cuda::CudaRuntime Class Reference

#include <CudaRuntime.h>

Inheritance diagram for cuda::CudaRuntime:

Collaboration diagram for cuda::CudaRuntime:

List of all members.

Public Member Functions
	CudaRuntime ()
	~CudaRuntime ()
virtual void **	cudaRegisterFatBinary (void *fatCubin)
virtual void	cudaUnregisterFatBinary (void **fatCubinHandle)
virtual void	cudaRegisterVar (void **fatCubinHandle, char *hostVar, char *deviceAddress, const char *deviceName, int ext, int size, int constant, int global)
	register a CUDA global variable
virtual void	cudaRegisterTexture (void **fatCubinHandle, const struct textureReference *hostVar, const void **deviceAddress, const char *deviceName, int dim, int norm, int ext)
	registers a CUDA texture reference
virtual void	cudaRegisterShared (void **fatCubinHandle, void **devicePtr)
virtual void	cudaRegisterSharedVar (void **fatCubinHandle, void **devicePtr, size_t size, size_t alignment, int storage)
virtual void	cudaRegisterFunction (void **fatCubinHandle, const char *hostFun, char *deviceFun, const char *deviceName, int thread_limit, uint3 *tid, uint3 *bid, dim3 *bDim, dim3 *gDim, int *wSize)
virtual cudaError_t	cudaGetExportTable (const void **ppExportTable, const cudaUUID_t *pExportTableId)
virtual cudaError_t	cudaMalloc (void **devPtr, size_t size)
virtual cudaError_t	cudaMallocHost (void **ptr, size_t size)
virtual cudaError_t	cudaMallocPitch (void **devPtr, size_t *pitch, size_t width, size_t height)
virtual cudaError_t	cudaMallocArray (struct cudaArray **array, const struct cudaChannelFormatDesc *desc, size_t width, size_t height=1)
virtual cudaError_t	cudaFree (void *devPtr)
virtual cudaError_t	cudaFreeHost (void *ptr)
virtual cudaError_t	cudaFreeArray (struct cudaArray *array)
virtual cudaError_t	cudaMalloc3D (struct cudaPitchedPtr *pitchedDevPtr, struct cudaExtent extent)
virtual cudaError_t	cudaMalloc3DArray (struct cudaArray **arrayPtr, const struct cudaChannelFormatDesc *desc, struct cudaExtent extent)
virtual cudaError_t	cudaHostAlloc (void **pHost, size_t bytes, unsigned int flags)
virtual cudaError_t	cudaHostGetDevicePointer (void **pDevice, void *pHost, unsigned int flags)
virtual cudaError_t	cudaHostGetFlags (unsigned int *pFlags, void *pHost)
virtual cudaError_t	cudaHostRegister (void *pHost, size_t bytes, unsigned int flags)
virtual cudaError_t	cudaHostUnregister (void *pHost)
virtual cudaError_t	cudaMemcpy (void *dst, const void *src, size_t count, enum cudaMemcpyKind kind)
virtual cudaError_t	cudaMemcpyToSymbol (const char *symbol, const void *src, size_t count, size_t offset, enum cudaMemcpyKind kind=cudaMemcpyHostToDevice)
virtual cudaError_t	cudaMemcpyFromSymbol (void *dst, const char *symbol, size_t count, size_t offset=0, enum cudaMemcpyKind kind=cudaMemcpyDeviceToHost)
virtual cudaError_t	cudaMemcpyAsync (void *dst, const void *src, size_t count, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t	cudaMemcpyToArray (struct cudaArray *dst, size_t wOffset, size_t hOffset, const void *src, size_t count, enum cudaMemcpyKind kind)
virtual cudaError_t	cudaMemcpyFromArray (void *dst, const struct cudaArray *src, size_t wOffset, size_t hOffset, size_t count, enum cudaMemcpyKind kind)
virtual cudaError_t	cudaMemcpyArrayToArray (struct cudaArray *dst, size_t wOffsetDst, size_t hOffsetDst, const struct cudaArray *src, size_t wOffsetSrc, size_t hOffsetSrc, size_t count, enum cudaMemcpyKind kind)
virtual cudaError_t	cudaMemcpy2D (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t height, enum cudaMemcpyKind kind)
	perform a 2D memcpy from a dense buffer
virtual cudaError_t	cudaMemcpy2DToArray (struct cudaArray *dst, size_t wOffset, size_t hOffset, const void *src, size_t spitch, size_t width, size_t height, enum cudaMemcpyKind kind)
	perform a 2D memcpy to an array
