experiments/STT-tensorflow
1
0
Fork 0
Code Issues Pull Requests Projects Releases Wiki Activity

Using IMAGE_BUFFER as buffer for writing.

Browse Source 
PiperOrigin-RevId: 269903377
This commit is contained in:
A. Unique TensorFlower 2019-09-18 15:23:30 -07:00 committed by TensorFlower Gardener
parent 6274f037d4
commit a6d8ab779d
26 changed files with 37 additions and 31 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

2
tensorflow/lite/delegates/gpu/cl/kernels/concat_xy.cc
View File

@ -111,7 +111,7 @@ Status ConcatXY::BindArguments() {
for (int i = 0; i < tensors_count_; ++i) {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[i]->GetMemoryPtr()));
}
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
int max_src_width = 0;
int max_src_height = 0;

2
tensorflow/lite/delegates/gpu/cl/kernels/concat_z.cc
View File

@ -189,7 +189,7 @@ Status ConcatZ::BindArguments() {
for (int i = 0; i < channels_.size(); ++i) {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[i]->GetMemoryPtr()));
}
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
for (int i = 0; i < channels_.size(); ++i) {
int4 size(src_[i]->Width(), src_[i]->Height(), channels_[i],

2
tensorflow/lite/delegates/gpu/cl/kernels/conv_buffer.cc
View File

@ -217,7 +217,7 @@ Status ConvBuffer::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
int4 src_size = int4(src_[0]->Width(), src_[0]->Height(),
src_[0]->Width() * src_[0]->Height(), src_[0]->Depth());
int4 dst_size = int4(dst_[0]->Width(), dst_[0]->Height(),

2
tensorflow/lite/delegates/gpu/cl/kernels/conv_buffer_1x1.cc
View File

@ -259,7 +259,7 @@ Status ConvBuffer1x1::BindArguments() {
RETURN_IF_ERROR(kernel->SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel->SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(kernel, linked_operations_));
RETURN_IF_ERROR(kernel->SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel->SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
int4 src_size = int4(src_[0]->Width(), src_[0]->Height(),
GetGridWidth(src_[0]->Width()) * src_[0]->Height(),
src_[0]->Depth());

2
tensorflow/lite/delegates/gpu/cl/kernels/conv_constants.cc
View File

@ -234,7 +234,7 @@ Status ConvConstants::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(stride_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(padding_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));

2
tensorflow/lite/delegates/gpu/cl/kernels/conv_powervr.cc
View File

@ -80,7 +80,7 @@ Status ConvPowerVR::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
if (!conv_params_.x_kernel_is_1 || !conv_params_.y_kernel_is_1) {
RETURN_IF_ERROR(kernel_.SetBytesAuto(stride_padding_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(kernel_dilation_));

2
tensorflow/lite/delegates/gpu/cl/kernels/conv_texture.cc
View File

@ -318,7 +318,7 @@ Status ConvTexture::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_3_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
if (!(kernel_size_.x == 1 && kernel_size_.y == 1)) {

2
tensorflow/lite/delegates/gpu/cl/kernels/convolution_transposed.cc
View File

@ -230,7 +230,7 @@ Status ConvolutionTransposed::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(kernel_size_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(stride_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(padding_));

2
tensorflow/lite/delegates/gpu/cl/kernels/convolution_transposed_3x3_thin.cc
View File

@ -211,7 +211,7 @@ Status ConvolutionTransposed3x3Thin::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
return OkStatus();

2
tensorflow/lite/delegates/gpu/cl/kernels/convolution_transposed_thin.cc
View File

@ -203,7 +203,7 @@ Status ConvolutionTransposedThin::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_buf_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(bias_value_));

2
tensorflow/lite/delegates/gpu/cl/kernels/depth_wise_conv.cc
View File

@ -222,7 +222,7 @@ Status DepthWiseConvolution::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(kernel_size_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(stride_));
RETURN_IF_ERROR(kernel_.SetBytesAuto(padding_));

2
tensorflow/lite/delegates/gpu/cl/kernels/depth_wise_conv_3x3.cc
View File

@ -312,7 +312,7 @@ Status DepthWiseConv3x3::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
return OkStatus();

2
tensorflow/lite/delegates/gpu/cl/kernels/fully_connected_texture.cc
View File

@ -158,7 +158,7 @@ Status FullyConnectedTexture::AddToQueue(CLCommandQueue* queue) {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(weights_.GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(biases_.GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_depth_x4));

2
tensorflow/lite/delegates/gpu/cl/kernels/gpu_operation.cc
View File

@ -143,7 +143,7 @@ Status ElementwiseOperation::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArguments(&kernel_));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
return OkStatus();

2
tensorflow/lite/delegates/gpu/cl/kernels/max_unpooling.cc
View File

@ -131,7 +131,7 @@ Status MaxUnpooling::BindArguments() {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[1]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(kernel_size_));

2
tensorflow/lite/delegates/gpu/cl/kernels/padding.cc
View File

@ -118,7 +118,7 @@ Status Padding::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(prepended_));

4
tensorflow/lite/delegates/gpu/cl/kernels/pooling.cc
View File

@ -228,9 +228,9 @@ Status Pooling::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
if (output_indices_) {
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[1]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[1]->GetMemoryPtrForWriting()));
}
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));

