Hopefully there is someone familiar with how LV calls into Frameworks from the call Library node.
(Note for those who are cross platform challenged, use "shared library" or .so file for linux, and "dll" for windows in place of "Framework" under Mac OS X)
With LV 8.5 there is a nice "Import Framework" into a project. This has a nice wizard interface. If I import a System Framework and find all the header files that are needed, I still get a bunch of "undefined variables". Here is a list below:
KERNEL;_ANSI_SOURCE;_APPLE_C_SOURCE;_NONSTD_SOURCE;_POSIX_C_SOURCE;_POSIX_SOURCE;_XOPEN_SOURCE;__APPLE_CC__;__DARWIN_UNIX03;__DBL_MANT_DIG__;__ENVIRONMENT_MAC_OS_X_VERSION_MIN_REQUIRED__;__GNUC_MINOR__;__GNUC__;__INTMAX_TYPE__;__LDBL_MANT_DIG__;__LONG_DOUBLE_128__;__LP64__;__MWERKS__;__SSE2__;__SSE__;__STDC_CONSTANT_MACROS;__STDC_LIMIT_MACROS;__STDC__;__STDDEF_H__;__STRICT_ANSI__;__UINTMAX_TYPE__;__VEC__;__WANT_LONG_DOUBLE_FORMAT__;__WCHAR_MAX__;__cplusplus;__i386__;__need_NULL;__need_ptrdiff_t;__need_size_t;__need_wchar_t;__need_wint_t;__ppc64__;__ppc__;__x86_64__;
These are all defined by the gcc compiler and so work just fine when I use these headers with gcc. But of course are not defined for LabVIEW. Is there a magic header file somewhere that contains all these definitions for my particular environment?
The particular problem I have is trying to call into the vector FFT accelerate library for performance gains.