virtual cudaError_t	cudaMemcpy2DFromArray (void *dst, size_t dpitch, const struct cudaArray *src, size_t wOffset, size_t hOffset, size_t width, size_t height, enum cudaMemcpyKind kind)
	perform a 2D memcpy from an array
virtual cudaError_t	cudaMemcpy3D (const struct cudaMemcpy3DParms *p)
virtual cudaError_t	cudaMemcpy3DAsync (const struct cudaMemcpy3DParms *p, cudaStream_t stream)
virtual cudaError_t	cudaMemset (void *devPtr, int value, size_t count)
virtual cudaError_t	cudaMemset2D (void *devPtr, size_t pitch, int value, size_t width, size_t height)
virtual cudaError_t	cudaMemset3D (struct cudaPitchedPtr pitchedDevPtr, int value, struct cudaExtent extent)
virtual cudaError_t	cudaGetSymbolAddress (void **devPtr, const char *symbol)
virtual cudaError_t	cudaGetSymbolSize (size_t *size, const char *symbol)
virtual cudaError_t	cudaGetDeviceCount (int *count)
virtual cudaError_t	cudaGetDeviceProperties (struct cudaDeviceProp *prop, int device)
virtual cudaError_t	cudaChooseDevice (int *device, const struct cudaDeviceProp *prop)
virtual cudaError_t	cudaSetDevice (int device)
virtual cudaError_t	cudaGetDevice (int *device)
virtual cudaError_t	cudaSetValidDevices (int *device_arr, int len)
virtual cudaError_t	cudaSetDeviceFlags (int flags)
virtual cudaError_t	cudaBindTexture (size_t *offset, const struct textureReference *texref, const void *devPtr, const struct cudaChannelFormatDesc *desc, size_t size=UINT_MAX)
	binds a texture to a reference and a CUDA memory block
virtual cudaError_t	cudaBindTexture2D (size_t *offset, const struct textureReference *texref, const void *devPtr, const struct cudaChannelFormatDesc *desc, size_t width, size_t height, size_t pitch)
virtual cudaError_t	cudaBindTextureToArray (const struct textureReference *texref, const struct cudaArray *array, const struct cudaChannelFormatDesc *desc)
virtual cudaError_t	cudaUnbindTexture (const struct textureReference *texref)
virtual cudaError_t	cudaGetTextureAlignmentOffset (size_t *offset, const struct textureReference *texref)
virtual cudaError_t	cudaGetTextureReference (const struct textureReference **texref, const char *symbol)
virtual cudaError_t	cudaGetChannelDesc (struct cudaChannelFormatDesc *desc, const struct cudaArray *array)
virtual struct cudaChannelFormatDesc	cudaCreateChannelDesc (int x, int y, int z, int w, enum cudaChannelFormatKind f)
virtual cudaError_t	cudaGetLastError (void)
virtual cudaError_t	cudaPeekAtLastError (void)
virtual cudaError_t	cudaConfigureCall (dim3 gridDim, dim3 blockDim, size_t sharedMem=0, cudaStream_t stream=0)
virtual cudaError_t	cudaSetupArgument (const void *arg, size_t size, size_t offset)
virtual cudaError_t	cudaLaunch (const char *entry)
virtual cudaError_t	cudaFuncGetAttributes (struct cudaFuncAttributes *attr, const char *func)
virtual cudaError_t	cudaFuncSetCacheConfig (const char *func, enum cudaFuncCache cacheConfig)
virtual cudaError_t	cudaEventCreate (cudaEvent_t *event)
virtual cudaError_t	cudaEventCreateWithFlags (cudaEvent_t *event, int flags)
virtual cudaError_t	cudaEventRecord (cudaEvent_t event, cudaStream_t stream)
virtual cudaError_t	cudaEventQuery (cudaEvent_t event)
virtual cudaError_t	cudaEventSynchronize (cudaEvent_t event)
virtual cudaError_t	cudaEventDestroy (cudaEvent_t event)
virtual cudaError_t	cudaEventElapsedTime (float *ms, cudaEvent_t start, cudaEvent_t end)
virtual cudaError_t	cudaStreamCreate (cudaStream_t *pStream)
virtual cudaError_t	cudaStreamDestroy (cudaStream_t stream)
virtual cudaError_t	cudaStreamSynchronize (cudaStream_t stream)
virtual cudaError_t	cudaStreamQuery (cudaStream_t stream)
virtual cudaError_t	cudaDriverGetVersion (int *driverVersion)
virtual cudaError_t	cudaRuntimeGetVersion (int *runtimeVersion)
virtual cudaError_t	cudaDeviceReset (void)
