4.9. Memory Management

This section describes the memory management functions of the CUDA runtime application programming interface.

Some functions have overloaded C++ API template versions documented separately in the C++ API Routines module.

Functions

__host__cudaError_t cudaArrayGetInfo ( cudaChannelFormatDesc* desc, cudaExtent* extent, unsigned int* flags, cudaArray_t array )
Gets info about the specified cudaArray.
__host____device__cudaError_t cudaFree ( void* devPtr )
Frees memory on the device.
__host__cudaError_t cudaFreeArray ( cudaArray_t array )
Frees an array on the device.
__host__cudaError_t cudaFreeHost ( void* ptr )
Frees page-locked memory.
__host__cudaError_t cudaFreeMipmappedArray ( cudaMipmappedArray_t mipmappedArray )
Frees a mipmapped array on the device.
__host__cudaError_t cudaGetMipmappedArrayLevel ( cudaArray_t* levelArray, cudaMipmappedArray_const_t mipmappedArray, unsigned int  level )
Gets a mipmap level of a CUDA mipmapped array.
__host__cudaError_t cudaGetSymbolAddress ( void** devPtr, const void* symbol )
Finds the address associated with a CUDA symbol.
__host__cudaError_t cudaGetSymbolSize ( size_t* size, const void* symbol )
Finds the size of the object associated with a CUDA symbol.
__host__cudaError_t cudaHostAlloc ( void** pHost, size_t size, unsigned int  flags )
Allocates page-locked memory on the host.
__host__cudaError_t cudaHostGetDevicePointer ( void** pDevice, void* pHost, unsigned int  flags )
Passes back device pointer of mapped host memory allocated by cudaHostAlloc or registered by cudaHostRegister.
__host__cudaError_t cudaHostGetFlags ( unsigned int* pFlags, void* pHost )
Passes back flags used to allocate pinned host memory allocated by cudaHostAlloc.
__host__cudaError_t cudaHostRegister ( void* ptr, size_t size, unsigned int  flags )
Registers an existing host memory range for use by CUDA.
__host__cudaError_t cudaHostUnregister ( void* ptr )
Unregisters a memory range that was registered with cudaHostRegister.
__host____device__cudaError_t cudaMalloc ( void** devPtr, size_t size )
Allocate memory on the device.
__host__cudaError_t cudaMalloc3D ( cudaPitchedPtr* pitchedDevPtr, cudaExtent extent )
Allocates logical 1D, 2D, or 3D memory objects on the device.
__host__cudaError_t cudaMalloc3DArray ( cudaArray_t* array, const cudaChannelFormatDesc* desc, cudaExtent extent, unsigned int  flags = 0 )
Allocate an array on the device.
__host__cudaError_t cudaMallocArray ( cudaArray_t* array, const cudaChannelFormatDesc* desc, size_t width, size_t height = 0, unsigned int  flags = 0 )
Allocate an array on the device.
__host__cudaError_t cudaMallocHost ( void** ptr, size_t size )
Allocates page-locked memory on the host.
__host__cudaError_t cudaMallocManaged ( void** devPtr, size_t size, unsigned int  flags = cudaMemAttachGlobal )
Allocates memory that will be automatically managed by the Unified Memory system.
__host__cudaError_t cudaMallocMipmappedArray ( cudaMipmappedArray_t* mipmappedArray, const cudaChannelFormatDesc* desc, cudaExtent extent, unsigned int  numLevels, unsigned int  flags = 0 )
Allocate a mipmapped array on the device.
__host__cudaError_t cudaMallocPitch ( void** devPtr, size_t* pitch, size_t width, size_t height )
Allocates pitched memory on the device.
__host__cudaError_t cudaMemAdvise ( const void* devPtr, size_t count, cudaMemoryAdvise advice, int  device )
Advise about the usage of a given memory range.
__host__cudaError_t cudaMemGetInfo ( size_t* free, size_t* total )
Gets free and total device memory.
__host__cudaError_t cudaMemPrefetchAsync ( const void* devPtr, size_t count, int  dstDevice, cudaStream_t stream = 0 )
Prefetches memory to the specified destination device.
__host__cudaError_t cudaMemRangeGetAttribute ( void* data, size_t dataSize, cudaMemRangeAttribute attribute, const void* devPtr, size_t count )
Query an attribute of a given memory range.
__host__cudaError_t cudaMemRangeGetAttributes ( void** data, size_t* dataSizes, cudaMemRangeAttribute ** attributes, size_t numAttributes, const void* devPtr, size_t count )
Query attributes of a given memory range.
__host__cudaError_t cudaMemcpy ( void* dst, const void* src, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2D ( void* dst, size_t dpitch, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2DArrayToArray ( cudaArray_t dst, size_t wOffsetDst, size_t hOffsetDst, cudaArray_const_t src, size_t wOffsetSrc, size_t hOffsetSrc, size_t width, size_t height, cudaMemcpyKind kind = cudaMemcpyDeviceToDevice )
Copies data between host and device.
__host____device__cudaError_t cudaMemcpy2DAsync ( void* dst, size_t dpitch, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2DFromArray ( void* dst, size_t dpitch, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2DFromArrayAsync ( void* dst, size_t dpitch, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2DToArray ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpy2DToArrayAsync ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpy3D ( const cudaMemcpy3DParms* p )
Copies data between 3D objects.
__host____device__cudaError_t cudaMemcpy3DAsync ( const cudaMemcpy3DParms* p, cudaStream_t stream = 0 )
Copies data between 3D objects.
__host__cudaError_t cudaMemcpy3DPeer ( const cudaMemcpy3DPeerParms* p )
Copies memory between devices.
__host__cudaError_t cudaMemcpy3DPeerAsync ( const cudaMemcpy3DPeerParms* p, cudaStream_t stream = 0 )
Copies memory between devices asynchronously.
__host__cudaError_t cudaMemcpyArrayToArray ( cudaArray_t dst, size_t wOffsetDst, size_t hOffsetDst, cudaArray_const_t src, size_t wOffsetSrc, size_t hOffsetSrc, size_t count, cudaMemcpyKind kind = cudaMemcpyDeviceToDevice )
Copies data between host and device.
__host____device__cudaError_t cudaMemcpyAsync ( void* dst, const void* src, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpyFromArray ( void* dst, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpyFromArrayAsync ( void* dst, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpyFromSymbol ( void* dst, const void* symbol, size_t count, size_t offset = 0, cudaMemcpyKind kind = cudaMemcpyDeviceToHost )
Copies data from the given symbol on the device.
__host__cudaError_t cudaMemcpyFromSymbolAsync ( void* dst, const void* symbol, size_t count, size_t offset, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data from the given symbol on the device.
__host__cudaError_t cudaMemcpyPeer ( void* dst, int  dstDevice, const void* src, int  srcDevice, size_t count )
Copies memory between two devices.
__host__cudaError_t cudaMemcpyPeerAsync ( void* dst, int  dstDevice, const void* src, int  srcDevice, size_t count, cudaStream_t stream = 0 )
Copies memory between two devices asynchronously.
__host__cudaError_t cudaMemcpyToArray ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
__host__cudaError_t cudaMemcpyToArrayAsync ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
__host__cudaError_t cudaMemcpyToSymbol ( const void* symbol, const void* src, size_t count, size_t offset = 0, cudaMemcpyKind kind = cudaMemcpyHostToDevice )
Copies data to the given symbol on the device.
__host__cudaError_t cudaMemcpyToSymbolAsync ( const void* symbol, const void* src, size_t count, size_t offset, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data to the given symbol on the device.
__host__cudaError_t cudaMemset ( void* devPtr, int  value, size_t count )
Initializes or sets device memory to a value.
__host__cudaError_t cudaMemset2D ( void* devPtr, size_t pitch, int  value, size_t width, size_t height )
Initializes or sets device memory to a value.
__host____device__cudaError_t cudaMemset2DAsync ( void* devPtr, size_t pitch, int  value, size_t width, size_t height, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
__host__cudaError_t cudaMemset3D ( cudaPitchedPtr pitchedDevPtr, int  value, cudaExtent extent )
Initializes or sets device memory to a value.
__host____device__cudaError_t cudaMemset3DAsync ( cudaPitchedPtr pitchedDevPtr, int  value, cudaExtent extent, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
__host____device__cudaError_t cudaMemsetAsync ( void* devPtr, int  value, size_t count, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
__host__cudaExtent make_cudaExtent ( size_t w, size_t h, size_t d )
Returns a cudaExtent based on input parameters.
__host__cudaPitchedPtr make_cudaPitchedPtr ( void* d, size_t p, size_t xsz, size_t ysz )
Returns a cudaPitchedPtr based on input parameters.
__host__cudaPos make_cudaPos ( size_t x, size_t y, size_t z )
Returns a cudaPos based on input parameters.

Functions

__host__cudaError_t cudaArrayGetInfo ( cudaChannelFormatDesc* desc, cudaExtent* extent, unsigned int* flags, cudaArray_t array )
Gets info about the specified cudaArray.
Parameters
desc
- Returned array type
extent
- Returned array shape. 2D arrays will have depth of zero
flags
- Returned array flags
array
- The cudaArray to get info for
Description

Returns in *desc, *extent and *flags respectively, the type, shape and flags of array.

Any of *desc, *extent and *flags may be specified as NULL.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cuArrayGetDescriptor, cuArray3DGetDescriptor

__host____device__cudaError_t cudaFree ( void* devPtr )
Frees memory on the device.
Parameters
devPtr
- Device pointer to memory to free
Description

Frees the memory space pointed to by devPtr, which must have been returned by a previous call to cudaMalloc() or cudaMallocPitch(). Otherwise, or if cudaFree(devPtr) has already been called before, an error is returned. If devPtr is 0, no operation is performed. cudaFree() returns cudaErrorInvalidDevicePointer in case of failure.

The device version of cudaFree cannot be used with a *devPtr allocated using the host API, and vice versa.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaMallocArray, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaMalloc3D, cudaMalloc3DArray, cudaHostAlloc, cuMemFree

__host__cudaError_t cudaFreeArray ( cudaArray_t array )
Frees an array on the device.
Parameters
array
- Pointer to array to free
Description

Frees the CUDA array array, which must have been * returned by a previous call to cudaMallocArray(). If cudaFreeArray(array) has already been called before, cudaErrorInvalidValue is returned. If devPtr is 0, no operation is performed.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaFree, cudaMallocArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, cuArrayDestroy

__host__cudaError_t cudaFreeHost ( void* ptr )
Frees page-locked memory.
Parameters
ptr
- Pointer to memory to free
Description

Frees the memory space pointed to by hostPtr, which must have been returned by a previous call to cudaMallocHost() or cudaHostAlloc().

