foreman
9ba3948388
SWDEV-124171 - adding support for p2p OCL in rocm stack - Add cl_amd_copy_buffer_p2p extension for P2P transfers. The extension adds a new API entry - clEnqueueCopyBufferP2PAMD() which allows to transfer CL buffers between different CL contexts on different GPUs. If P2P isn't possible, then double copy performed - Also the app can query the P2P support capabilities for the device. A list of P2P accessible devices can be returned for the current device http://ocltc.amd.com/reviews/r/12913/ Affected files ... ... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_context.cpp#54 edit ... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_device.cpp#62 edit ... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_p2p_amd.cpp#1 add ... //depot/stg/opencl/drivers/opencl/api/opencl/amdocl/cl_p2p_amd.h#1 add ... //depot/stg/opencl/drivers/opencl/api/opencl/khronos/headers/opencl2.0/CL/cl_ext.h#29 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/cpu/cpuvirtual.hpp#14 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/device.hpp#287 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/gpu/gpuvirtual.hpp#141 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/pal/palvirtual.hpp#26 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocblit.cpp#19 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.cpp#54 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocdevice.hpp#22 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocmemory.cpp#24 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocsettings.cpp#19 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocvirtual.cpp#39 edit ... //depot/stg/opencl/drivers/opencl/runtime/device/rocm/rocvirtual.hpp#12 edit ... //depot/stg/opencl/drivers/opencl/runtime/platform/command.cpp#79 edit ... //depot/stg/opencl/drivers/opencl/runtime/platform/command.hpp#84 edit
70 строки
2.8 KiB
C++
70 строки
2.8 KiB
C++
//
|
|
// Copyright (c) 2011 Advanced Micro Devices, Inc. All rights reserved.
|
|
//
|
|
|
|
#ifndef CPUVIRTUAL_HPP_
|
|
#define CPUVIRTUAL_HPP_
|
|
|
|
#include "top.hpp"
|
|
#include "device/device.hpp"
|
|
#include "thread/atomic.hpp"
|
|
#include "thread/thread.hpp"
|
|
#include "platform/ndrange.hpp"
|
|
|
|
//! \namespace cpu CPU Device Implementation
|
|
namespace cpu {
|
|
|
|
class WorkerThread;
|
|
class Device;
|
|
|
|
class VirtualCPU : public device::VirtualDevice {
|
|
private:
|
|
WorkerThread** cores_; //!< Pointer to array of Worker threads
|
|
static amd::Atomic<size_t> numWorkerThreads_; //!< Current Worker Threads number
|
|
bool acceptingCommands_;
|
|
|
|
public:
|
|
VirtualCPU(cpu::Device& device);
|
|
~VirtualCPU();
|
|
bool terminate();
|
|
|
|
WorkerThread* getWorkerThread(size_t id) { return cores_[id]; }
|
|
|
|
bool acceptingCommands() const { return acceptingCommands_; }
|
|
|
|
virtual void submitReadMemory(amd::ReadMemoryCommand& command);
|
|
virtual void submitWriteMemory(amd::WriteMemoryCommand& command);
|
|
virtual void submitCopyMemory(amd::CopyMemoryCommand& command);
|
|
virtual void submitCopyMemoryP2P(amd::CopyMemoryP2PCommand& command) {}
|
|
virtual void submitMapMemory(amd::MapMemoryCommand& command);
|
|
virtual void submitUnmapMemory(amd::UnmapMemoryCommand& command);
|
|
virtual void submitKernel(amd::NDRangeKernelCommand& command);
|
|
virtual void submitNativeFn(amd::NativeFnCommand& command);
|
|
virtual void submitMarker(amd::Marker& command);
|
|
virtual void submitFillMemory(amd::FillMemoryCommand& command);
|
|
virtual void submitMigrateMemObjects(amd::MigrateMemObjectsCommand& cmd) {}
|
|
virtual void submitAcquireExtObjects(amd::AcquireExtObjectsCommand& cmd);
|
|
virtual void submitReleaseExtObjects(amd::ReleaseExtObjectsCommand& cmd);
|
|
virtual void submitPerfCounter(amd::PerfCounterCommand& cmd);
|
|
virtual void submitThreadTraceMemObjects(amd::ThreadTraceMemObjectsCommand& cmd);
|
|
virtual void submitThreadTrace(amd::ThreadTraceCommand& cmd);
|
|
virtual void flush(amd::Command* list = NULL, bool wait = false);
|
|
virtual void submitSignal(amd::SignalCommand& cmd);
|
|
virtual void submitMakeBuffersResident(amd::MakeBuffersResidentCommand& cmd);
|
|
virtual void submitSvmFreeMemory(amd::SvmFreeMemoryCommand& cmd);
|
|
virtual void submitSvmCopyMemory(amd::SvmCopyMemoryCommand& cmd);
|
|
virtual void submitSvmFillMemory(amd::SvmFillMemoryCommand& cmd);
|
|
virtual void submitSvmMapMemory(amd::SvmMapMemoryCommand& cmd);
|
|
virtual void submitSvmUnmapMemory(amd::SvmUnmapMemoryCommand& cmd);
|
|
|
|
virtual void computeLocalSizes(amd::NDRangeKernelCommand& command, amd::NDRange& local);
|
|
|
|
static bool fillImage(amd::Image& image, address fillMem, const void* pattern,
|
|
const amd::Coord3D& origin, const amd::Coord3D& region, size_t rowPitch,
|
|
size_t slicePitch, size_t elementSize);
|
|
};
|
|
|
|
} // namespace cpu
|
|
|
|
#endif // CPUVIRTUAL_HPP_
|