src/share/replaygain_synthesis/ : Remove old fast float math hack.

Old code was for the Intel compiler, had no appreciable performance benefits
and doesn't actually work with current version of that compiler.

Patch-from: lvqcl <lvqcl.mail@gmail.com>

src/share/Makefile.am

@@ -31,7 +31,6 @@ EXTRA_DIST = \
 	replaygain_analysis/replaygain_analysis_static.vcproj \
 	replaygain_synthesis/Makefile.lite \
 	replaygain_synthesis/replaygain_synthesis_static.vcproj \
-	replaygain_synthesis/include/private/fast_float_math_hack.h \
 	utf8/Makefile.lite \
 	utf8/charmaps.h \
 	utf8/makemap.c \

src/share/replaygain_synthesis/include/private/fast_float_math_hack.h

@@ -1,39 +0,0 @@
-#   ifdef __ICL /* only Intel C compiler has fmath ??? */
-
-    #include <mathf.h>
-
-/* Nearest integer, absolute value, etc. */
-
-    #define ceil ceilf
-    #define fabs fabsf
-    #define floor floorf
-    #define fmod fmodf
-    #define rint rintf
-    #define hypot hypotf
-
-/* Power functions */
-
-    #define pow powf
-    #define sqrt sqrtf
-
-/* Exponential and logarithmic functions */
-
-    #define exp expf
-    #define log logf
-    #define log10 log10f
-
-/* Trigonometric functions */
-
-    #define acos acosf
-    #define asin asinf
-    #define atan atanf
-    #define cos cosf
-    #define sin sinf
-    #define tan tanf
-
-/* Hyperbolic functions */
-    #define cosh coshf
-    #define sinh sinhf
-    #define tanh tanhf
-
-#   endif

src/share/replaygain_synthesis/replaygain_synthesis.c

@@ -41,7 +41,6 @@
 
 #include <string.h> /* for memset() */
 #include <math.h>
-#include "private/fast_float_math_hack.h"
 #include "replaygain_synthesis.h"
 #include "FLAC/assert.h"
 

src/share/replaygain_synthesis/replaygain_synthesis_static.vcproj

@@ -155,10 +155,6 @@
 			Filter="h;hpp;hxx;hm;inl;inc;xsd"
 			UniqueIdentifier="{93995380-89BD-4b04-88EB-625FBE52EBFB}"
 			>
-			<File
-				RelativePath=".\include\private\fast_float_math_hack.h"
-				>
-			</File>
 		</Filter>
 		<Filter
 			Name="Public Header Files"