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaFree, cudaMallocArray, cudaFreeArray, cudaMallocHost ( C API), cudaMalloc3D, cudaMalloc3DArray, cudaHostAlloc, cuMemFreeHost

__host__cudaError_t cudaFreeMipmappedArray ( cudaMipmappedArray_t mipmappedArray )
Frees a mipmapped array on the device.
Parameters
mipmappedArray
- Pointer to mipmapped array to free
Description

Frees the CUDA mipmapped array mipmappedArray, which must have been returned by a previous call to cudaMallocMipmappedArray(). If cudaFreeMipmappedArray(mipmappedArray) has already been called before, cudaErrorInvalidValue is returned.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaFree, cudaMallocArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, cuMipmappedArrayDestroy

__host__cudaError_t cudaGetMipmappedArrayLevel ( cudaArray_t* levelArray, cudaMipmappedArray_const_t mipmappedArray, unsigned int  level )
Gets a mipmap level of a CUDA mipmapped array.
Parameters
levelArray
- Returned mipmap level CUDA array
mipmappedArray
- CUDA mipmapped array
level
- Mipmap level
Description

Returns in *levelArray a CUDA array that represents a single mipmap level of the CUDA mipmapped array mipmappedArray.

If level is greater than the maximum number of levels in this mipmapped array, cudaErrorInvalidValue is returned.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc3D, cudaMalloc, cudaMallocPitch, cudaFree, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, make_cudaExtent, cuMipmappedArrayGetLevel

__host__cudaError_t cudaGetSymbolAddress ( void** devPtr, const void* symbol )
Finds the address associated with a CUDA symbol.
Parameters
devPtr
- Return device pointer associated with symbol
symbol
- Device symbol address
Description

Returns in *devPtr the address of symbol symbol on the device. symbol is a variable that resides in global or constant memory space. If symbol cannot be found, or if symbol is not declared in the global or constant memory space, *devPtr is unchanged and the error cudaErrorInvalidSymbol is returned.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaGetSymbolAddress ( C++ API), cudaGetSymbolSize ( C API), cuModuleGetGlobal

__host__cudaError_t cudaGetSymbolSize ( size_t* size, const void* symbol )
Finds the size of the object associated with a CUDA symbol.
Parameters
size
- Size of object associated with symbol
symbol
- Device symbol address
Description

Returns in *size the size of symbol symbol. symbol is a variable that resides in global or constant memory space. If symbol cannot be found, or if symbol is not declared in global or constant memory space, *size is unchanged and the error cudaErrorInvalidSymbol is returned.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaGetSymbolAddress ( C API), cudaGetSymbolSize ( C++ API), cuModuleGetGlobal

__host__cudaError_t cudaHostAlloc ( void** pHost, size_t size, unsigned int  flags )
Allocates page-locked memory on the host.
Parameters
pHost
- Device pointer to allocated memory
size
- Requested allocation size in bytes
flags
- Requested properties of allocated memory
Description

Allocates size bytes of host memory that is page-locked and accessible to the device. The driver tracks the virtual memory ranges allocated with this function and automatically accelerates calls to functions such as cudaMemcpy(). Since the memory can be accessed directly by the device, it can be read or written with much higher bandwidth than pageable memory obtained with functions such as malloc(). Allocating excessive amounts of pinned memory may degrade system performance, since it reduces the amount of memory available to the system for paging. As a result, this function is best used sparingly to allocate staging areas for data exchange between host and device.

The flags parameter enables different options to be specified that affect the allocation, as follows.

  • cudaHostAllocDefault: This flag's value is defined to be 0 and causes cudaHostAlloc() to emulate cudaMallocHost().

  • cudaHostAllocPortable: The memory returned by this call will be considered as pinned memory by all CUDA contexts, not just the one that performed the allocation.

  • cudaHostAllocMapped: Maps the allocation into the CUDA address space. The device pointer to the memory may be obtained by calling cudaHostGetDevicePointer().

  • cudaHostAllocWriteCombined: Allocates the memory as write-combined (WC). WC memory can be transferred across the PCI Express bus more quickly on some system configurations, but cannot be read efficiently by most CPUs. WC memory is a good option for buffers that will be written by the CPU and read by the device via mapped pinned memory or host->device transfers.

All of these flags are orthogonal to one another: a developer may allocate memory that is portable, mapped and/or write-combined with no restrictions.

cudaSetDeviceFlags() must have been called with the cudaDeviceMapHost flag in order for the cudaHostAllocMapped flag to have any effect.

The cudaHostAllocMapped flag may be specified on CUDA contexts for devices that do not support mapped pinned memory. The failure is deferred to cudaHostGetDevicePointer() because the memory may be mapped into other CUDA contexts via the cudaHostAllocPortable flag.

Memory allocated by this function must be freed with cudaFreeHost().

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaSetDeviceFlags, cudaMallocHost ( C API), cudaFreeHost, cuMemHostAlloc

__host__cudaError_t cudaHostGetDevicePointer ( void** pDevice, void* pHost, unsigned int  flags )
Passes back device pointer of mapped host memory allocated by cudaHostAlloc or registered by cudaHostRegister.
Parameters
pDevice
- Returned device pointer for mapped memory
pHost
- Requested host pointer mapping
flags
- Flags for extensions (must be 0 for now)
Description

Passes back the device pointer corresponding to the mapped, pinned host buffer allocated by cudaHostAlloc() or registered by cudaHostRegister().

cudaHostGetDevicePointer() will fail if the cudaDeviceMapHost flag was not specified before deferred context creation occurred, or if called on a device that does not support mapped, pinned memory.

For devices that have a non-zero value for the device attribute cudaDevAttrCanUseHostPointerForRegisteredMem, the memory can also be accessed from the device using the host pointer pHost. The device pointer returned by cudaHostGetDevicePointer() may or may not match the original host pointer pHost and depends on the devices visible to the application. If all devices visible to the application have a non-zero value for the device attribute, the device pointer returned by cudaHostGetDevicePointer() will match the original pointer pHost. If any device visible to the application has a zero value for the device attribute, the device pointer returned by cudaHostGetDevicePointer() will not match the original host pointer pHost, but it will be suitable for use on all devices provided Unified Virtual Addressing is enabled. In such systems, it is valid to access the memory using either pointer on devices that have a non-zero value for the device attribute. Note however that such devices should access the memory using only of the two pointers and not both.

flags provides for future releases. For now, it must be set to 0.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaSetDeviceFlags, cudaHostAlloc, cuMemHostGetDevicePointer

__host__cudaError_t cudaHostGetFlags ( unsigned int* pFlags, void* pHost )
Passes back flags used to allocate pinned host memory allocated by cudaHostAlloc.
Parameters
pFlags
- Returned flags word
pHost
- Host pointer
Description

cudaHostGetFlags() will fail if the input pointer does not reside in an address range allocated by cudaHostAlloc().

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaHostAlloc, cuMemHostGetFlags

__host__cudaError_t cudaHostRegister ( void* ptr, size_t size, unsigned int  flags )
Registers an existing host memory range for use by CUDA.
Parameters
ptr
- Host pointer to memory to page-lock
size
- Size in bytes of the address range to page-lock in bytes
flags
- Flags for allocation request
Description

Page-locks the memory range specified by ptr and size and maps it for the device(s) as specified by flags. This memory range also is added to the same tracking mechanism as cudaHostAlloc() to automatically accelerate calls to functions such as cudaMemcpy(). Since the memory can be accessed directly by the device, it can be read or written with much higher bandwidth than pageable memory that has not been registered. Page-locking excessive amounts of memory may degrade system performance, since it reduces the amount of memory available to the system for paging. As a result, this function is best used sparingly to register staging areas for data exchange between host and device.

cudaHostRegister is not supported on non I/O coherent devices.

The flags parameter enables different options to be specified that affect the allocation, as follows.

  • cudaHostRegisterDefault: On a system with unified virtual addressing, the memory will be both mapped and portable. On a system with no unified virtual addressing, the memory will be neither mapped nor portable.

  • cudaHostRegisterPortable: The memory returned by this call will be considered as pinned memory by all CUDA contexts, not just the one that performed the allocation.

  • cudaHostRegisterIoMemory: The passed memory pointer is treated as pointing to some memory-mapped I/O space, e.g. belonging to a third-party PCIe device, and it will marked as non cache-coherent and contiguous.

All of these flags are orthogonal to one another: a developer may page-lock memory that is portable or mapped with no restrictions.

The CUDA context must have been created with the cudaMapHost flag in order for the cudaHostRegisterMapped flag to have any effect.

The cudaHostRegisterMapped flag may be specified on CUDA contexts for devices that do not support mapped pinned memory. The failure is deferred to cudaHostGetDevicePointer() because the memory may be mapped into other CUDA contexts via the cudaHostRegisterPortable flag.

For devices that have a non-zero value for the device attribute cudaDevAttrCanUseHostPointerForRegisteredMem, the memory can also be accessed from the device using the host pointer ptr. The device pointer returned by cudaHostGetDevicePointer() may or may not match the original host pointer ptr and depends on the devices visible to the application. If all devices visible to the application have a non-zero value for the device attribute, the device pointer returned by cudaHostGetDevicePointer() will match the original pointer ptr. If any device visible to the application has a zero value for the device attribute, the device pointer returned by cudaHostGetDevicePointer() will not match the original host pointer ptr, but it will be suitable for use on all devices provided Unified Virtual Addressing is enabled. In such systems, it is valid to access the memory using either pointer on devices that have a non-zero value for the device attribute. Note however that such devices should access the memory using only of the two pointers and not both.

The memory page-locked by this function must be unregistered with cudaHostUnregister().

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaHostUnregister, cudaHostGetFlags, cudaHostGetDevicePointer, cuMemHostRegister

__host__cudaError_t cudaHostUnregister ( void* ptr )
Unregisters a memory range that was registered with cudaHostRegister.
Parameters
ptr
- Host pointer to memory to unregister
Description

Unmaps the memory range whose base address is specified by ptr, and makes it pageable again.

The base address must be the same one specified to cudaHostRegister().