virtual cudaError_t	cudaDeviceSynchronize (void)
virtual cudaError_t	cudaDeviceSetLimit (enum cudaLimit limit, size_t value)
virtual cudaError_t	cudaDeviceGetLimit (size_t *pValue, enum cudaLimit limit)
virtual cudaError_t	cudaDeviceGetCacheConfig (enum cudaFuncCache *pCacheConfig)
virtual cudaError_t	cudaDeviceSetCacheConfig (enum cudaFuncCache cacheConfig)
virtual cudaError_t	cudaThreadExit (void)
virtual cudaError_t	cudaThreadSynchronize (void)
virtual cudaError_t	cudaGLMapBufferObject (void **devPtr, GLuint bufObj)
virtual cudaError_t	cudaGLMapBufferObjectAsync (void **devPtr, GLuint bufObj, cudaStream_t stream)
virtual cudaError_t	cudaGLRegisterBufferObject (GLuint bufObj)
virtual cudaError_t	cudaGLSetBufferObjectMapFlags (GLuint bufObj, unsigned int flags)
virtual cudaError_t	cudaGLSetGLDevice (int device)
virtual cudaError_t	cudaGLUnmapBufferObject (GLuint bufObj)
virtual cudaError_t	cudaGLUnmapBufferObjectAsync (GLuint bufObj, cudaStream_t stream)
virtual cudaError_t	cudaGLUnregisterBufferObject (GLuint bufObj)
virtual cudaError_t	cudaGraphicsGLRegisterBuffer (struct cudaGraphicsResource **resource, GLuint buffer, unsigned int flags)
virtual cudaError_t	cudaGraphicsGLRegisterImage (struct cudaGraphicsResource **resource, GLuint image, int target, unsigned int flags)
virtual cudaError_t	cudaGraphicsUnregisterResource (struct cudaGraphicsResource *resource)
virtual cudaError_t	cudaGraphicsResourceSetMapFlags (struct cudaGraphicsResource *resource, unsigned int flags)
virtual cudaError_t	cudaGraphicsMapResources (int count, struct cudaGraphicsResource **resources, cudaStream_t stream=0)
virtual cudaError_t	cudaGraphicsUnmapResources (int count, struct cudaGraphicsResource **resources, cudaStream_t stream=0)
virtual cudaError_t	cudaGraphicsResourceGetMappedPointer (void **devPtr, size_t *size, struct cudaGraphicsResource *resource)
virtual cudaError_t	cudaGraphicsSubResourceGetMappedArray (struct cudaArray **arrayPtr, struct cudaGraphicsResource *resource, unsigned int arrayIndex, unsigned int mipLevel)
virtual void	addTraceGenerator (trace::TraceGenerator &gen, bool persistent=false)
	Adds a trace generator for the next kernel invocation.
virtual void	clearTraceGenerators ()
	Clear all trace generators.
virtual void	addPTXPass (transforms::Pass &pass)
	Adds a PTX->PTX pass for the next *Module load*.
virtual void	removePTXPass (transforms::Pass &pass)
	removes the specified pass
virtual void	clearPTXPasses ()
	clears all PTX->PTX passes
virtual void	limitWorkerThreads (unsigned int limit=1024)
	Sets a limit on the number of host worker threads to launch when executing a CUDA kernel on a Multi-Core CPU.
virtual void	registerPTXModule (std::istream &stream, const std::string &name)
	Register an istream containing a PTX module.
virtual void	registerTexture (const void *texref, const std::string &moduleName, const std::string &textureName, bool normalize)
	Register a texture with the cuda runtime.
virtual void	clearErrors ()
	Clear all errors in the Cuda Runtime.
virtual void	reset ()
	Reset all CUDA runtime state.
virtual ocelot::PointerMap	contextSwitch (unsigned int destinationDevice, unsigned int sourceDevice)
	Perform a device context switch.
virtual void	unregisterModule (const std::string &name)
	Unregister a module, either PTX or LLVM, not a fatbinary.
virtual void	launch (const std::string &moduleName, const std::string &kernelName)
	Launch a cuda kernel by name.
virtual void	setOptimizationLevel (translator::Translator::OptimizationLevel l)
	Set the optimization level.
virtual void	registerExternalFunction (const std::string &name, void *function)
	Register a callable host function with Ocelot.
virtual void	removeExternalFunction (const std::string &name)
	Remove a previously registered host function.

