GPUOcelot
Public Member Functions | Static Public Member Functions | Static Public Attributes | Protected Attributes

cuda::CudaRuntimeInterface Class Reference

#include <CudaRuntimeInterface.h>

Inheritance diagram for cuda::CudaRuntimeInterface:
Inheritance graph
[legend]
Collaboration diagram for cuda::CudaRuntimeInterface:
Collaboration graph
[legend]

List of all members.

Public Member Functions

 CudaRuntimeInterface ()
virtual ~CudaRuntimeInterface ()
const ocelot::OcelotRuntimeocelot () const
 gets the Ocelot runtime object
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)
virtual void cudaRegisterTexture (void **fatCubinHandle, const struct textureReference *hostVar, const void **deviceAddress, const char *deviceName, int dim, int norm, int ext)
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 cudaMalloc3D (struct cudaPitchedPtr *pitchedDevPtr, struct cudaExtent extent)
virtual cudaError_t cudaMalloc3DArray (struct cudaArray **arrayPtr, const struct cudaChannelFormatDesc *desc, struct cudaExtent extent)
virtual cudaError_t cudaMemset3D (struct cudaPitchedPtr pitchedDevPtr, int value, struct cudaExtent extent)
virtual cudaError_t cudaMemcpy3D (const struct cudaMemcpy3DParms *p)
virtual cudaError_t cudaMemcpy3DAsync (const struct cudaMemcpy3DParms *p, cudaStream_t stream)
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 cudaHostRegister (void *pHost, size_t bytes, unsigned int flags)
virtual cudaError_t cudaHostUnregister (void *pHost)
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 cudaMemcpy (void *dst, const void *src, size_t count, enum cudaMemcpyKind kind)
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=cudaMemcpyDeviceToDevice)
virtual cudaError_t cudaMemcpy2D (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t height, enum cudaMemcpyKind kind)
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)
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)
virtual cudaError_t cudaMemcpy2DArrayToArray (struct cudaArray *dst, size_t wOffsetDst, size_t hOffsetDst, const struct cudaArray *src, size_t wOffsetSrc, size_t hOffsetSrc, size_t width, size_t height, enum cudaMemcpyKind kind=cudaMemcpyDeviceToDevice)
virtual cudaError_t cudaMemcpyToSymbol (const char *symbol, const void *src, size_t count, size_t offset=0, 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 cudaMemcpyToArrayAsync (struct cudaArray *dst, size_t wOffset, size_t hOffset, const void *src, size_t count, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpyFromArrayAsync (void *dst, const struct cudaArray *src, size_t wOffset, size_t hOffset, size_t count, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpy2DAsync (void *dst, size_t dpitch, const void *src, size_t spitch, size_t width, size_t height, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpy2DToArrayAsync (struct cudaArray *dst, size_t wOffset, size_t hOffset, const void *src, size_t spitch, size_t width, size_t height, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpy2DFromArrayAsync (void *dst, size_t dpitch, const struct cudaArray *src, size_t wOffset, size_t hOffset, size_t width, size_t height, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpyToSymbolAsync (const char *symbol, const void *src, size_t count, size_t offset, enum cudaMemcpyKind kind, cudaStream_t stream)
virtual cudaError_t cudaMemcpyFromSymbolAsync (void *dst, const char *symbol, size_t count, size_t offset, enum cudaMemcpyKind kind, 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 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)
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 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 cudaStreamWaitEvent (cudaStream_t stream, cudaEvent_t event, unsigned int flags)
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 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 cudaError_t cudaSetDoubleForDevice (double *d)
virtual cudaError_t cudaSetDoubleForHost (double *d)
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 cudaThreadSetLimit (enum cudaLimit limit, size_t value)
virtual cudaError_t cudaDriverGetVersion (int *driverVersion)
virtual cudaError_t cudaRuntimeGetVersion (int *runtimeVersion)
virtual cudaError_t cudaGetExportTable (const void **ppExportTable, const cudaUUID_t *pExportTableId)
virtual void cudaMutexOperation (int lock)
virtual int cudaSynchronizeThreads (void **one, void *two)
virtual void cudaTextureFetch (const void *tex, void *index, int integer, void *val)
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.