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaHostUnregister, cuMemHostUnregister

__host____device__cudaError_t cudaMalloc ( void** devPtr, size_t size )
Allocate memory on the device.
Parameters
devPtr
- Pointer to allocated device memory
size
- Requested allocation size in bytes
Description

Allocates size bytes of linear memory on the device and returns in *devPtr a pointer to the allocated memory. The allocated memory is suitably aligned for any kind of variable. The memory is not cleared. cudaMalloc() returns cudaErrorMemoryAllocation in case of failure.

The device version of cudaFree cannot be used with a *devPtr allocated using the host API, and vice versa.

See also:

cudaMallocPitch, cudaFree, cudaMallocArray, cudaFreeArray, cudaMalloc3D, cudaMalloc3DArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, cuMemAlloc

__host__cudaError_t cudaMalloc3D ( cudaPitchedPtr* pitchedDevPtr, cudaExtent extent )
Allocates logical 1D, 2D, or 3D memory objects on the device.
Parameters
pitchedDevPtr
- Pointer to allocated pitched device memory
extent
- Requested allocation size (width field in bytes)
Description

Allocates at least width * height * depth bytes of linear memory on the device and returns a cudaPitchedPtr in which ptr is a pointer to the allocated memory. The function may pad the allocation to ensure hardware alignment requirements are met. The pitch returned in the pitch field of pitchedDevPtr is the width in bytes of the allocation.

The returned cudaPitchedPtr contains additional fields xsize and ysize, the logical width and height of the allocation, which are equivalent to the width and heightextent parameters provided by the programmer during allocation.

For allocations of 2D and 3D objects, it is highly recommended that programmers perform allocations using cudaMalloc3D() or cudaMallocPitch(). Due to alignment restrictions in the hardware, this is especially true if the application will be performing memory copies involving 2D or 3D objects (whether linear memory or CUDA arrays).

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMallocPitch, cudaFree, cudaMemcpy3D, cudaMemset3D, cudaMalloc3DArray, cudaMallocArray, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, make_cudaPitchedPtr, make_cudaExtent, cuMemAllocPitch

__host__cudaError_t cudaMalloc3DArray ( cudaArray_t* array, const cudaChannelFormatDesc* desc, cudaExtent extent, unsigned int  flags = 0 )
Allocate an array on the device.
Parameters
array
- Pointer to allocated array in device memory
desc
- Requested channel format
extent
- Requested allocation size (width field in elements)
flags
- Flags for extensions
Description

Allocates a CUDA array according to the cudaChannelFormatDesc structure desc and returns a handle to the new CUDA array in *array.

The cudaChannelFormatDesc is defined as:

‎    struct cudaChannelFormatDesc {
              int x, y, z, w;
              enum cudaChannelFormatKind 
                  f;
          };
where cudaChannelFormatKind is one of cudaChannelFormatKindSigned, cudaChannelFormatKindUnsigned, or cudaChannelFormatKindFloat.

cudaMalloc3DArray() can allocate the following:

  • A 1D array is allocated if the height and depth extents are both zero.

  • A 2D array is allocated if only the depth extent is zero.

  • A 3D array is allocated if all three extents are non-zero.

  • A 1D layered CUDA array is allocated if only the height extent is zero and the cudaArrayLayered flag is set. Each layer is a 1D array. The number of layers is determined by the depth extent.

  • A 2D layered CUDA array is allocated if all three extents are non-zero and the cudaArrayLayered flag is set. Each layer is a 2D array. The number of layers is determined by the depth extent.

  • A cubemap CUDA array is allocated if all three extents are non-zero and the cudaArrayCubemap flag is set. Width must be equal to height, and depth must be six. A cubemap is a special type of 2D layered CUDA array, where the six layers represent the six faces of a cube. The order of the six layers in memory is the same as that listed in cudaGraphicsCubeFace.

  • A cubemap layered CUDA array is allocated if all three extents are non-zero, and both, cudaArrayCubemap and cudaArrayLayered flags are set. Width must be equal to height, and depth must be a multiple of six. A cubemap layered CUDA array is a special type of 2D layered CUDA array that consists of a collection of cubemaps. The first six layers represent the first cubemap, the next six layers form the second cubemap, and so on.

The flags parameter enables different options to be specified that affect the allocation, as follows.

  • cudaArrayDefault: This flag's value is defined to be 0 and provides default array allocation

  • cudaArrayLayered: Allocates a layered CUDA array, with the depth extent indicating the number of layers

  • cudaArrayCubemap: Allocates a cubemap CUDA array. Width must be equal to height, and depth must be six. If the cudaArrayLayered flag is also set, depth must be a multiple of six.

  • cudaArraySurfaceLoadStore: Allocates a CUDA array that could be read from or written to using a surface reference.

  • cudaArrayTextureGather: This flag indicates that texture gather operations will be performed on the CUDA array. Texture gather can only be performed on 2D CUDA arrays.

The width, height and depth extents must meet certain size requirements as listed in the following table. All values are specified in elements.

Note that 2D CUDA arrays have different size requirements if the cudaArrayTextureGather flag is set. In that case, the valid range for (width, height, depth) is ((1,maxTexture2DGather[0]), (1,maxTexture2DGather[1]), 0).

CUDA array type Valid extents that must always be met {(width range in elements), (height range), (depth range)} Valid extents with cudaArraySurfaceLoadStore set {(width range in elements), (height range), (depth range)}
1D { (1,maxTexture1D), 0, 0 } { (1,maxSurface1D), 0, 0 }
2D { (1,maxTexture2D[0]), (1,maxTexture2D[1]), 0 } { (1,maxSurface2D[0]), (1,maxSurface2D[1]), 0 }
3D { (1,maxTexture3D[0]), (1,maxTexture3D[1]), (1,maxTexture3D[2]) } OR { (1,maxTexture3DAlt[0]), (1,maxTexture3DAlt[1]), (1,maxTexture3DAlt[2]) } { (1,maxSurface3D[0]), (1,maxSurface3D[1]), (1,maxSurface3D[2]) }
1D Layered { (1,maxTexture1DLayered[0]), 0, (1,maxTexture1DLayered[1]) } { (1,maxSurface1DLayered[0]), 0, (1,maxSurface1DLayered[1]) }
2D Layered { (1,maxTexture2DLayered[0]), (1,maxTexture2DLayered[1]), (1,maxTexture2DLayered[2]) } { (1,maxSurface2DLayered[0]), (1,maxSurface2DLayered[1]), (1,maxSurface2DLayered[2]) }
Cubemap { (1,maxTextureCubemap), (1,maxTextureCubemap), 6 } { (1,maxSurfaceCubemap), (1,maxSurfaceCubemap), 6 }
Cubemap Layered { (1,maxTextureCubemapLayered[0]), (1,maxTextureCubemapLayered[0]), (1,maxTextureCubemapLayered[1]) } { (1,maxSurfaceCubemapLayered[0]), (1,maxSurfaceCubemapLayered[0]), (1,maxSurfaceCubemapLayered[1]) }

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc3D, cudaMalloc, cudaMallocPitch, cudaFree, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, make_cudaExtent, cuArray3DCreate

__host__cudaError_t cudaMallocArray ( cudaArray_t* array, const cudaChannelFormatDesc* desc, size_t width, size_t height = 0, unsigned int  flags = 0 )
Allocate an array on the device.
Parameters
array
- Pointer to allocated array in device memory
desc
- Requested channel format
width
- Requested array allocation width
height
- Requested array allocation height
flags
- Requested properties of allocated array
Description

Allocates a CUDA array according to the cudaChannelFormatDesc structure desc and returns a handle to the new CUDA array in *array.

The cudaChannelFormatDesc is defined as:

‎    struct cudaChannelFormatDesc {
              int x, y, z, w;
          enum cudaChannelFormatKind 
                  f;
          };
where cudaChannelFormatKind is one of cudaChannelFormatKindSigned, cudaChannelFormatKindUnsigned, or cudaChannelFormatKindFloat.

The flags parameter enables different options to be specified that affect the allocation, as follows.

width and height must meet certain size requirements. See cudaMalloc3DArray() for more details.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaFree, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaMalloc3D, cudaMalloc3DArray, cudaHostAlloc, cuArrayCreate

__host__cudaError_t cudaMallocHost ( void** ptr, size_t size )
Allocates page-locked memory on the host.
Parameters
ptr
- Pointer to allocated host memory
size
- Requested allocation size in bytes
Description

Allocates size bytes of host memory that is page-locked and accessible to the device. The driver tracks the virtual memory ranges allocated with this function and automatically accelerates calls to functions such as cudaMemcpy*(). Since the memory can be accessed directly by the device, it can be read or written with much higher bandwidth than pageable memory obtained with functions such as malloc(). Allocating excessive amounts of memory with cudaMallocHost() may degrade system performance, since it reduces the amount of memory available to the system for paging. As a result, this function is best used sparingly to allocate staging areas for data exchange between host and device.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaMallocPitch, cudaMallocArray, cudaMalloc3D, cudaMalloc3DArray, cudaHostAlloc, cudaFree, cudaFreeArray, cudaMallocHost ( C++ API), cudaFreeHost, cudaHostAlloc, cuMemAllocHost

__host__cudaError_t cudaMallocManaged ( void** devPtr, size_t size, unsigned int  flags = cudaMemAttachGlobal )
Allocates memory that will be automatically managed by the Unified Memory system.
Parameters
devPtr
- Pointer to allocated device memory
size
- Requested allocation size in bytes
flags
- Must be either cudaMemAttachGlobal or cudaMemAttachHost (defaults to cudaMemAttachGlobal)
Description

Allocates size bytes of managed memory on the device and returns in *devPtr a pointer to the allocated memory. If the device doesn't support allocating managed memory, cudaErrorNotSupported is returned. Support for managed memory can be queried using the device attribute cudaDevAttrManagedMemory. The allocated memory is suitably aligned for any kind of variable. The memory is not cleared. If size is 0, cudaMallocManaged returns cudaErrorInvalidValue. The pointer is valid on the CPU and on all GPUs in the system that support managed memory. All accesses to this pointer must obey the Unified Memory programming model.

flags specifies the default stream association for this allocation. flags must be one of cudaMemAttachGlobal or cudaMemAttachHost. The default value for flags is cudaMemAttachGlobal. If cudaMemAttachGlobal is specified, then this memory is accessible from any stream on any device. If cudaMemAttachHost is specified, then the allocation should not be accessed from devices that have a zero value for the device attribute cudaDevAttrConcurrentManagedAccess; an explicit call to cudaStreamAttachMemAsync will be required to enable access on such devices.

