walberla::GeneratedGPUPackInfo Class Reference

Detailed Description

Data packing/unpacking for ghost layer based communication of a GPUField.

Encapsulate information on how to extract data from blocks that should be communicated to neighboring blocks (see pack()) and how to inject this data in a receiving block (see unpack()). This involves a memory buffer and two memory copy operations.

A special method exists for communication between two blocks which are allocated on the same process (see communicateLocal()). In this case the data does not have be communicated via a buffer, but can be copied directly.

Data that is packed in direction "dir" at one block is unpacked in direction "stencil::inverseDir[dir]" at the neighboring block. This behavior must be implemented in communicateLocal()!

#include <GeneratedGPUPackInfo.h>

Inheritance diagram for walberla::GeneratedGPUPackInfo:

Public Member Functions
	GeneratedGPUPackInfo ()=default
virtual	~GeneratedGPUPackInfo ()=default
virtual void	pack (stencil::Direction dir, unsigned char *buffer, IBlock *block, gpuStream_t stream)=0
	Pack data from a block into a send buffer.
virtual void	communicateLocal (stencil::Direction dir, const IBlock *sender, IBlock *receiver, gpuStream_t stream)=0
	Copy data from one local block to another local block.
virtual void	unpack (stencil::Direction dir, unsigned char *buffer, IBlock *block, gpuStream_t stream)=0
	Unpack data from a receive buffer into a block.
virtual uint_t	size (stencil::Direction dir, IBlock *block)=0

Constructor & Destructor Documentation

GeneratedGPUPackInfo()

walberla::gpu::GeneratedGPUPackInfo::GeneratedGPUPackInfo ( )

default

~GeneratedGPUPackInfo()

virtual walberla::gpu::GeneratedGPUPackInfo::~GeneratedGPUPackInfo ( )

virtualdefault

Member Function Documentation

communicateLocal()

virtual void walberla::gpu::GeneratedGPUPackInfo::communicateLocal ( stencil::Direction dir, const IBlock * sender, IBlock * receiver, gpuStream_t stream )

pure virtual

Copy data from one local block to another local block.

Both blocks are allocated on the same MPI rank.

Parameters

dir	the direction of the communication (from sender to receiver)
sender	id of block where the data should be copied from
receiver	id of block where the data should be copied to
stream	GPU stream

pack()

virtual void walberla::gpu::GeneratedGPUPackInfo::pack ( stencil::Direction dir, unsigned char * buffer, IBlock * block, gpuStream_t stream )

pure virtual

Pack data from a block into a send buffer.

Parameters

dir	pack data for neighbor in this direction
buffer	buffer for writing the data into
block	the block whose data should be packed into a buffer
stream	GPU stream

size()

virtual uint_t walberla::gpu::GeneratedGPUPackInfo::size ( stencil::Direction dir, IBlock * block )

pure virtual

unpack()

virtual void walberla::gpu::GeneratedGPUPackInfo::unpack ( stencil::Direction dir, unsigned char * buffer, IBlock * block, gpuStream_t stream )

pure virtual

Unpack data from a receive buffer into a block.

Parameters

dir	receive data from neighbor in this direction
buffer	buffer for reading the data from
block	the block where the unpacked data should be stored into
stream	GPU stream

---

The documentation for this class was generated from the following file:

• /builds/administration/walberla-website/walberla/src/gpu/communication/GeneratedGPUPackInfo.h