Static Public Member Functions

static CudaRuntimeInterfaceget ()

Static Public Attributes

static CudaRuntimeInterfaceinstance = 0

Protected Attributes

ocelot::OcelotRuntime ocelotRuntime
 Ocelot runtime object containing state related to Ocelot.

Detailed Description

Singleton object called directly by CUDA Runtime API wrapper


Constructor & Destructor Documentation

cuda::CudaRuntimeInterface::CudaRuntimeInterface ( )
cuda::CudaRuntimeInterface::~CudaRuntimeInterface ( ) [virtual]

Member Function Documentation

void cuda::CudaRuntimeInterface::addPTXPass ( transforms::Pass pass) [virtual]

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

Reimplemented in cuda::CudaRuntime.

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

Adds a trace generator for the next kernel invocation.

Parameters:
genA reference to the generator being added, it must not be destroyed until the next kernel is executed.
persistentThe trace generator will be associated with all subsequent kernels until clear is called, otherwise it will only be associated with the next kernel.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::clearErrors ( ) [virtual]

Clear all errors in the Cuda Runtime.

Reimplemented in cuda::CudaRuntime, and cuda::TraceGeneratingCudaRuntime.

void cuda::CudaRuntimeInterface::clearPTXPasses ( ) [virtual]

clears all PTX->PTX passes

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::clearTraceGenerators ( ) [virtual]

Clear all trace generators.

Reimplemented in cuda::CudaRuntime.

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

Perform a device context switch.

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::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 in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaConfigureCall ( dim3  gridDim,
dim3  blockDim,
size_t  sharedMem = 0,
cudaStream_t  stream = 0 
) [virtual]
struct cudaChannelFormatDesc cuda::CudaRuntimeInterface::cudaCreateChannelDesc ( int  x,
int  y,
int  z,
int  w,
enum cudaChannelFormatKind  f 
) [read, virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaDeviceGetCacheConfig ( enum cudaFuncCache pCacheConfig) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaDeviceReset ( void  ) [virtual]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaDeviceSetCacheConfig ( enum cudaFuncCache  cacheConfig) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaDeviceSynchronize ( void  ) [virtual]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaDriverGetVersion ( int *  driverVersion) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventCreate ( cudaEvent_t event) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventCreateWithFlags ( cudaEvent_t event,
int  flags 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventDestroy ( cudaEvent_t  event) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventElapsedTime ( float *  ms,
cudaEvent_t  start,
cudaEvent_t  end 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventQuery ( cudaEvent_t  event) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventRecord ( cudaEvent_t  event,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaEventSynchronize ( cudaEvent_t  event) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaFree ( void *  devPtr) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaFreeArray ( struct cudaArray *  array) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaFreeHost ( void *  ptr) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaFuncGetAttributes ( struct cudaFuncAttributes attr,
const char *  func 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaFuncSetCacheConfig ( const char *  func,
enum cudaFuncCache  cacheConfig 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaGetDevice ( int *  device) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetDeviceCount ( int *  count) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetDeviceProperties ( struct cudaDeviceProp prop,
int  device 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetExportTable ( const void **  ppExportTable,
const cudaUUID_t pExportTableId 
) [virtual]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaGetLastError ( void  ) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetSymbolAddress ( void **  devPtr,
const char *  symbol 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetSymbolSize ( size_t size,
const char *  symbol 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGetTextureAlignmentOffset ( size_t offset,
const struct textureReference texref 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaGLSetGLDevice ( int  device) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGLUnmapBufferObject ( GLuint  bufObj) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaGLUnmapBufferObjectAsync ( GLuint  bufObj,
cudaStream_t  stream 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaGraphicsUnregisterResource ( struct cudaGraphicsResource *  resource) [virtual]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaHostAlloc ( void **  pHost,
size_t  bytes,
unsigned int  flags 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaHostGetDevicePointer ( void **  pDevice,
void *  pHost,
unsigned int  flags 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaHostGetFlags ( unsigned int *  pFlags,
void *  pHost 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaHostUnregister ( void *  pHost) [virtual]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaLaunch ( const char *  entry) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMalloc ( void **  devPtr,
size_t  size 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMalloc3D ( struct cudaPitchedPtr pitchedDevPtr,
struct cudaExtent  extent 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMalloc3DArray ( struct cudaArray **  arrayPtr,
const struct cudaChannelFormatDesc desc,
struct cudaExtent  extent 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaMallocHost ( void **  ptr,
size_t  size 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMallocPitch ( void **  devPtr,
size_t pitch,
size_t  width,
size_t  height 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy ( void *  dst,
const void *  src,
size_t  count,
enum cudaMemcpyKind  kind 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy2D ( void *  dst,
size_t  dpitch,
const void *  src,
size_t  spitch,
size_t  width,
size_t  height,
enum cudaMemcpyKind  kind 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy2DArrayToArray ( struct cudaArray *  dst,
size_t  wOffsetDst,
size_t  hOffsetDst,
const struct cudaArray *  src,
size_t  wOffsetSrc,
size_t  hOffsetSrc,
size_t  width,
size_t  height,
enum cudaMemcpyKind  kind = cudaMemcpyDeviceToDevice 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy2DAsync ( void *  dst,
size_t  dpitch,
const void *  src,
size_t  spitch,
size_t  width,
size_t  height,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::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]