If the association is later changed via cudaStreamAttachMemAsync to a single stream, the default association, as specifed during cudaMallocManaged, is restored when that stream is destroyed. For __managed__ variables, the default association is always cudaMemAttachGlobal. Note that destroying a stream is an asynchronous operation, and as a result, the change to default association won't happen until all work in the stream has completed.

Memory allocated with cudaMallocManaged should be released with cudaFree.

Device memory oversubscription is possible for GPUs that have a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess. Managed memory on such GPUs may be evicted from device memory to host memory at any time by the Unified Memory driver in order to make room for other allocations.

In a multi-GPU system where all GPUs have a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess, managed memory may not be populated when this API returns and instead may be populated on access. In such systems, managed memory can migrate to any processor's memory at any time. The Unified Memory driver will employ heuristics to maintain data locality and prevent excessive page faults to the extent possible. The application can also guide the driver about memory usage patterns via cudaMemAdvise. The application can also explicitly migrate memory to a desired processor's memory via cudaMemPrefetchAsync.

In a multi-GPU system where all of the GPUs have a zero value for the device attribute cudaDevAttrConcurrentManagedAccess and all the GPUs have peer-to-peer support with each other, the physical storage for managed memory is created on the GPU which is active at the time cudaMallocManaged is called. All other GPUs will reference the data at reduced bandwidth via peer mappings over the PCIe bus. The Unified Memory driver does not migrate memory among such GPUs.

In a multi-GPU system where not all GPUs have peer-to-peer support with each other and where the value of the device attribute cudaDevAttrConcurrentManagedAccess is zero for at least one of those GPUs, the location chosen for physical storage of managed memory is system-dependent.

  • On Linux, the location chosen will be device memory as long as the current set of active contexts are on devices that either have peer-to-peer support with each other or have a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess. If there is an active context on a GPU that does not have a non-zero value for that device attribute and it does not have peer-to-peer support with the other devices that have active contexts on them, then the location for physical storage will be 'zero-copy' or host memory. Note that this means that managed memory that is located in device memory is migrated to host memory if a new context is created on a GPU that doesn't have a non-zero value for the device attribute and does not support peer-to-peer with at least one of the other devices that has an active context. This in turn implies that context creation may fail if there is insufficient host memory to migrate all managed allocations.

  • On Windows, the physical storage is always created in 'zero-copy' or host memory. All GPUs will reference the data at reduced bandwidth over the PCIe bus. In these circumstances, use of the environment variable CUDA_VISIBLE_DEVICES is recommended to restrict CUDA to only use those GPUs that have peer-to-peer support. Alternatively, users can also set CUDA_MANAGED_FORCE_DEVICE_ALLOC to a non-zero value to force the driver to always use device memory for physical storage. When this environment variable is set to a non-zero value, all devices used in that process that support managed memory have to be peer-to-peer compatible with each other. The error cudaErrorInvalidDevice will be returned if a device that supports managed memory is used and it is not peer-to-peer compatible with any of the other managed memory supporting devices that were previously used in that process, even if cudaDeviceReset has been called on those devices. These environment variables are described in the CUDA programming guide under the "CUDA environment variables" section.

See also:

cudaMallocPitch, cudaFree, cudaMallocArray, cudaFreeArray, cudaMalloc3D, cudaMalloc3DArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, cudaDeviceGetAttribute, cudaStreamAttachMemAsync, cuMemAllocManaged

__host__cudaError_t cudaMallocMipmappedArray ( cudaMipmappedArray_t* mipmappedArray, const cudaChannelFormatDesc* desc, cudaExtent extent, unsigned int  numLevels, unsigned int  flags = 0 )
Allocate a mipmapped array on the device.
Parameters
mipmappedArray
- Pointer to allocated mipmapped array in device memory
desc
- Requested channel format
extent
- Requested allocation size (width field in elements)
numLevels
- Number of mipmap levels to allocate
flags
- Flags for extensions
Description

Allocates a CUDA mipmapped array according to the cudaChannelFormatDesc structure desc and returns a handle to the new CUDA mipmapped array in *mipmappedArray. numLevels specifies the number of mipmap levels to be allocated. This value is clamped to the range [1, 1 + floor(log2(max(width, height, depth)))].

The cudaChannelFormatDesc is defined as:

‎    struct cudaChannelFormatDesc {
              int x, y, z, w;
              enum cudaChannelFormatKind 
                  f;
          };
where cudaChannelFormatKind is one of cudaChannelFormatKindSigned, cudaChannelFormatKindUnsigned, or cudaChannelFormatKindFloat.

cudaMallocMipmappedArray() can allocate the following:

  • A 1D mipmapped array is allocated if the height and depth extents are both zero.

  • A 2D mipmapped array is allocated if only the depth extent is zero.

  • A 3D mipmapped array is allocated if all three extents are non-zero.

  • A 1D layered CUDA mipmapped array is allocated if only the height extent is zero and the cudaArrayLayered flag is set. Each layer is a 1D mipmapped array. The number of layers is determined by the depth extent.

  • A 2D layered CUDA mipmapped array is allocated if all three extents are non-zero and the cudaArrayLayered flag is set. Each layer is a 2D mipmapped array. The number of layers is determined by the depth extent.

  • A cubemap CUDA mipmapped array is allocated if all three extents are non-zero and the cudaArrayCubemap flag is set. Width must be equal to height, and depth must be six. The order of the six layers in memory is the same as that listed in cudaGraphicsCubeFace.

  • A cubemap layered CUDA mipmapped array is allocated if all three extents are non-zero, and both, cudaArrayCubemap and cudaArrayLayered flags are set. Width must be equal to height, and depth must be a multiple of six. A cubemap layered CUDA mipmapped array is a special type of 2D layered CUDA mipmapped array that consists of a collection of cubemap mipmapped arrays. The first six layers represent the first cubemap mipmapped array, the next six layers form the second cubemap mipmapped array, and so on.

The flags parameter enables different options to be specified that affect the allocation, as follows.

  • cudaArrayDefault: This flag's value is defined to be 0 and provides default mipmapped array allocation

  • cudaArrayLayered: Allocates a layered CUDA mipmapped array, with the depth extent indicating the number of layers

  • cudaArrayCubemap: Allocates a cubemap CUDA mipmapped array. Width must be equal to height, and depth must be six. If the cudaArrayLayered flag is also set, depth must be a multiple of six.

  • cudaArraySurfaceLoadStore: This flag indicates that individual mipmap levels of the CUDA mipmapped array will be read from or written to using a surface reference.

  • cudaArrayTextureGather: This flag indicates that texture gather operations will be performed on the CUDA array. Texture gather can only be performed on 2D CUDA mipmapped arrays, and the gather operations are performed only on the most detailed mipmap level.

The width, height and depth extents must meet certain size requirements as listed in the following table. All values are specified in elements.

CUDA array type Valid extents that must always be met {(width range in elements), (height range), (depth range)} Valid extents with cudaArraySurfaceLoadStore set {(width range in elements), (height range), (depth range)}
1D { (1,maxTexture1DMipmap), 0, 0 } { (1,maxSurface1D), 0, 0 }
2D { (1,maxTexture2DMipmap[0]), (1,maxTexture2DMipmap[1]), 0 } { (1,maxSurface2D[0]), (1,maxSurface2D[1]), 0 }
3D { (1,maxTexture3D[0]), (1,maxTexture3D[1]), (1,maxTexture3D[2]) } OR { (1,maxTexture3DAlt[0]), (1,maxTexture3DAlt[1]), (1,maxTexture3DAlt[2]) } { (1,maxSurface3D[0]), (1,maxSurface3D[1]), (1,maxSurface3D[2]) }
1D Layered { (1,maxTexture1DLayered[0]), 0, (1,maxTexture1DLayered[1]) } { (1,maxSurface1DLayered[0]), 0, (1,maxSurface1DLayered[1]) }
2D Layered { (1,maxTexture2DLayered[0]), (1,maxTexture2DLayered[1]), (1,maxTexture2DLayered[2]) } { (1,maxSurface2DLayered[0]), (1,maxSurface2DLayered[1]), (1,maxSurface2DLayered[2]) }
Cubemap { (1,maxTextureCubemap), (1,maxTextureCubemap), 6 } { (1,maxSurfaceCubemap), (1,maxSurfaceCubemap), 6 }
Cubemap Layered { (1,maxTextureCubemapLayered[0]), (1,maxTextureCubemapLayered[0]), (1,maxTextureCubemapLayered[1]) } { (1,maxSurfaceCubemapLayered[0]), (1,maxSurfaceCubemapLayered[0]), (1,maxSurfaceCubemapLayered[1]) }

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc3D, cudaMalloc, cudaMallocPitch, cudaFree, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaHostAlloc, make_cudaExtent, cuMipmappedArrayCreate

__host__cudaError_t cudaMallocPitch ( void** devPtr, size_t* pitch, size_t width, size_t height )
Allocates pitched memory on the device.
Parameters
devPtr
- Pointer to allocated pitched device memory
pitch
- Pitch for allocation
width
- Requested pitched allocation width (in bytes)
height
- Requested pitched allocation height
Description

Allocates at least width (in bytes) * height bytes of linear memory on the device and returns in *devPtr a pointer to the allocated memory. The function may pad the allocation to ensure that corresponding pointers in any given row will continue to meet the alignment requirements for coalescing as the address is updated from row to row. The pitch returned in *pitch by cudaMallocPitch() is the width in bytes of the allocation. The intended usage of pitch is as a separate parameter of the allocation, used to compute addresses within the 2D array. Given the row and column of an array element of type T, the address is computed as:

‎    T* pElement = (T*)((char*)BaseAddress + Row * pitch) + Column;

For allocations of 2D arrays, it is recommended that programmers consider performing pitch allocations using cudaMallocPitch(). Due to pitch alignment restrictions in the hardware, this is especially true if the application will be performing 2D memory copies between different regions of device memory (whether linear memory or CUDA arrays).

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMalloc, cudaFree, cudaMallocArray, cudaFreeArray, cudaMallocHost ( C API), cudaFreeHost, cudaMalloc3D, cudaMalloc3DArray, cudaHostAlloc, cuMemAllocPitch

__host__cudaError_t cudaMemAdvise ( const void* devPtr, size_t count, cudaMemoryAdvise advice, int  device )
Advise about the usage of a given memory range.
Parameters
devPtr
- Pointer to memory to set the advice for
count
- Size in bytes of the memory range
advice
- Advice to be applied for the specified memory range
device
- Device to apply the advice for
Description

Advise the Unified Memory subsystem about the usage pattern for the memory range starting at devPtr with a size of count bytes. The start address and end address of the memory range will be rounded down and rounded up respectively to be aligned to CPU page size before the advice is applied. The memory range must refer to managed memory allocated via cudaMallocManaged or declared via __managed__ variables.