---

Detailed Description

Cuda runtime context

---

Constructor & Destructor Documentation

cuda::CudaRuntime::CudaRuntime

(

)

cuda::CudaRuntime::~CudaRuntime

(

)

---

Member Function Documentation

void cuda::CudaRuntime::addPTXPass

(

transforms::Pass &

pass

)

[virtual]

Adds a PTX->PTX pass for the next *Module load*.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::addTraceGenerator	(	trace::TraceGenerator &	gen,
		bool	persistent = false
	)		[virtual]

Adds a trace generator for the next kernel invocation.

Parameters:

gen	A reference to the generator being added, it must not be destroyed until the next kernel is executed.
persistent	The trace generator will be associated with all subsequent kernels until clear is called, otherwise it will only be associated with the next kernel.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::clearErrors

(

)

[virtual]

Clear all errors in the Cuda Runtime.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::clearPTXPasses

(

)

[virtual]

clears all PTX->PTX passes

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::clearTraceGenerators

(

)

[virtual]

Clear all trace generators.

Reimplemented from cuda::CudaRuntimeInterface.

ocelot::PointerMap cuda::CudaRuntime::contextSwitch	(	unsigned int	destinationDevice,
		unsigned int	sourceDevice
	)		[virtual]

Perform a device context switch.

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaBindTexture	(	size_t *	offset,
		const struct textureReference *	texref,
		const void *	devPtr,
		const struct cudaChannelFormatDesc *	desc,
		size_t	size = UINT_MAX
	)		[virtual]

binds a texture to a reference and a CUDA memory block

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaBindTexture2D	(	size_t *	offset,
		const struct textureReference *	texref,
		const void *	devPtr,
		const struct cudaChannelFormatDesc *	desc,
		size_t	width,
		size_t	height,
		size_t	pitch
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaBindTextureToArray	(	const struct textureReference *	texref,
		const struct cudaArray *	array,
		const struct cudaChannelFormatDesc *	desc
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaChooseDevice	(	int *	device,
		const struct cudaDeviceProp *	prop
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaConfigureCall	(	dim3	gridDim,
		dim3	blockDim,
		size_t	sharedMem = 0,
		cudaStream_t	stream = 0
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

struct cudaChannelFormatDesc cuda::CudaRuntime::cudaCreateChannelDesc	(	int	x,
		int	y,
		int	z,
		int	w,
		enum cudaChannelFormatKind	f
	)		[read, virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceGetCacheConfig