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy2DFromArrayAsync ( void *  dst,
size_t  dpitch,
const struct cudaArray *  src,
size_t  wOffset,
size_t  hOffset,
size_t  width,
size_t  height,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::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]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy2DToArrayAsync ( struct cudaArray *  dst,
size_t  wOffset,
size_t  hOffset,
const void *  src,
size_t  spitch,
size_t  width,
size_t  height,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpy3D ( const struct cudaMemcpy3DParms p) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::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 = cudaMemcpyDeviceToDevice 
) [virtual]

Reimplemented in cuda::CudaRuntime.

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

Reimplemented in cuda::CudaRuntime.

cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyFromArrayAsync ( void *  dst,
const struct cudaArray *  src,
size_t  wOffset,
size_t  hOffset,
size_t  count,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyFromSymbol ( void *  dst,
const char *  symbol,
size_t  count,
size_t  offset = 0,
enum cudaMemcpyKind  kind = cudaMemcpyDeviceToHost 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyFromSymbolAsync ( void *  dst,
const char *  symbol,
size_t  count,
size_t  offset,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyToArray ( struct cudaArray *  dst,
size_t  wOffset,
size_t  hOffset,
const void *  src,
size_t  count,
enum cudaMemcpyKind  kind 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyToArrayAsync ( struct cudaArray *  dst,
size_t  wOffset,
size_t  hOffset,
const void *  src,
size_t  count,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyToSymbol ( const char *  symbol,
const void *  src,
size_t  count,
size_t  offset = 0,
enum cudaMemcpyKind  kind = cudaMemcpyHostToDevice 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemcpyToSymbolAsync ( const char *  symbol,
const void *  src,
size_t  count,
size_t  offset,
enum cudaMemcpyKind  kind,
cudaStream_t  stream 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemset ( void *  devPtr,
int  value,
size_t  count 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemset2D ( void *  devPtr,
size_t  pitch,
int  value,
size_t  width,
size_t  height 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaMemset3D ( struct cudaPitchedPtr  pitchedDevPtr,
int  value,
struct cudaExtent  extent 
) [virtual]
void cuda::CudaRuntimeInterface::cudaMutexOperation ( int  lock) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaPeekAtLastError ( void  ) [virtual]

Reimplemented in cuda::CudaRuntime.