The advice parameter can take the following values:

  • cudaMemAdviseSetReadMostly: This implies that the data is mostly going to be read from and only occasionally written to. Any read accesses from any processor to this region will create a read-only copy of at least the accessed pages in that processor's memory. Additionally, if cudaMemPrefetchAsync is called on this region, it will create a read-only copy of the data on the destination processor. If any processor writes to this region, all copies of the corresponding page will be invalidated except for the one where the write occurred. The device argument is ignored for this advice. Note that for a page to be read-duplicated, the accessing processor must either be the CPU or a GPU that has a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess. Also, if a context is created on a device that does not have the device attribute cudaDevAttrConcurrentManagedAccess set, then read-duplication will not occur until all such contexts are destroyed.

  • cudaMemAdviceUnsetReadMostly: Undoes the effect of cudaMemAdviceReadMostly and also prevents the Unified Memory driver from attempting heuristic read-duplication on the memory range. Any read-duplicated copies of the data will be collapsed into a single copy. The location for the collapsed copy will be the preferred location if the page has a preferred location and one of the read-duplicated copies was resident at that location. Otherwise, the location chosen is arbitrary.

  • cudaMemAdviseSetPreferredLocation: This advice sets the preferred location for the data to be the memory belonging to device. Passing in cudaCpuDeviceId for device sets the preferred location as host memory. If device is a GPU, then it must have a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess. Setting the preferred location does not cause data to migrate to that location immediately. Instead, it guides the migration policy when a fault occurs on that memory region. If the data is already in its preferred location and the faulting processor can establish a mapping without requiring the data to be migrated, then data migration will be avoided. On the other hand, if the data is not in its preferred location or if a direct mapping cannot be established, then it will be migrated to the processor accessing it. It is important to note that setting the preferred location does not prevent data prefetching done using cudaMemPrefetchAsync. Having a preferred location can override the page thrash detection and resolution logic in the Unified Memory driver. Normally, if a page is detected to be constantly thrashing between for example host and device memory, the page may eventually be pinned to host memory by the Unified Memory driver. But if the preferred location is set as device memory, then the page will continue to thrash indefinitely. If cudaMemAdviseSetReadMostly is also set on this memory region or any subset of it, then the policies associated with that advice will override the policies of this advice.

  • cudaMemAdviseUnsetPreferredLocation: Undoes the effect of cudaMemAdviseSetPreferredLocation and changes the preferred location to none.

  • cudaMemAdviseSetAccessedBy: This advice implies that the data will be accessed by device. Passing in cudaCpuDeviceId for device will set the advice for the CPU. If device is a GPU, then the device attribute cudaDevAttrConcurrentManagedAccess must be non-zero. This advice does not cause data migration and has no impact on the location of the data per se. Instead, it causes the data to always be mapped in the specified processor's page tables, as long as the location of the data permits a mapping to be established. If the data gets migrated for any reason, the mappings are updated accordingly. This advice is recommended in scenarios where data locality is not important, but avoiding faults is. Consider for example a system containing multiple GPUs with peer-to-peer access enabled, where the data located on one GPU is occasionally accessed by peer GPUs. In such scenarios, migrating data over to the other GPUs is not as important because the accesses are infrequent and the overhead of migration may be too high. But preventing faults can still help improve performance, and so having a mapping set up in advance is useful. Note that on CPU access of this data, the data may be migrated to host memory because the CPU typically cannot access device memory directly. Any GPU that had the cudaMemAdviceSetAccessedBy flag set for this data will now have its mapping updated to point to the page in host memory. If cudaMemAdviseSetReadMostly is also set on this memory region or any subset of it, then the policies associated with that advice will override the policies of this advice. Additionally, if the preferred location of this memory region or any subset of it is also device, then the policies associated with cudaMemAdviseSetPreferredLocation will override the policies of this advice.

  • cudaMemAdviseUnsetAccessedBy: Undoes the effect of cudaMemAdviseSetAccessedBy. Any mappings to the data from device may be removed at any time causing accesses to result in non-fatal page faults.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpyPeer, cudaMemcpyAsync, cudaMemcpy3DPeerAsync, cudaMemPrefetchAsync, cuMemAdvise

__host__cudaError_t cudaMemGetInfo ( size_t* free, size_t* total )
Gets free and total device memory.
Parameters
free
- Returned free memory in bytes
total
- Returned total memory in bytes
Description

Returns in *free and *total respectively, the free and total amount of memory available for allocation by the device in bytes.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cuMemGetInfo

__host__cudaError_t cudaMemPrefetchAsync ( const void* devPtr, size_t count, int  dstDevice, cudaStream_t stream = 0 )
Prefetches memory to the specified destination device.
Parameters
devPtr
- Pointer to be prefetched
count
- Size in bytes
dstDevice
- Destination device to prefetch to
stream
- Stream to enqueue prefetch operation
Description

Prefetches memory to the specified destination device. devPtr is the base device pointer of the memory to be prefetched and dstDevice is the destination device. count specifies the number of bytes to copy. stream is the stream in which the operation is enqueued. The memory range must refer to managed memory allocated via cudaMallocManaged or declared via __managed__ variables.

Passing in cudaCpuDeviceId for dstDevice will prefetch the data to host memory. If dstDevice is a GPU, then the device attribute cudaDevAttrConcurrentManagedAccess must be non-zero. Additionally, stream must be associated with a device that has a non-zero value for the device attribute cudaDevAttrConcurrentManagedAccess.

The start address and end address of the memory range will be rounded down and rounded up respectively to be aligned to CPU page size before the prefetch operation is enqueued in the stream.

If no physical memory has been allocated for this region, then this memory region will be populated and mapped on the destination device. If there's insufficient memory to prefetch the desired region, the Unified Memory driver may evict pages from other cudaMallocManaged allocations to host memory in order to make room. Device memory allocated using cudaMalloc or cudaMallocArray will not be evicted.

By default, any mappings to the previous location of the migrated pages are removed and mappings for the new location are only setup on dstDevice. The exact behavior however also depends on the settings applied to this memory range via cudaMemAdvise as described below:

If cudaMemAdviseSetReadMostly was set on any subset of this memory range, then that subset will create a read-only copy of the pages on dstDevice.

If cudaMemAdviseSetPreferredLocation was called on any subset of this memory range, then the pages will be migrated to dstDevice even if dstDevice is not the preferred location of any pages in the memory range.

If cudaMemAdviseSetAccessedBy was called on any subset of this memory range, then mappings to those pages from all the appropriate processors are updated to refer to the new location if establishing such a mapping is possible. Otherwise, those mappings are cleared.

Note that this API is not required for functionality and only serves to improve performance by allowing the application to migrate data to a suitable location before it is accessed. Memory accesses to this range are always coherent and are allowed even when the data is actively being migrated.

Note that this function is asynchronous with respect to the host and all work on other devices.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpyPeer, cudaMemcpyAsync, cudaMemcpy3DPeerAsync, cudaMemAdvise, cuMemPrefetchAsync

__host__cudaError_t cudaMemRangeGetAttribute ( void* data, size_t dataSize, cudaMemRangeAttribute attribute, const void* devPtr, size_t count )
Query an attribute of a given memory range.
Parameters
data
- A pointers to a memory location where the result of each attribute query will be written to.
dataSize
- Array containing the size of data
attribute
- The attribute to query
devPtr
- Start of the range to query
count
- Size of the range to query
Description

Query an attribute about the memory range starting at devPtr with a size of count bytes. The memory range must refer to managed memory allocated via cudaMallocManaged or declared via __managed__ variables.

The attribute parameter can take the following values:

  • cudaMemRangeAttributeReadMostly: If this attribute is specified, data will be interpreted as a 32-bit integer, and dataSize must be 4. The result returned will be 1 if all pages in the given memory range have read-duplication enabled, or 0 otherwise.

  • cudaMemRangeAttributePreferredLocation: If this attribute is specified, data will be interpreted as a 32-bit integer, and dataSize must be 4. The result returned will be a GPU device id if all pages in the memory range have that GPU as their preferred location, or it will be cudaCpuDeviceId if all pages in the memory range have the CPU as their preferred location, or it will be cudaInvalidDeviceId if either all the pages don't have the same preferred location or some of the pages don't have a preferred location at all. Note that the actual location of the pages in the memory range at the time of the query may be different from the preferred location.

  • cudaMemRangeAttributeAccessedBy: If this attribute is specified, data will be interpreted as an array of 32-bit integers, and dataSize must be a non-zero multiple of 4. The result returned will be a list of device ids that had cudaMemAdviceSetAccessedBy set for that entire memory range. If any device does not have that advice set for the entire memory range, that device will not be included. If data is larger than the number of devices that have that advice set for that memory range, cudaInvalidDeviceId will be returned in all the extra space provided. For ex., if dataSize is 12 (i.e. data has 3 elements) and only device 0 has the advice set, then the result returned will be { 0, cudaInvalidDeviceId, cudaInvalidDeviceId }. If data is smaller than the number of devices that have that advice set, then only as many devices will be returned as can fit in the array. There is no guarantee on which specific devices will be returned, however.

  • cudaMemRangeAttributeLastPrefetchLocation: If this attribute is specified, data will be interpreted as a 32-bit integer, and dataSize must be 4. The result returned will be the last location to which all pages in the memory range were prefetched explicitly via cudaMemPrefetchAsync. This will either be a GPU id or cudaCpuDeviceId depending on whether the last location for prefetch was a GPU or the CPU respectively. If any page in the memory range was never explicitly prefetched or if all pages were not prefetched to the same location, cudaInvalidDeviceId will be returned. Note that this simply returns the last location that the applicaton requested to prefetch the memory range to. It gives no indication as to whether the prefetch operation to that location has completed or even begun.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemRangeGetAttributes, cudaMemPrefetchAsync, cudaMemAdvise, cuMemRangeGetAttribute

__host__cudaError_t cudaMemRangeGetAttributes ( void** data, size_t* dataSizes, cudaMemRangeAttribute ** attributes, size_t numAttributes, const void* devPtr, size_t count )
Query attributes of a given memory range.
Parameters
data
- A two-dimensional array containing pointers to memory locations where the result of each attribute query will be written to.
dataSizes
- Array containing the sizes of each result
attributes
- An array of attributes to query (numAttributes and the number of attributes in this array should match)
numAttributes
- Number of attributes to query
devPtr
- Start of the range to query
count
- Size of the range to query
Description

Query attributes of the memory range starting at devPtr with a size of count bytes. The memory range must refer to managed memory allocated via cudaMallocManaged or declared via __managed__ variables. The attributes array will be interpreted to have numAttributes entries. The dataSizes array will also be interpreted to have numAttributes entries. The results of the query will be stored in data.