(

enum cudaFuncCache *

pCacheConfig

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceGetLimit	(	size_t *	pValue,
		enum cudaLimit	limit
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceReset

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceSetCacheConfig

(

enum cudaFuncCache

cacheConfig

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceSetLimit	(	enum cudaLimit	limit,
		size_t	value
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDeviceSynchronize

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaDriverGetVersion

(

int *

driverVersion

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventCreate

(

cudaEvent_t *

event

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventCreateWithFlags	(	cudaEvent_t *	event,
		int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventDestroy

(

cudaEvent_t

event

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventElapsedTime	(	float *	ms,
		cudaEvent_t	start,
		cudaEvent_t	end
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventQuery

(

cudaEvent_t

event

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventRecord	(	cudaEvent_t	event,
		cudaStream_t	stream
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaEventSynchronize

(

cudaEvent_t

event

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaFree

(

void *

devPtr

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaFreeArray

(

struct cudaArray *

array

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaFreeHost

(

void *

ptr

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaFuncGetAttributes	(	struct cudaFuncAttributes *	attr,
		const char *	func
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaFuncSetCacheConfig	(	const char *	func,
		enum cudaFuncCache	cacheConfig
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetChannelDesc	(	struct cudaChannelFormatDesc *	desc,
		const struct cudaArray *	array
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetDevice

(

int *

device

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetDeviceCount

(

int *

count

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetDeviceProperties	(	struct cudaDeviceProp *	prop,
		int	device
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetExportTable	(	const void **	ppExportTable,
		const cudaUUID_t *	pExportTableId
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetLastError

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetSymbolAddress	(	void **	devPtr,
		const char *	symbol
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetSymbolSize	(	size_t *	size,
		const char *	symbol
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetTextureAlignmentOffset	(	size_t *	offset,
		const struct textureReference *	texref
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGetTextureReference	(	const struct textureReference **	texref,
		const char *	symbol
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLMapBufferObject	(	void **	devPtr,
		GLuint	bufObj
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLMapBufferObjectAsync	(	void **	devPtr,
		GLuint	bufObj,
		cudaStream_t	stream
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLRegisterBufferObject

(

GLuint

bufObj

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLSetBufferObjectMapFlags	(	GLuint	bufObj,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLSetGLDevice

(

int

device

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLUnmapBufferObject

(

GLuint

bufObj

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLUnmapBufferObjectAsync	(	GLuint	bufObj,
		cudaStream_t	stream
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGLUnregisterBufferObject

(

GLuint

bufObj

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsGLRegisterBuffer	(	struct cudaGraphicsResource **	resource,
		GLuint	buffer,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsGLRegisterImage	(	struct cudaGraphicsResource **	resource,
		GLuint	image,
		int	target,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsMapResources	(	int	count,
		struct cudaGraphicsResource **	resources,
		cudaStream_t	stream = 0
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsResourceGetMappedPointer	(	void **	devPtr,
		size_t *	size,
		struct cudaGraphicsResource *	resource
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsResourceSetMapFlags	(	struct cudaGraphicsResource *	resource,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsSubResourceGetMappedArray	(	struct cudaArray **	arrayPtr,
		struct cudaGraphicsResource *	resource,
		unsigned int	arrayIndex,
		unsigned int	mipLevel
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsUnmapResources	(	int	count,
		struct cudaGraphicsResource **	resources,
		cudaStream_t	stream = 0
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaGraphicsUnregisterResource

(

struct cudaGraphicsResource *

resource

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaHostAlloc	(	void **	pHost,
		size_t	bytes,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaHostGetDevicePointer	(	void **	pDevice,
		void *	pHost,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaHostGetFlags	(	unsigned int *	pFlags,
		void *	pHost
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaHostRegister	(	void *	pHost,
		size_t	bytes,
		unsigned int	flags
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaHostUnregister

