khanat-opennel-code/code/nel/include/nel/misc/contiguous_block_allocator.h
2012-12-11 11:59:54 +01:00

83 lines
3.2 KiB
C++

// NeL - MMORPG Framework <http://dev.ryzom.com/projects/nel/>
// Copyright (C) 2010 Winch Gate Property Limited
//
// This program is free software: you can redistribute it and/or modify
// it under the terms of the GNU Affero General Public License as
// published by the Free Software Foundation, either version 3 of the
// License, or (at your option) any later version.
//
// This program is distributed in the hope that it will be useful,
// but WITHOUT ANY WARRANTY; without even the implied warranty of
// MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
// GNU Affero General Public License for more details.
//
// You should have received a copy of the GNU Affero General Public License
// along with this program. If not, see <http://www.gnu.org/licenses/>.
#ifndef NL_CONTIGUOUS_BLOCK_ALLOCATOR_H
#define NL_CONTIGUOUS_BLOCK_ALLOCATOR_H
#include "types_nl.h"
namespace NLMISC
{
/** One of the simplest scheme of allocation around, but very useful in some situations.
* This allocator is just provided with a single block of memory at start (possibly of size 0). Each alloc get a new block in that big block, by
* simply advancing a pointer. When not enough space is available, the default stl allocator is used.
* When a block is deallocated, nothing happens, unless the block was allocated using the default stl allocator is used, in which case deallocate() is called.
*
* The typical use is when an object makes a lot of allocations at init, but in a predictable way, and if it doesn't make alloc / realloc
* later. In this case the caller can measure the amount of memory needed to create the object, and can create this allocator with the good amount of
* memory. Subsequent allocations will then be very fast even for very differently sized blocks, with no fragmentation inside the allocated block
* and no memory overhead per allocated bloc.
*
* Obviously, if the quantity of memory to be allocated can't be predicted (or if no max bytes can be forseen), then other allocators may
* be best suited.
*
* \author Nicolas Vizerie
* \author Nevrax France
* \date 2004
*
*/
class CContiguousBlockAllocator
{
public:
// ctor
CContiguousBlockAllocator();
// dtor
~CContiguousBlockAllocator();
// Init the allocator with the given size. Previous allocations become invalid, so when calling init again, the user must have
// freed all memory he allocated
void init(uint numBytes = 0);
// synonymous to init(0)
void release() { init(0); }
// allocated a block of n bytes
void *alloc(uint numBytes);
// deallocate a block
void free(void *block, uint numBytes);
// compute the total number of bytes allocated since init
// NB : freed block are not subtracted from that total !!
uint getNumAllocatedBytes() const { return _NumAllocatedBytes; }
#ifdef NL_DEBUG
// get number of calls to alloc since last init
uint getNumAlloc() const { return _NumAlloc; }
// get number of calls to free since last init
uint getNumFree() const { return _NumFree; }
#endif
private:
uint8 *_BlockStart;
uint8 *_BlockEnd;
uint8 *_NextAvailablePos;
uint _NumAllocatedBytes;
std::allocator<uint8> _DefaultAlloc;
#ifdef NL_DEBUG
uint _NumAlloc;
uint _NumFree;
#endif
};
}
#endif