void ** cuda::CudaRuntimeInterface::cudaRegisterFatBinary ( void *  fatCubin) [virtual]
void cuda::CudaRuntimeInterface::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]
void cuda::CudaRuntimeInterface::cudaRegisterShared ( void **  fatCubinHandle,
void **  devicePtr 
) [virtual]
void cuda::CudaRuntimeInterface::cudaRegisterSharedVar ( void **  fatCubinHandle,
void **  devicePtr,
size_t  size,
size_t  alignment,
int  storage 
) [virtual]
void cuda::CudaRuntimeInterface::cudaRegisterTexture ( void **  fatCubinHandle,
const struct textureReference hostVar,
const void **  deviceAddress,
const char *  deviceName,
int  dim,
int  norm,
int  ext 
) [virtual]
void cuda::CudaRuntimeInterface::cudaRegisterVar ( void **  fatCubinHandle,
char *  hostVar,
char *  deviceAddress,
const char *  deviceName,
int  ext,
int  size,
int  constant,
int  global 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaRuntimeGetVersion ( int *  runtimeVersion) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetDevice ( int  device) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetDeviceFlags ( int  flags) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetDoubleForDevice ( double *  d) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetDoubleForHost ( double *  d) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetupArgument ( const void *  arg,
size_t  size,
size_t  offset 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaSetValidDevices ( int *  device_arr,
int  len 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaStreamCreate ( cudaStream_t pStream) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaStreamDestroy ( cudaStream_t  stream) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaStreamQuery ( cudaStream_t  stream) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaStreamSynchronize ( cudaStream_t  stream) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaStreamWaitEvent ( cudaStream_t  stream,
cudaEvent_t  event,
unsigned int  flags 
) [virtual]
int cuda::CudaRuntimeInterface::cudaSynchronizeThreads ( void **  one,
void *  two 
) [virtual]
void cuda::CudaRuntimeInterface::cudaTextureFetch ( const void *  tex,
void *  index,
int  integer,
void *  val 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaThreadExit ( void  ) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaThreadSetLimit ( enum cudaLimit  limit,
size_t  value 
) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaThreadSynchronize ( void  ) [virtual]
cudaError_t cuda::CudaRuntimeInterface::cudaUnbindTexture ( const struct textureReference texref) [virtual]

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::cudaUnregisterFatBinary ( void **  fatCubinHandle) [virtual]
cuda::CudaRuntimeInterface * cuda::CudaRuntimeInterface::get ( ) [static]
void cuda::CudaRuntimeInterface::launch ( const std::string &  moduleName,
const std::string &  kernelName 
) [virtual]

Launch a cuda kernel by name.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::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:
limitThe max number of worker threads to launch per kernel.

Reimplemented in cuda::CudaRuntime, and cuda::TraceGeneratingCudaRuntime.

const ocelot::OcelotRuntime& cuda::CudaRuntimeInterface::ocelot ( ) const

gets the Ocelot runtime object

void cuda::CudaRuntimeInterface::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 in cuda::CudaRuntime.

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

Register an istream containing a PTX module.

Parameters:
streamAn input stream containing a PTX module
Thename of the module being registered. Must be Unique.

Reimplemented in cuda::CudaRuntime, and cuda::TraceGeneratingCudaRuntime.

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

Register a texture with the cuda runtime.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::removeExternalFunction ( const std::string &  name) [virtual]

Remove a previously registered host function.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::removePTXPass ( transforms::Pass pass) [virtual]

removes the specified pass

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::reset ( ) [virtual]

Reset all CUDA runtime state.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::setOptimizationLevel ( translator::Translator::OptimizationLevel  l) [virtual]

Set the optimization level.

Reimplemented in cuda::CudaRuntime.

void cuda::CudaRuntimeInterface::unregisterModule ( const std::string &  name) [virtual]

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

Reimplemented in cuda::CudaRuntime.


Member Data Documentation

Ocelot runtime object containing state related to Ocelot.


The documentation for this class was generated from the following files:
 All Classes Namespaces Files Functions Variables Typedefs Enumerations Enumerator Friends Defines