(

void *

pHost

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaLaunch

(

const char *

entry

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMalloc	(	void **	devPtr,
		size_t	size
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMalloc3D	(	struct cudaPitchedPtr *	pitchedDevPtr,
		struct cudaExtent	extent
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMalloc3DArray	(	struct cudaArray **	arrayPtr,
		const struct cudaChannelFormatDesc *	desc,
		struct cudaExtent	extent
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMallocArray	(	struct cudaArray **	array,
		const struct cudaChannelFormatDesc *	desc,
		size_t	width,
		size_t	height = 1
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMallocHost	(	void **	ptr,
		size_t	size
	)		[virtual]

constructs a host-side allocation, returns pointer to mapped region - this allocation is referenced by the mappedPointer()

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMallocPitch	(	void **	devPtr,
		size_t *	pitch,
		size_t	width,
		size_t	height
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy	(	void *	dst,
		const void *	src,
		size_t	count,
		enum cudaMemcpyKind	kind
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy2D	(	void *	dst,
		size_t	dpitch,
		const void *	src,
		size_t	spitch,
		size_t	width,
		size_t	height,
		enum cudaMemcpyKind	kind
	)		[virtual]

perform a 2D memcpy from a dense buffer

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy2DFromArray	(	void *	dst,
		size_t	dpitch,
		const struct cudaArray *	src,
		size_t	wOffset,
		size_t	hOffset,
		size_t	width,
		size_t	height,
		enum cudaMemcpyKind	kind
	)		[virtual]

perform a 2D memcpy from an array

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy2DToArray	(	struct cudaArray *	dst,
		size_t	wOffset,
		size_t	hOffset,
		const void *	src,
		size_t	spitch,
		size_t	width,
		size_t	height,
		enum cudaMemcpyKind	kind
	)		[virtual]

perform a 2D memcpy to an array

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy3D

(

const struct cudaMemcpy3DParms *

p

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpy3DAsync	(	const struct cudaMemcpy3DParms *	p,
		cudaStream_t	stream
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyArrayToArray	(	struct cudaArray *	dst,
		size_t	wOffsetDst,
		size_t	hOffsetDst,
		const struct cudaArray *	src,
		size_t	wOffsetSrc,
		size_t	hOffsetSrc,
		size_t	count,
		enum cudaMemcpyKind	kind
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyAsync	(	void *	dst,
		const void *	src,
		size_t	count,
		enum cudaMemcpyKind	kind,
		cudaStream_t	stream
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyFromArray	(	void *	dst,
		const struct cudaArray *	src,
		size_t	wOffset,
		size_t	hOffset,
		size_t	count,
		enum cudaMemcpyKind	kind
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyFromSymbol	(	void *	dst,
		const char *	symbol,
		size_t	count,
		size_t	offset = 0,
		enum cudaMemcpyKind	kind = cudaMemcpyDeviceToHost
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyToArray	(	struct cudaArray *	dst,
		size_t	wOffset,
		size_t	hOffset,
		const void *	src,
		size_t	count,
		enum cudaMemcpyKind	kind
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemcpyToSymbol	(	const char *	symbol,
		const void *	src,
		size_t	count,
		size_t	offset,
		enum cudaMemcpyKind	kind = cudaMemcpyHostToDevice
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemset	(	void *	devPtr,
		int	value,
		size_t	count
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemset2D	(	void *	devPtr,
		size_t	pitch,
		int	value,
		size_t	width,
		size_t	height
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaMemset3D	(	struct cudaPitchedPtr	pitchedDevPtr,
		int	value,
		struct cudaExtent	extent
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaPeekAtLastError

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

void ** cuda::CudaRuntime::cudaRegisterFatBinary

(

void *

fatCubin

)