The list of supported attributes are given below. Please refer to cudaMemRangeGetAttribute for attribute descriptions and restrictions.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMemRangeGetAttribute, cudaMemAdvisecudaMemPrefetchAsync, cuMemRangeGetAttributes

__host__cudaError_t cudaMemcpy ( void* dst, const void* src, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
src
- Source memory address
count
- Size in bytes to copy
kind
- Type of transfer
Description

Copies count bytes from the memory area pointed to by src to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. Calling cudaMemcpy() with dst and src pointers that do not match the direction of the copy results in an undefined behavior.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyDtoH, cuMemcpyHtoD, cuMemcpyDtoD, cuMemcpy

__host__cudaError_t cudaMemcpy2D ( void* dst, size_t dpitch, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
dpitch
- Pitch of destination memory
src
- Source memory address
spitch
- Pitch of source memory
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
Description

Copies a matrix (height rows of width bytes each) from the memory area pointed to by src to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. dpitch and spitch are the widths in memory in bytes of the 2D arrays pointed to by dst and src, including any padding added to the end of each row. The memory areas may not overlap. width must not exceed either dpitch or spitch. Calling cudaMemcpy2D() with dst and src pointers that do not match the direction of the copy results in an undefined behavior. cudaMemcpy2D() returns an error if dpitch or spitch exceeds the maximum allowed.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMemcpy, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2D, cuMemcpy2DUnaligned

__host__cudaError_t cudaMemcpy2DArrayToArray ( cudaArray_t dst, size_t wOffsetDst, size_t hOffsetDst, cudaArray_const_t src, size_t wOffsetSrc, size_t hOffsetSrc, size_t width, size_t height, cudaMemcpyKind kind = cudaMemcpyDeviceToDevice )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffsetDst
- Destination starting X offset
hOffsetDst
- Destination starting Y offset
src
- Source memory address
wOffsetSrc
- Source starting X offset
hOffsetSrc
- Source starting Y offset
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
Description

Copies a matrix (height rows of width bytes each) from the CUDA array srcArray starting at the upper left corner (wOffsetSrc, hOffsetSrc) to the CUDA array dst starting at the upper left corner (wOffsetDst, hOffsetDst), where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. wOffsetDst + width must not exceed the width of the CUDA array dst. wOffsetSrc + width must not exceed the width of the CUDA array src.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2D, cuMemcpy2DUnaligned

__host____device__cudaError_t cudaMemcpy2DAsync ( void* dst, size_t dpitch, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
dpitch
- Pitch of destination memory
src
- Source memory address
spitch
- Pitch of source memory
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
stream
- Stream identifier
Description

Copies a matrix (height rows of width bytes each) from the memory area pointed to by src to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. dpitch and spitch are the widths in memory in bytes of the 2D arrays pointed to by dst and src, including any padding added to the end of each row. The memory areas may not overlap. width must not exceed either dpitch or spitch.

Calling cudaMemcpy2DAsync() with dst and src pointers that do not match the direction of the copy results in an undefined behavior. cudaMemcpy2DAsync() returns an error if dpitch or spitch is greater than the maximum allowed.

cudaMemcpy2DAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2DAsync

__host__cudaError_t cudaMemcpy2DFromArray ( void* dst, size_t dpitch, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
dpitch
- Pitch of destination memory
src
- Source memory address
wOffset
- Source starting X offset
hOffset
- Source starting Y offset
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
Description

Copies a matrix (height rows of width bytes each) from the CUDA array srcArray starting at the upper left corner (wOffset, hOffset) to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. dpitch is the width in memory in bytes of the 2D array pointed to by dst, including any padding added to the end of each row. wOffset + width must not exceed the width of the CUDA array src. width must not exceed dpitch. cudaMemcpy2DFromArray() returns an error if dpitch exceeds the maximum allowed.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2D, cuMemcpy2DUnaligned

__host__cudaError_t cudaMemcpy2DFromArrayAsync ( void* dst, size_t dpitch, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
dpitch
- Pitch of destination memory
src
- Source memory address
wOffset
- Source starting X offset
hOffset
- Source starting Y offset
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
stream
- Stream identifier
Description

Copies a matrix (height rows of width bytes each) from the CUDA array srcArray starting at the upper left corner (wOffset, hOffset) to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. dpitch is the width in memory in bytes of the 2D array pointed to by dst, including any padding added to the end of each row. wOffset + width must not exceed the width of the CUDA array src. width must not exceed dpitch. cudaMemcpy2DFromArrayAsync() returns an error if dpitch exceeds the maximum allowed.

cudaMemcpy2DFromArrayAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2DAsync

__host__cudaError_t cudaMemcpy2DToArray ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffset
- Destination starting X offset
hOffset
- Destination starting Y offset
src
- Source memory address
spitch
- Pitch of source memory
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
Description

Copies a matrix (height rows of width bytes each) from the memory area pointed to by src to the CUDA array dst starting at the upper left corner (wOffset, hOffset) where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. spitch is the width in memory in bytes of the 2D array pointed to by src, including any padding added to the end of each row. wOffset + width must not exceed the width of the CUDA array dst. width must not exceed spitch. cudaMemcpy2DToArray() returns an error if spitch exceeds the maximum allowed.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2D, cuMemcpy2DUnaligned

__host__cudaError_t cudaMemcpy2DToArrayAsync ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t spitch, size_t width, size_t height, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffset
- Destination starting X offset
hOffset
- Destination starting Y offset
src
- Source memory address
spitch
- Pitch of source memory
width
- Width of matrix transfer (columns in bytes)
height
- Height of matrix transfer (rows)
kind
- Type of transfer
stream
- Stream identifier
Description

Copies a matrix (height rows of width bytes each) from the memory area pointed to by src to the CUDA array dst starting at the upper left corner (wOffset, hOffset) where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. spitch is the width in memory in bytes of the 2D array pointed to by src, including any padding added to the end of each row. wOffset + width must not exceed the width of the CUDA array dst. width must not exceed spitch. cudaMemcpy2DToArrayAsync() returns an error if spitch exceeds the maximum allowed.

cudaMemcpy2DToArrayAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy2DAsync

__host__cudaError_t cudaMemcpy3D ( const cudaMemcpy3DParms* p )
Copies data between 3D objects.
Parameters
p
- 3D memory copy parameters
Description

‎struct cudaExtent {
        size_t width;
        size_t height;
        size_t depth;
      };
      struct cudaExtent 
                  make_cudaExtent(size_t w, size_t h, size_t d);
      
      struct cudaPos {
        size_t x;
        size_t y;
        size_t z;
      };
      struct cudaPos 
                  make_cudaPos(size_t x, size_t y, size_t z);
      
      struct cudaMemcpy3DParms {
        cudaArray_t           
                  srcArray;
        struct cudaPos        
                  srcPos;
        struct cudaPitchedPtr 
                  srcPtr;
        cudaArray_t           
                  dstArray;
        struct cudaPos        
                  dstPos;
        struct cudaPitchedPtr 
                  dstPtr;
        struct cudaExtent     
                  extent;
        enum cudaMemcpyKind   
                  kind;
      };

cudaMemcpy3D() copies data betwen two 3D objects. The source and destination objects may be in either host memory, device memory, or a CUDA array. The source, destination, extent, and kind of copy performed is specified by the cudaMemcpy3DParms struct which should be initialized to zero before use:

cudaMemcpy3DParms myParms = {0};

The struct passed to cudaMemcpy3D() must specify one of srcArray or srcPtr and one of dstArray or dstPtr. Passing more than one non-zero source or destination will cause cudaMemcpy3D() to return an error.

The srcPos and dstPos fields are optional offsets into the source and destination objects and are defined in units of each object's elements. The element for a host or device pointer is assumed to be unsigned char.

The extent field defines the dimensions of the transferred area in elements. If a CUDA array is participating in the copy, the extent is defined in terms of that array's elements. If no CUDA array is participating in the copy then the extents are defined in elements of unsigned char.

The kind field defines the direction of the copy. It must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. For cudaMemcpyHostToHost or cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost passed as kind and cudaArray type passed as source or destination, if the kind implies cudaArray type to be present on the host, cudaMemcpy3D() will disregard that implication and silently correct the kind based on the fact that cudaArray type can only be present on the device.

If the source and destination are both arrays, cudaMemcpy3D() will return an error if they do not have the same element size.

The source and destination object may not overlap. If overlapping source and destination objects are specified, undefined behavior will result.

The source object must lie entirely within the region defined by srcPos and extent. The destination object must lie entirely within the region defined by dstPos and extent.

cudaMemcpy3D() returns an error if the pitch of srcPtr or dstPtr exceeds the maximum allowed. The pitch of a cudaPitchedPtr allocated with cudaMalloc3D() will always be valid.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMalloc3D, cudaMalloc3DArray, cudaMemset3D, cudaMemcpy3DAsync, cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, make_cudaExtent, make_cudaPos, cuMemcpy3D

__host____device__cudaError_t cudaMemcpy3DAsync ( const cudaMemcpy3DParms* p, cudaStream_t stream = 0 )
Copies data between 3D objects.
Parameters
p
- 3D memory copy parameters
stream
- Stream identifier
Description

‎struct cudaExtent {
        size_t width;
        size_t height;
        size_t depth;
      };
      struct cudaExtent 
                  make_cudaExtent(size_t w, size_t h, size_t d);
      
      struct cudaPos {
        size_t x;
        size_t y;
        size_t z;
      };
      struct cudaPos 
                  make_cudaPos(size_t x, size_t y, size_t z);
      
      struct cudaMemcpy3DParms {
        cudaArray_t           
                  srcArray;
        struct cudaPos        
                  srcPos;
        struct cudaPitchedPtr 
                  srcPtr;
        cudaArray_t           
                  dstArray;
        struct cudaPos        
                  dstPos;
        struct cudaPitchedPtr 
                  dstPtr;
        struct cudaExtent     
                  extent;
        enum cudaMemcpyKind   
                  kind;
      };

cudaMemcpy3DAsync() copies data betwen two 3D objects. The source and destination objects may be in either host memory, device memory, or a CUDA array. The source, destination, extent, and kind of copy performed is specified by the cudaMemcpy3DParms struct which should be initialized to zero before use:

cudaMemcpy3DParms myParms = {0};

The struct passed to cudaMemcpy3DAsync() must specify one of srcArray or srcPtr and one of dstArray or dstPtr. Passing more than one non-zero source or destination will cause cudaMemcpy3DAsync() to return an error.

