MallocResource Class Reference

Synchronous device memory resource backed by cudaMalloc/cudaFree. Models only the Resource concept: it never touches stream-ordered allocation, so it works on devices without memory-pool support (cudaDevAttrMemoryPoolsSupported == 0), where DeviceResource's cudaMallocAsync path fails by design. Pair with AsyncFromSync to drive the stream-ordered builders on such a device. More...

#include <nanovdb/cuda/DeviceResource.h>

Public Member Functions
void *	allocate (size_t bytes, size_t)
	Allocates bytes with cudaMalloc; valid on every stream when this returns. A zero request returns nullptr.
void	deallocate (void *p, size_t, size_t)
	Frees p with cudaFree; the caller guarantees that device work touching the memory has completed.

Static Public Attributes
static constexpr size_t	DEFAULT_ALIGNMENT = 256

Friends
bool	operator== (const MallocResource &, const MallocResource &)
	Stateless: memory allocated through one instance may be deallocated through any other, so every two instances compare equal (the equality cuda::mr's resource concepts require).
bool	operator!= (const MallocResource &, const MallocResource &)

Detailed Description

Synchronous device memory resource backed by cudaMalloc/cudaFree. Models only the Resource concept: it never touches stream-ordered allocation, so it works on devices without memory-pool support (cudaDevAttrMemoryPoolsSupported == 0), where DeviceResource's cudaMallocAsync path fails by design. Pair with AsyncFromSync to drive the stream-ordered builders on such a device.

Member Function Documentation

allocate()

void * allocate ( size_t bytes, size_t  )

inline

Allocates bytes with cudaMalloc; valid on every stream when this returns. A zero request returns nullptr.

deallocate()

void deallocate ( void * p, size_t , size_t  )

inline

Frees p with cudaFree; the caller guarantees that device work touching the memory has completed.

Friends And Related Symbol Documentation

operator!=

bool operator!= ( const MallocResource & , const MallocResource &  )

friend

operator==

bool operator== ( const MallocResource & , const MallocResource &  )

friend

Stateless: memory allocated through one instance may be deallocated through any other, so every two instances compare equal (the equality cuda::mr's resource concepts require).

Member Data Documentation

DEFAULT_ALIGNMENT

size_t DEFAULT_ALIGNMENT = 256

staticconstexpr