2
tensorflow/lite/delegates/gpu/cl/kernels/reshape.cc
View File

@ -102,7 +102,7 @@ Status Reshape::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
const int2 plane_size = int2(src_[0]->Width() * src_[0]->Channels(),

2
tensorflow/lite/delegates/gpu/cl/kernels/reshapex4.cc
View File

@ -87,7 +87,7 @@ Status Reshapex4::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
const int2 plane_size = int2(src_[0]->Width() * src_[0]->Depth(),

2
tensorflow/lite/delegates/gpu/cl/kernels/softmax.cc
View File

@ -101,7 +101,7 @@ Status Softmax::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(
kernel_.SetBytesAuto(GetMaskForLastPlane(src_[0]->Channels())));

2
tensorflow/lite/delegates/gpu/cl/kernels/softmax1x1.cc
View File

@ -121,7 +121,7 @@ Status Softmax1x1::AddToQueue(CLCommandQueue* queue) {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
const int depth = src_[0]->Depth();
RETURN_IF_ERROR(

2
tensorflow/lite/delegates/gpu/cl/kernels/strided_slice.cc
View File

@ -153,7 +153,7 @@ Status StridedSlice::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
int3 offset = GetOffset(attributes_, src_[0]->Width(), src_[0]->Height(),
src_[0]->Channels());
RETURN_IF_ERROR(kernel_.SetBytesAuto(int4(offset.x, offset.y, offset.z, 1)));

2
tensorflow/lite/delegates/gpu/cl/kernels/upsample.cc
View File

@ -105,7 +105,7 @@ Status Upsample::BindArguments() {
kernel_.ResetBindingCounter();
RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtrForWriting()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
RETURN_IF_ERROR(kernel_.SetBytesAuto(dst_[0]->GetSizeWithDepth()));
float2 scale_factor =

14
tensorflow/lite/delegates/gpu/cl/kernels/util.cc
View File

@ -228,14 +228,10 @@ std::string WriteGlobalFLT4(TensorStorageType storage_type, DataType data_type,
const std::string& y, const std::string& z) {
switch (storage_type) {
case TensorStorageType::BUFFER:
case TensorStorageType::IMAGE_BUFFER:
return absl::StrCat(tensor_name, "[((", z, ") * ", size_name, ".y + (", y,
")) * ", size_name, ".x + (", x, ")] = ", var_name,
";\n");
case TensorStorageType::IMAGE_BUFFER:
return absl::StrCat(GetWriteImageFromDataType(data_type), "(",
tensor_name, ", ((", z, ") * ", size_name, ".y + (",
y, ")) * ", size_name, ".x + (", x, "), ", var_name,
");\n");
case TensorStorageType::TEXTURE_2D:
return absl::StrCat(GetWriteImageFromDataType(data_type), "(",
tensor_name, ", (int2)((", x, "), (", y, ") * ",
@ -259,12 +255,12 @@ std::string WriteGlobalFLT4(TensorStorageType storage_type, DataType data_type,
const std::string& global_address) {
switch (storage_type) {
case TensorStorageType::BUFFER:
case TensorStorageType::IMAGE_BUFFER:
return absl::StrCat(tensor_name, "[", global_address, "] = ", var_name,
";\n");
case TensorStorageType::TEXTURE_2D:
case TensorStorageType::SINGLE_TEXTURE_2D:
case TensorStorageType::TEXTURE_ARRAY:
case TensorStorageType::IMAGE_BUFFER:
return absl::StrCat(GetWriteImageFromDataType(data_type), "(",
tensor_name, ", ", global_address, ", ", var_name,
");\n");
@ -308,7 +304,11 @@ std::string GetTensorDeclaration(TensorStorageType storage_type,
case TensorStorageType::TEXTURE_ARRAY:
return GetImageModifier(access) + " image2d_array_t";
case TensorStorageType::IMAGE_BUFFER:
return GetImageModifier(access) + " image1d_buffer_t";
if (access == AccessType::WRITE) {
return absl::StrCat("__global ", GetDataType4(data_type), "*");
} else {
return GetImageModifier(access) + " image1d_buffer_t";
}
case TensorStorageType::UNKNOWN:
return "";
}

2
tensorflow/lite/delegates/gpu/cl/tensor.cc
View File

@ -161,6 +161,8 @@ cl_mem Tensor::GetMemoryPtr() const {
: memory_;
}
cl_mem Tensor::GetMemoryPtrForWriting() const { return memory_; }
Status Tensor::WriteDataBHWC(absl::Span<const float> in,
CLCommandQueue* queue) {
if (in.size() != channels_ * width_ * height_) {

4
tensorflow/lite/delegates/gpu/cl/tensor.h
View File

@ -67,6 +67,10 @@ class Tensor {
}
cl_mem GetMemoryPtr() const;
// This function returns buffer memory ptr for IMAGE_BUFFER instead of image
// memory ptr.
cl_mem GetMemoryPtrForWriting() const;
Status WriteDataBHWC(absl::Span<const float> in, CLCommandQueue* queue);
Status ReadDataBHWC(absl::Span<float> out, CLCommandQueue* queue) const;