There are compiler directives for generating code in a specific linker segment instead of the default linker segment for processing audio in the AVRT library. This is an internal feature of Microsoft. Segments exist, probably for optimization or security. I found this header file that shows the value:
ftp://ns432777.ip-37-59-27.eu/Program%20Files%20%28x86%29/Windows%20Kits/8.0/Include/um/baseaudioprocessingobject.h
#define AVRT_CODE_BEGIN code_seg( push, "RT_CODE" )
#define AVRT_DATA_BEGIN data_seg( push, "RT_DATA" )
#define AVRT_BSS_BEGIN bss_seg( push, "RT_BSS" )
#define AVRT_CONST_BEGIN const_seg( push, "RT_CONST" )
#define AVRT_VTABLES_BEGIN AVRT_CONST_BEGIN
#define AVRT_CODE_END code_seg( pop )
#define AVRT_DATA_END data_seg( pop )
#define AVRT_BSS_END bss_seg( pop )
#define AVRT_CONST_END const_seg( pop )
#define AVRT_VTABLES_END AVRT_CONST_END
#define AVRT_DATA __declspec(allocate("RT_DATA"))
#define AVRT_BSS __declspec(allocate("RT_BSS"))
#define AVRT_CONST __declspec(allocate("RT_CONST"))