[virtual]

registers a CUDA fatbinary and returns a handle for referencing the fat binary

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaRegisterFunction	(	void **	fatCubinHandle,
		const char *	hostFun,
		char *	deviceFun,
		const char *	deviceName,
		int	thread_limit,
		uint3 *	tid,
		uint3 *	bid,
		dim3 *	bDim,
		dim3 *	gDim,
		int *	wSize
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaRegisterShared	(	void **	fatCubinHandle,
		void **	devicePtr
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaRegisterSharedVar	(	void **	fatCubinHandle,
		void **	devicePtr,
		size_t	size,
		size_t	alignment,
		int	storage
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaRegisterTexture	(	void **	fatCubinHandle,
		const struct textureReference *	hostVar,
		const void **	deviceAddress,
		const char *	deviceName,
		int	dim,
		int	norm,
		int	ext
	)		[virtual]

registers a CUDA texture reference

Parameters:

fatCubinHandle
hostVar
deviceAddress
deviceName
dim
norm
ext

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaRegisterVar	(	void **	fatCubinHandle,
		char *	hostVar,
		char *	deviceAddress,
		const char *	deviceName,
		int	ext,
		int	size,
		int	constant,
		int	global
	)		[virtual]

register a CUDA global variable

Parameters:

fatCubinHandle
hostVar
deviceAddress
deviceName
ext
size
constant
global

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaRuntimeGetVersion

(

int *

runtimeVersion

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaSetDevice

(

int

device

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaSetDeviceFlags

(

int

flags

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaSetupArgument	(	const void *	arg,
		size_t	size,
		size_t	offset
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaSetValidDevices	(	int *	device_arr,
		int	len
	)		[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaStreamCreate

(

cudaStream_t *

pStream

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaStreamDestroy

(

cudaStream_t

stream

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaStreamQuery

(

cudaStream_t

stream

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaStreamSynchronize

(

cudaStream_t

stream

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaThreadExit

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaThreadSynchronize

(

void

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

cudaError_t cuda::CudaRuntime::cudaUnbindTexture

(

const struct textureReference *

texref

)

[virtual]

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::cudaUnregisterFatBinary

(

void **

fatCubinHandle

)

[virtual]

unregister a cuda fat binary

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::launch	(	const std::string &	moduleName,
		const std::string &	kernelName
	)		[virtual]

Launch a cuda kernel by name.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::limitWorkerThreads

(

unsigned int

limit = 1024

)

[virtual]

Sets a limit on the number of host worker threads to launch when executing a CUDA kernel on a Multi-Core CPU.

Parameters:

limit

The max number of worker threads to launch per kernel.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::registerExternalFunction	(	const std::string &	name,
		void *	function
	)		[virtual]

Register a callable host function with Ocelot.

This function will be callable as a PTX function.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::registerPTXModule	(	std::istream &	stream,
		const std::string &	name
	)		[virtual]

Register an istream containing a PTX module.

Parameters:

stream	An input stream containing a PTX module
The	name of the module being registered. Must be Unique.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::registerTexture	(	const void *	texref,
		const std::string &	moduleName,
		const std::string &	textureName,
		bool	normalize
	)		[virtual]

Register a texture with the cuda runtime.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::removeExternalFunction

(

const std::string &

name

)

[virtual]

Remove a previously registered host function.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::removePTXPass

(

transforms::Pass &

pass

)

[virtual]

removes the specified pass

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::reset

(

)

[virtual]

Reset all CUDA runtime state.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::setOptimizationLevel

(

translator::Translator::OptimizationLevel

l

)

[virtual]

Set the optimization level.

Reimplemented from cuda::CudaRuntimeInterface.

void cuda::CudaRuntime::unregisterModule

(

const std::string &

name

)

[virtual]

Unregister a module, either PTX or LLVM, not a fatbinary.

Reimplemented from cuda::CudaRuntimeInterface.

---

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

• ocelot/cuda/interface/CudaRuntime.h
• ocelot/cuda/implementation/CudaRuntime.cpp