0]);
Expand All @@ -869,7 +875,7 @@ class LayerNormKernel<platform::CUDADeviceContext, T>

switch (GetDesiredBlockDim(feature_size)) {
FIXED_BLOCK_DIM_CASE(
LayerNormForward<T, LayerNormParamType<T>,
LayerNormForward<T, U,
kBlockDim><<<batch_size, kBlockDim, 0, stream>>>(
x_data, scale_data, bias_data, y_data, mean_data, var_data,
epsilon, feature_size));
Expand Down