diff --git a/include/bx/allocator.h b/include/bx/allocator.h
new file mode 100644
index 0000000..9dd9d8b
--- /dev/null
+++ b/include/bx/allocator.h
@@ -0,0 +1,224 @@
+/*
+ * Copyright 2010-2013 Branimir Karadzic. All rights reserved.
+ * License: http://www.opensource.org/licenses/BSD-2-Clause
+ */
+ 
+#ifndef __BX_ALLOCATOR_H__
+#define __BX_ALLOCATOR_H__
+
+#include "bx.h"
+
+#if BX_CONFIG_ALLOCATOR_CRT
+#	include <malloc.h>
+#endif // BX_CONFIG_ALLOCATOR_CRT
+
+#if BX_CONFIG_ALLOCATOR_DEBUG
+#	define BX_ALLOC(_allocator, _size)                 bx::alloc(_allocator, _size, __FILE__, __LINE__)
+#	define BX_REALLOC(_allocator, _ptr, _size)         bx::realloc(_allocator, _ptr, _size, __FILE__, __LINE__)
+#	define BX_FREE(_allocator, _ptr)                   bx::free(_allocator, _ptr, __FILE__, __LINE__)
+#	define BX_ALIGNED_ALLOC(_allocator, _size)         bx::alignedAlloc(_allocator, _size, __FILE__, __LINE__)
+#	define BX_ALIGNED_REALLOC(_allocator, _ptr, _size) bx::alignedRealloc(_allocator, _ptr, _size, __FILE__, __LINE__)
+#	define BX_ALIGNED_FREE(_allocator, _ptr)           bx::alignedFree(_allocator, _ptr, __FILE__, __LINE__)
+#else
+#	define BX_ALLOC(_allocator, _size)                 bx::alloc(_allocator, _size)
+#	define BX_REALLOC(_allocator, _ptr, _size)         bx::realloc(_allocator, _ptr, _size)
+#	define BX_FREE(_allocator, _ptr)                   bx::free(_allocator, _ptr)
+#	define BX_ALIGNED_ALLOC(_allocator, _size)         bx::alignedAlloc(_allocator, _size)
+#	define BX_ALIGNED_REALLOC(_allocator, _ptr, _size) bx::alignedRealloc(_allocator, _ptr, _size)
+#	define BX_ALIGNED_FREE(_allocator, _ptr)           bx::alignedFree(_allocator, _ptr)
+#endif // BX_CONFIG_DEBUG_ALLOC
+
+namespace bx
+{
+	struct BX_NO_VTABLE AllocatorI
+	{
+		virtual ~AllocatorI() = 0;
+		virtual void* alloc(size_t _size, const char* _file, uint32_t _line) = 0;
+		virtual void free(void* _ptr, const char* _file, uint32_t _line) = 0;
+	};
+
+	inline AllocatorI::~AllocatorI()
+	{
+	}
+
+	struct BX_NO_VTABLE ReallocatorI : public AllocatorI
+	{
+		virtual void* realloc(void* _ptr, size_t _size, const char* _file, uint32_t _line) = 0;
+	};
+
+	struct BX_NO_INLINE AlignedAllocatorI
+	{
+		virtual void* alignedAlloc(size_t _size, size_t _align, const char* _file, uint32_t _line) = 0;
+		virtual void alignedFree(void* _ptr, const char* _file, uint32_t _line) = 0;
+	};
+
+	struct BX_NO_VTABLE AlignedReallocatorI : public AlignedAllocatorI
+	{
+		virtual void* alignedRealloc(void* _ptr, size_t _size, size_t _align, const char* _file, uint32_t _line) = 0;
+	};
+
+	inline void* alloc(AllocatorI* _allocator, size_t _size, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->alloc(_size, _file, _line);
+	}
+
+	inline void free(AllocatorI* _allocator, void* _ptr, const char* _file = NULL, uint32_t _line = 0)
+	{
+		_allocator->free(_ptr, _file, _line);
+	}
+
+	inline void* alloc(AlignedAllocatorI* _allocator, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->alignedAlloc(_size, _align, _file, _line);
+	}
+