The srcPos and dstPos fields are optional offsets into the source and destination objects and are defined in units of each object's elements. The element for a host or device pointer is assumed to be unsigned char. For CUDA arrays, positions must be in the range [0, 2048) for any dimension.

The extent field defines the dimensions of the transferred area in elements. If a CUDA array is participating in the copy, the extent is defined in terms of that array's elements. If no CUDA array is participating in the copy then the extents are defined in elements of unsigned char.

The kind field defines the direction of the copy. It must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing. For cudaMemcpyHostToHost or cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost passed as kind and cudaArray type passed as source or destination, if the kind implies cudaArray type to be present on the host, cudaMemcpy3DAsync() will disregard that implication and silently correct the kind based on the fact that cudaArray type can only be present on the device.

If the source and destination are both arrays, cudaMemcpy3DAsync() will return an error if they do not have the same element size.

The source and destination object may not overlap. If overlapping source and destination objects are specified, undefined behavior will result.

The source object must lie entirely within the region defined by srcPos and extent. The destination object must lie entirely within the region defined by dstPos and extent.

cudaMemcpy3DAsync() returns an error if the pitch of srcPtr or dstPtr exceeds the maximum allowed. The pitch of a cudaPitchedPtr allocated with cudaMalloc3D() will always be valid.

cudaMemcpy3DAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMalloc3D, cudaMalloc3DArray, cudaMemset3D, cudaMemcpy3D, cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, make_cudaExtent, make_cudaPos, cuMemcpy3DAsync

__host__cudaError_t cudaMemcpy3DPeer ( const cudaMemcpy3DPeerParms* p )
Copies memory between devices.
Parameters
p
- Parameters for the memory copy
Description

Perform a 3D memory copy according to the parameters specified in p. See the definition of the cudaMemcpy3DPeerParms structure for documentation of its parameters.

Note that this function is synchronous with respect to the host only if the source or destination of the transfer is host memory. Note also that this copy is serialized with respect to all pending and future asynchronous work in to the current device, the copy's source device, and the copy's destination device (use cudaMemcpy3DPeerAsync to avoid this synchronization).

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpyPeer, cudaMemcpyAsync, cudaMemcpyPeerAsync, cudaMemcpy3DPeerAsync, cuMemcpy3DPeer

__host__cudaError_t cudaMemcpy3DPeerAsync ( const cudaMemcpy3DPeerParms* p, cudaStream_t stream = 0 )
Copies memory between devices asynchronously.
Parameters
p
- Parameters for the memory copy
stream
- Stream identifier
Description

Perform a 3D memory copy according to the parameters specified in p. See the definition of the cudaMemcpy3DPeerParms structure for documentation of its parameters.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpyPeer, cudaMemcpyAsync, cudaMemcpyPeerAsync, cudaMemcpy3DPeerAsync, cuMemcpy3DPeerAsync

__host__cudaError_t cudaMemcpyArrayToArray ( cudaArray_t dst, size_t wOffsetDst, size_t hOffsetDst, cudaArray_const_t src, size_t wOffsetSrc, size_t hOffsetSrc, size_t count, cudaMemcpyKind kind = cudaMemcpyDeviceToDevice )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffsetDst
- Destination starting X offset
hOffsetDst
- Destination starting Y offset
src
- Source memory address
wOffsetSrc
- Source starting X offset
hOffsetSrc
- Source starting Y offset
count
- Size in bytes to copy
kind
- Type of transfer
Description

Copies count bytes from the CUDA array src starting at the upper left corner (wOffsetSrc, hOffsetSrc) to the CUDA array dst starting at the upper left corner (wOffsetDst, hOffsetDst) where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

Note:

Note that this function may also return error codes from previous, asynchronous launches.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyAtoA

__host____device__cudaError_t cudaMemcpyAsync ( void* dst, const void* src, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
src
- Source memory address
count
- Size in bytes to copy
kind
- Type of transfer
stream
- Stream identifier
Description

Copies count bytes from the memory area pointed to by src to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

The memory areas may not overlap. Calling cudaMemcpyAsync() with dst and src pointers that do not match the direction of the copy results in an undefined behavior.

cudaMemcpyAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and the stream is non-zero, the copy may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsynccuMemcpyAsync, cuMemcpyDtoHAsync, cuMemcpyHtoDAsync, cuMemcpyDtoDAsync

__host__cudaError_t cudaMemcpyFromArray ( void* dst, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
src
- Source memory address
wOffset
- Source starting X offset
hOffset
- Source starting Y offset
count
- Size in bytes to copy
kind
- Type of transfer
Description

Copies count bytes from the CUDA array src starting at the upper left corner (wOffset, hOffset) to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyAtoH, cuMemcpyAtoD

__host__cudaError_t cudaMemcpyFromArrayAsync ( void* dst, cudaArray_const_t src, size_t wOffset, size_t hOffset, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
src
- Source memory address
wOffset
- Source starting X offset
hOffset
- Source starting Y offset
count
- Size in bytes to copy
kind
- Type of transfer
stream
- Stream identifier
Description

Copies count bytes from the CUDA array src starting at the upper left corner (wOffset, hOffset) to the memory area pointed to by dst, where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

cudaMemcpyFromArrayAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyAtoHAsync, cuMemcpy2DAsync

__host__cudaError_t cudaMemcpyFromSymbol ( void* dst, const void* symbol, size_t count, size_t offset = 0, cudaMemcpyKind kind = cudaMemcpyDeviceToHost )
Copies data from the given symbol on the device.
Parameters
dst
- Destination memory address
symbol
- Device symbol address
count
- Size in bytes to copy
offset
- Offset from start of symbol in bytes
kind
- Type of transfer
Description

Copies count bytes from the memory area pointed to by offset bytes from the start of symbol symbol to the memory area pointed to by dst. The memory areas may not overlap. symbol is a variable that resides in global or constant memory space. kind can be either cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy, cuMemcpyDtoH, cuMemcpyDtoD

__host__cudaError_t cudaMemcpyFromSymbolAsync ( void* dst, const void* symbol, size_t count, size_t offset, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data from the given symbol on the device.
Parameters
dst
- Destination memory address
symbol
- Device symbol address
count
- Size in bytes to copy
offset
- Offset from start of symbol in bytes
kind
- Type of transfer
stream
- Stream identifier
Description

Copies count bytes from the memory area pointed to by offset bytes from the start of symbol symbol to the memory area pointed to by dst. The memory areas may not overlap. symbol is a variable that resides in global or constant memory space. kind can be either cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

cudaMemcpyFromSymbolAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cuMemcpyAsync, cuMemcpyDtoHAsync, cuMemcpyDtoDAsync

__host__cudaError_t cudaMemcpyPeer ( void* dst, int  dstDevice, const void* src, int  srcDevice, size_t count )
Copies memory between two devices.
Parameters
dst
- Destination device pointer
dstDevice
- Destination device
src
- Source device pointer
srcDevice
- Source device
count
- Size of memory copy in bytes
Description

Copies memory from one device to memory on another device. dst is the base device pointer of the destination memory and dstDevice is the destination device. src is the base device pointer of the source memory and srcDevice is the source device. count specifies the number of bytes to copy.

Note that this function is asynchronous with respect to the host, but serialized with respect all pending and future asynchronous work in to the current device, srcDevice, and dstDevice (use cudaMemcpyPeerAsync to avoid this synchronization).

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpyAsync, cudaMemcpyPeerAsync, cudaMemcpy3DPeerAsync, cuMemcpyPeer

__host__cudaError_t cudaMemcpyPeerAsync ( void* dst, int  dstDevice, const void* src, int  srcDevice, size_t count, cudaStream_t stream = 0 )
Copies memory between two devices asynchronously.
Parameters
dst
- Destination device pointer
dstDevice
- Destination device
src
- Source device pointer
srcDevice
- Source device
count
- Size of memory copy in bytes
stream
- Stream identifier
Description

Copies memory from one device to memory on another device. dst is the base device pointer of the destination memory and dstDevice is the destination device. src is the base device pointer of the source memory and srcDevice is the source device. count specifies the number of bytes to copy.

Note that this function is asynchronous with respect to the host and all work on other devices.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpyPeer, cudaMemcpyAsync, cudaMemcpy3DPeerAsync, cuMemcpyPeerAsync

__host__cudaError_t cudaMemcpyToArray ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t count, cudaMemcpyKind kind )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffset
- Destination starting X offset
hOffset
- Destination starting Y offset
src
- Source memory address
count
- Size in bytes to copy
kind
- Type of transfer
Description

Copies count bytes from the memory area pointed to by src to the CUDA array dst starting at the upper left corner (wOffset, hOffset), where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyHtoA, cuMemcpyDtoA

__host__cudaError_t cudaMemcpyToArrayAsync ( cudaArray_t dst, size_t wOffset, size_t hOffset, const void* src, size_t count, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data between host and device.
Parameters
dst
- Destination memory address
wOffset
- Destination starting X offset
hOffset
- Destination starting Y offset
src
- Source memory address
count
- Size in bytes to copy
kind
- Type of transfer
stream
- Stream identifier
Description

Copies count bytes from the memory area pointed to by src to the CUDA array dst starting at the upper left corner (wOffset, hOffset), where kind specifies the direction of the copy, and must be one of cudaMemcpyHostToHost, cudaMemcpyHostToDevice, cudaMemcpyDeviceToHost, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

cudaMemcpyToArrayAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice or cudaMemcpyDeviceToHost and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpyHtoAAsync, cuMemcpy2DAsync

__host__cudaError_t cudaMemcpyToSymbol ( const void* symbol, const void* src, size_t count, size_t offset = 0, cudaMemcpyKind kind = cudaMemcpyHostToDevice )
Copies data to the given symbol on the device.
Parameters
symbol
- Device symbol address
src
- Source memory address
count
- Size in bytes to copy
offset
- Offset from start of symbol in bytes
kind
- Type of transfer
Description

