wrapCudaMemcpyAsync

Fully qualified name: isaacsim::sensors::rtx::wrapCudaMemcpyAsync

template<class T>
void isaacsim::sensors::rtx::wrapCudaMemcpyAsync(

T *dst,

const T *src,

uint32_t startLoc,

uint32_t num,

uint32_t numSpillover,

cudaMemcpyKind kind,

)

Wraps CUDA asynchronous memory copy with spillover handling.

Template function for performing CUDA memory copies that can handle data spillover when the destination buffer is smaller than the source data, implementing a circular buffer pattern.

Template Parameters:

T – Data type being copied

Parameters:

• dst – [out] Destination pointer for the copied data

• src – [in] Source pointer containing data to copy

• startLoc – [in] Starting location in destination buffer

• num – [in] Number of elements to copy to primary location

• numSpillover – [in] Number of elements to copy to beginning of buffer (spillover)

• kind – [in] Type of memory copy operation (host-to-device, device-to-host, etc.)