+	inline void free(AlignedAllocatorI* _allocator, void* _ptr, const char* _file = NULL, uint32_t _line = 0)
+	{
+		_allocator->alignedFree(_ptr, _file, _line);
+	}
+
+	inline void* realloc(ReallocatorI* _allocator, void* _ptr, size_t _size, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->realloc(_ptr, _size, _file, _line);
+	}
+
+	inline void* realloc(AlignedReallocatorI* _allocator, void* _ptr, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->alignedRealloc(_ptr, _size, _align, _file, _line);
+	}
+
+	inline void* alignPtr(void* _ptr, size_t _align)
+	{
+		union { void* ptr; size_t addr; } un;
+		un.ptr = _ptr;
+		size_t unaligned = un.addr + sizeof(uint32_t); // space for header
+		size_t mask = _align-1;
+		size_t aligned = BX_ALIGN_MASK(unaligned, mask);
+		un.addr = aligned;
+		return un.ptr;	
+	}
+
+	static inline void* alignedAlloc(AllocatorI* _allocator, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		size_t total = _size + _align;
+		uint8_t* ptr = (uint8_t*)alloc(_allocator, total, _file, _line);
+		uint8_t* aligned = (uint8_t*)alignPtr(ptr, _align);
+		uint32_t* header = (uint32_t*)aligned - 1;
+		*header = uint32_t(aligned - ptr);
+		return aligned;
+	}
+
+	static inline void alignedFree(AllocatorI* _allocator, void* _ptr, const char* _file = NULL, uint32_t _line = 0)
+	{
+		uint8_t* aligned = (uint8_t*)_ptr;
+		uint32_t* header = (uint32_t*)aligned - 1;
+		uint8_t* ptr = aligned - *header;
+		free(_allocator, ptr, _file, _line);
+	}
+
+	static inline void* alignedRealloc(ReallocatorI* _allocator, void* _ptr, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		if (NULL == _ptr)
+		{
+			return alignedAlloc(_allocator, _size, _align, _file, _line);
+		}
+
+		uint8_t* aligned = (uint8_t*)_ptr;
+		uint32_t offset = *( (uint32_t*)aligned - 1);
+		uint8_t* ptr = aligned - offset;
+		size_t total = _size + _align;
+		ptr = (uint8_t*)realloc(_allocator, ptr, total, _file, _line);
+		uint8_t* newAligned = (uint8_t*)alignPtr(ptr, _align);
+
+		if (newAligned == aligned)
+		{
+			return aligned;
+		}
+
+		aligned = ptr + offset;
+		::memmove(newAligned, aligned, _size);
+		uint32_t* header = (uint32_t*)newAligned - 1;
+		*header = uint32_t(newAligned - ptr);
+		return newAligned;
+	}
+
+	inline void* alignedAlloc(AlignedAllocatorI* _allocator, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->alignedAlloc(_size, _align, _file, _line);
+	}
+
+	inline void alignedFree(AlignedAllocatorI* _allocator, void* _ptr, const char* _file = NULL, uint32_t _line = 0)
+	{
+		_allocator->alignedFree(_ptr, _file, _line);
+	}
+
+	inline void* alignedRealloc(AlignedReallocatorI* _allocator, void* _ptr, size_t _size, size_t _align, const char* _file = NULL, uint32_t _line = 0)
+	{
+		return _allocator->alignedRealloc(_ptr, _size, _align, _file, _line);
+	}
+
+#if BX_CONFIG_ALLOCATOR_CRT
+	class CrtAllocator : public ReallocatorI, public AlignedReallocatorI
+	{
+	public:
+		CrtAllocator()
+		{
+		}
+
+		virtual ~CrtAllocator()
+		{
+		}
+
+		virtual void* alloc(size_t _size, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+			BX_UNUSED(_file, _line);
+			return ::malloc(_size);
