Fixes compilation on GCC/MinGW

Author: Gravecat

include/fast_float/float_common.h

@@ -407,7 +407,7 @@ full_multiplication(uint64_t a, uint64_t b) {
   answer.high = __umulh(a, b);
   answer.low = a * b;
 #elif defined(FASTFLOAT_32BIT) ||                                              \
-    (defined(_WIN64) && !defined(__clang__) && !defined(_M_ARM64))
+    (defined(_WIN64) && !defined(__clang__) && !defined(_M_ARM64) && !defined(__GNUC__))
   answer.low = _umul128(a, b, &answer.high); // _umul128 not available on ARM64
 #elif defined(FASTFLOAT_64BIT) && defined(__SIZEOF_INT128__)
   __uint128_t r = ((__uint128_t)a) * b;