MCPcopy Create free account
hub / github.com/DeepRec-AI/DeepRec / AddToQueue

Method AddToQueue

tensorflow/lite/delegates/gpu/cl/kernels/softmax1x1.cc:120–133  ·  view source on GitHub ↗

Source from the content-addressed store, hash-verified

118}
119
120Status Softmax1x1::AddToQueue(CLCommandQueue* queue) {
121 kernel_.ResetBindingCounter();
122 RETURN_IF_ERROR(kernel_.SetMemoryAuto(src_[0]->GetMemoryPtr()));
123 RETURN_IF_ERROR(BindArgs(&kernel_, linked_operations_));
124 RETURN_IF_ERROR(kernel_.SetMemoryAuto(dst_[0]->GetMemoryPtr()));
125 RETURN_IF_ERROR(kernel_.SetBytesAuto(src_[0]->GetSizeWithDepth()));
126 const int depth = src_[0]->Depth();
127 RETURN_IF_ERROR(
128 kernel_.SetBytesAuto(int2(depth, IntegralDivideRoundUp(depth, 32))));
129 RETURN_IF_ERROR(
130 kernel_.SetBytesAuto(GetMaskForLastPlane(src_[0]->Channels())));
131
132 return queue->DispatchImplicit(kernel_, {32u, 1u, 1u}, {32u, 1u, 1u});
133}
134
135Softmax1x1 CreateSoftmax1x1(const OperationDef& definition) {
136 return Softmax1x1(definition);

Callers

nothing calls this directly

Calls 11

BindArgsFunction · 0.85
GetMaskForLastPlaneFunction · 0.85
ResetBindingCounterMethod · 0.80
SetMemoryAutoMethod · 0.80
GetSizeWithDepthMethod · 0.80
DepthMethod · 0.80
ChannelsMethod · 0.80
DispatchImplicitMethod · 0.80
IntegralDivideRoundUpFunction · 0.50
GetMemoryPtrMethod · 0.45
SetBytesAutoMethod · 0.45

Tested by

no test coverage detected