+		}
+
+		virtual void free(void* _ptr, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+			BX_UNUSED(_file, _line);
+			::free(_ptr);
+		}
+
+		virtual void* realloc(void* _ptr, size_t _size, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+			BX_UNUSED(_file, _line);
+			return ::realloc(_ptr, _size);
+		}
+
+		virtual void* alignedAlloc(size_t _size, size_t _align, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+#	if BX_COMPILER_MSVC
+			BX_UNUSED(_file, _line);
+			return _aligned_malloc(_size, _align);
+#	else
+			return alignedAlloc(static_cast<AlignedReallocatorI*>this, _size, _align, _file, _line);
+#	endif // BX_
+		}
+
+		virtual void alignedFree(void* _ptr, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+#	if BX_COMPILER_MSVC
+			BX_UNUSED(_file, _line);
+			return _aligned_free(_ptr);
+#	else
+			return alignedFree(static_cast<AlignedReallocatorI*>this, _ptr, _align, _file, _line);
+#	endif // BX_
+		}
+
+		virtual void* alignedRealloc(void* _ptr, size_t _size, size_t _align, const char* _file, uint32_t _line) BX_OVERRIDE
+		{
+#	if BX_COMPILER_MSVC
+			BX_UNUSED(_file, _line);
+			return _aligned_realloc(_ptr, _size, _align);
+#	else
+			return alignedRealloc(static_cast<AlignedReallocatorI*>this, _ptr, _size, _align, _file, _line);
+#	endif // BX_
+		}
+	};
+#endif // BX_CONFIG_ALLOCATOR_CRT
+
+} // namespace bx
+
+#endif // __BX_ALLOCATOR_H__
diff --git a/include/bx/bx.h b/include/bx/bx.h
index ccba49f..c2c9688 100644
--- a/include/bx/bx.h
+++ b/include/bx/bx.h
@@ -12,6 +12,10 @@
 
 namespace bx
 {
-}// namespace bx
+	// http://cnicholson.net/2011/01/stupid-c-tricks-a-better-sizeof_array/
+	template<typename T, size_t N> char (&COUNTOF_REQUIRES_ARRAY_ARGUMENT(const T(&)[N]) )[N];
+#define BX_COUNTOF(x) sizeof(bx::COUNTOF_REQUIRES_ARRAY_ARGUMENT(x) )
+
+} // namespace bx
 
 #endif // __BX_H__ 
diff --git a/include/bx/countof.h b/include/bx/countof.h
deleted file mode 100644
index 85ae260..0000000
--- a/include/bx/countof.h
+++ /dev/null
@@ -1,19 +0,0 @@
-/*
- * Copyright 2010-2013 Branimir Karadzic. All rights reserved.
- * License: http://www.opensource.org/licenses/BSD-2-Clause
- */
-
-#ifndef __BX_COUNTOF_H__
-#define __BX_COUNTOF_H__
-
-#include "bx.h"
-
-namespace bx
-{
-	// http://cnicholson.net/2011/01/stupid-c-tricks-a-better-sizeof_array/
-	template<typename T, size_t N> char (&COUNTOF_REQUIRES_ARRAY_ARGUMENT(const T(&)[N]) )[N];
-#define countof(x) sizeof(bx::COUNTOF_REQUIRES_ARRAY_ARGUMENT(x) )
-
-} // namespace bx
-
-#endif // __BX_COUNTOF_H__
diff --git a/include/bx/macros.h b/include/bx/macros.h
index 76f150e..5add4ed 100644
--- a/include/bx/macros.h
+++ b/include/bx/macros.h
@@ -98,6 +98,14 @@
 #	define BX_WARN(_condition, ...) do {} while(0)
 #endif // BX_CHECK
 
+#ifndef BX_CONFIG_ALLOCATOR_DEBUG
+#	define BX_CONFIG_ALLOCATOR_DEBUG 0
+#endif // BX_CONFIG_DEBUG_ALLOC
+
+#ifndef BX_CONFIG_ALLOCATOR_CRT
+#	define BX_CONFIG_ALLOCATOR_CRT 1
+#endif // BX_CONFIG_ALLOCATOR_CRT
+
 #ifndef  BX_CONFIG_SPSCQUEUE_USE_MUTEX
 #	define BX_CONFIG_SPSCQUEUE_USE_MUTEX 0
 #endif // BX_CONFIG_SPSCQUEUE_USE_MUTEX