Functions
template<TensorDataType TDataType>
void	copyHostToHostImpl (const void src_data, void dst_data, size_t count)
	Fast raw-memory copy for host-accessible tensors of same abstract type.
template<TensorDataType TSrcDataType, TensorDataType TDstDataType>
void	copyHostToHostWithConversionImpl (const void src_data, void dst_data, size_t count)
	Element-wise host copy with conversion between abstract data types.

Function Documentation

◆ copyHostToHostImpl()

template<TensorDataType TDataType>

void Mila::Dnn::Compute::Cpu::Detail::copyHostToHostImpl	(	const void *	src_data,
		void *	dst_data,
		size_t	count )

inline

Fast raw-memory copy for host-accessible tensors of same abstract type.

Copies count logical elements from src_data to dst_data using a byte-wise memcpy based on the abstract element size for TDataType.

Template Parameters

TDataType Abstract tensor data type for src and dst

Parameters

src_data	Pointer to source memory (may be null)
dst_data	Pointer to destination memory (may be null)
count	Number of logical elements to copy

Note: No operation performed if either pointer is null or count is zero.

Here is the call graph for this function:

Here is the caller graph for this function:

◆ copyHostToHostWithConversionImpl()

template<TensorDataType TSrcDataType, TensorDataType TDstDataType>

void Mila::Dnn::Compute::Cpu::Detail::copyHostToHostWithConversionImpl	(	const void *	src_data,
		void *	dst_data,
		size_t	count )

inline

Element-wise host copy with conversion between abstract data types.

Copies count logical elements from src_data (type TSrcDataType) to dst_data (type TDstDataType) performing a per-element conversion via static_cast to the native CPU types.

If the source and destination abstract types are the same, this forwards to copyHostToHostImpl.