Copies count bytes from the memory area pointed to by src to the memory area pointed to by offset bytes from the start of symbol symbol. The memory areas may not overlap. symbol is a variable that resides in global or constant memory space. kind can be either cudaMemcpyHostToDevice, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits synchronous behavior for most use cases.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyToSymbolAsync, cudaMemcpyFromSymbolAsync, cuMemcpy, cuMemcpyHtoD, cuMemcpyDtoD

__host__cudaError_t cudaMemcpyToSymbolAsync ( const void* symbol, const void* src, size_t count, size_t offset, cudaMemcpyKind kind, cudaStream_t stream = 0 )
Copies data to the given symbol on the device.
Parameters
symbol
- Device symbol address
src
- Source memory address
count
- Size in bytes to copy
offset
- Offset from start of symbol in bytes
kind
- Type of transfer
stream
- Stream identifier
Description

Copies count bytes from the memory area pointed to by src to the memory area pointed to by offset bytes from the start of symbol symbol. The memory areas may not overlap. symbol is a variable that resides in global or constant memory space. kind can be either cudaMemcpyHostToDevice, cudaMemcpyDeviceToDevice, or cudaMemcpyDefault. Passing cudaMemcpyDefault is recommended, in which case the type of transfer is inferred from the pointer values. However, cudaMemcpyDefault is only allowed on systems that support unified virtual addressing.

cudaMemcpyToSymbolAsync() is asynchronous with respect to the host, so the call may return before the copy is complete. The copy can optionally be associated to a stream by passing a non-zero stream argument. If kind is cudaMemcpyHostToDevice and stream is non-zero, the copy may overlap with operations in other streams.

Note:
  • Note that this function may also return error codes from previous, asynchronous launches.

  • This function exhibits asynchronous behavior for most use cases.

  • This function uses standard default stream semantics.

  • Use of a string naming a variable as the symbol paramater was deprecated in CUDA 4.1 and removed in CUDA 5.0.

See also:

cudaMemcpy, cudaMemcpy2D, cudaMemcpyToArray, cudaMemcpy2DToArray, cudaMemcpyFromArray, cudaMemcpy2DFromArray, cudaMemcpyArrayToArray, cudaMemcpy2DArrayToArray, cudaMemcpyToSymbol, cudaMemcpyFromSymbol, cudaMemcpyAsync, cudaMemcpy2DAsync, cudaMemcpyToArrayAsync, cudaMemcpy2DToArrayAsync, cudaMemcpyFromArrayAsync, cudaMemcpy2DFromArrayAsync, cudaMemcpyFromSymbolAsync, cuMemcpyAsync, cuMemcpyHtoDAsync, cuMemcpyDtoDAsync

__host__cudaError_t cudaMemset ( void* devPtr, int  value, size_t count )
Initializes or sets device memory to a value.
Parameters
devPtr
- Pointer to device memory
value
- Value to set for each byte of specified memory
count
- Size in bytes to set
Description

Fills the first count bytes of the memory area pointed to by devPtr with the constant byte value value.

Note that this function is asynchronous with respect to the host unless devPtr refers to pinned host memory.

Note:

See also:

cuMemsetD8, cuMemsetD16, cuMemsetD32

__host__cudaError_t cudaMemset2D ( void* devPtr, size_t pitch, int  value, size_t width, size_t height )
Initializes or sets device memory to a value.
Parameters
devPtr
- Pointer to 2D device memory
pitch
- Pitch in bytes of 2D device memory
value
- Value to set for each byte of specified memory
width
- Width of matrix set (columns in bytes)
height
- Height of matrix set (rows)
Description

Sets to the specified value value a matrix (height rows of width bytes each) pointed to by dstPtr. pitch is the width in bytes of the 2D array pointed to by dstPtr, including any padding added to the end of each row. This function performs fastest when the pitch is one that has been passed back by cudaMallocPitch().

Note that this function is asynchronous with respect to the host unless devPtr refers to pinned host memory.

Note:

See also:

cudaMemset, cudaMemset3D, cudaMemsetAsync, cudaMemset2DAsync, cudaMemset3DAsync, cuMemsetD2D8, cuMemsetD2D16, cuMemsetD2D32

__host____device__cudaError_t cudaMemset2DAsync ( void* devPtr, size_t pitch, int  value, size_t width, size_t height, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
Parameters
devPtr
- Pointer to 2D device memory
pitch
- Pitch in bytes of 2D device memory
value
- Value to set for each byte of specified memory
width
- Width of matrix set (columns in bytes)
height
- Height of matrix set (rows)
stream
- Stream identifier
Description

Sets to the specified value value a matrix (height rows of width bytes each) pointed to by dstPtr. pitch is the width in bytes of the 2D array pointed to by dstPtr, including any padding added to the end of each row. This function performs fastest when the pitch is one that has been passed back by cudaMallocPitch().

cudaMemset2DAsync() is asynchronous with respect to the host, so the call may return before the memset is complete. The operation can optionally be associated to a stream by passing a non-zero stream argument. If stream is non-zero, the operation may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:

See also:

cudaMemset, cudaMemset2D, cudaMemset3D, cudaMemsetAsync, cudaMemset3DAsync, cuMemsetD2D8Async, cuMemsetD2D16Async, cuMemsetD2D32Async

__host__cudaError_t cudaMemset3D ( cudaPitchedPtr pitchedDevPtr, int  value, cudaExtent extent )
Initializes or sets device memory to a value.
Parameters
pitchedDevPtr
- Pointer to pitched device memory
value
- Value to set for each byte of specified memory
extent
- Size parameters for where to set device memory (width field in bytes)
Description

Initializes each element of a 3D array to the specified value value. The object to initialize is defined by pitchedDevPtr. The pitch field of pitchedDevPtr is the width in memory in bytes of the 3D array pointed to by pitchedDevPtr, including any padding added to the end of each row. The xsize field specifies the logical width of each row in bytes, while the ysize field specifies the height of each 2D slice in rows.

The extents of the initialized region are specified as a width in bytes, a height in rows, and a depth in slices.

Extents with width greater than or equal to the xsize of pitchedDevPtr may perform significantly faster than extents narrower than the xsize. Secondarily, extents with height equal to the ysize of pitchedDevPtr will perform faster than when the height is shorter than the ysize.

This function performs fastest when the pitchedDevPtr has been allocated by cudaMalloc3D().

Note that this function is asynchronous with respect to the host unless pitchedDevPtr refers to pinned host memory.

Note:

See also:

cudaMemset, cudaMemset2D, cudaMemsetAsync, cudaMemset2DAsync, cudaMemset3DAsync, cudaMalloc3D, make_cudaPitchedPtr, make_cudaExtent

__host____device__cudaError_t cudaMemset3DAsync ( cudaPitchedPtr pitchedDevPtr, int  value, cudaExtent extent, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
Parameters
pitchedDevPtr
- Pointer to pitched device memory
value
- Value to set for each byte of specified memory
extent
- Size parameters for where to set device memory (width field in bytes)
stream
- Stream identifier
Description

Initializes each element of a 3D array to the specified value value. The object to initialize is defined by pitchedDevPtr. The pitch field of pitchedDevPtr is the width in memory in bytes of the 3D array pointed to by pitchedDevPtr, including any padding added to the end of each row. The xsize field specifies the logical width of each row in bytes, while the ysize field specifies the height of each 2D slice in rows.

The extents of the initialized region are specified as a width in bytes, a height in rows, and a depth in slices.

Extents with width greater than or equal to the xsize of pitchedDevPtr may perform significantly faster than extents narrower than the xsize. Secondarily, extents with height equal to the ysize of pitchedDevPtr will perform faster than when the height is shorter than the ysize.

This function performs fastest when the pitchedDevPtr has been allocated by cudaMalloc3D().

cudaMemset3DAsync() is asynchronous with respect to the host, so the call may return before the memset is complete. The operation can optionally be associated to a stream by passing a non-zero stream argument. If stream is non-zero, the operation may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:

See also:

cudaMemset, cudaMemset2D, cudaMemset3D, cudaMemsetAsync, cudaMemset2DAsync, cudaMalloc3D, make_cudaPitchedPtr, make_cudaExtent

__host____device__cudaError_t cudaMemsetAsync ( void* devPtr, int  value, size_t count, cudaStream_t stream = 0 )
Initializes or sets device memory to a value.
Parameters
devPtr
- Pointer to device memory
value
- Value to set for each byte of specified memory
count
- Size in bytes to set
stream
- Stream identifier
Description

Fills the first count bytes of the memory area pointed to by devPtr with the constant byte value value.

cudaMemsetAsync() is asynchronous with respect to the host, so the call may return before the memset is complete. The operation can optionally be associated to a stream by passing a non-zero stream argument. If stream is non-zero, the operation may overlap with operations in other streams.

The device version of this function only handles device to device copies and cannot be given local or shared pointers.

Note:

See also:

cudaMemset, cudaMemset2D, cudaMemset3D, cudaMemset2DAsync, cudaMemset3DAsync, cuMemsetD8Async, cuMemsetD16Async, cuMemsetD32Async

__host__cudaExtent make_cudaExtent ( size_t w, size_t h, size_t d )
Returns a cudaExtent based on input parameters.
Parameters
w
- Width in elements when referring to array memory, in bytes when referring to linear memory
h
- Height in elements
d
- Depth in elements
Returns

cudaExtent specified by w, h, and d

Description

Returns a cudaExtent based on the specified input parameters w, h, and d.

See also:

make_cudaPitchedPtr, make_cudaPos

__host__cudaPitchedPtr make_cudaPitchedPtr ( void* d, size_t p, size_t xsz, size_t ysz )
Returns a cudaPitchedPtr based on input parameters.
Parameters
d
- Pointer to allocated memory
p
- Pitch of allocated memory in bytes
xsz
- Logical width of allocation in elements
ysz
- Logical height of allocation in elements
Returns

cudaPitchedPtr specified by d, p, xsz, and ysz

Description

Returns a cudaPitchedPtr based on the specified input parameters d, p, xsz, and ysz.

See also:

make_cudaExtent, make_cudaPos

__host__cudaPos make_cudaPos ( size_t x, size_t y, size_t z )
Returns a cudaPos based on input parameters.
Parameters
x
- X position
y
- Y position
z
- Z position
Returns

cudaPos specified by x, y, and z

Description

Returns a cudaPos based on the specified input parameters x, y, and z.

See also:

make_cudaExtent, make_cudaPitchedPtr