Use new impl in decimal32_fast

Author: mborland

include/boost/decimal/decimal32_fast.hpp

@@ -820,17 +820,10 @@ constexpr auto operator-(decimal32_fast lhs, decimal32_fast rhs) noexcept -> dec
     }
     #endif
 
-    if (!lhs.isneg() && rhs.isneg())
-    {
-        return lhs + (-rhs);
-    }
-
-    const bool abs_lhs_bigger {abs(lhs) > abs(rhs)};
-
-    return detail::sub_impl<decimal32_fast>(
+    return detail::new_sub_impl<decimal32_fast>(
             lhs.significand_, lhs.biased_exponent(), lhs.sign_,
             rhs.significand_, rhs.biased_exponent(), rhs.sign_,
-            abs_lhs_bigger
+            abs(lhs) > abs(rhs)
             );
 }
 
@@ -850,18 +843,13 @@ constexpr auto operator-(decimal32_fast lhs, Integer rhs) noexcept
 
     auto sig_rhs {static_cast<promoted_significand_type>(detail::make_positive_unsigned(rhs))};
 
-    if (!lhs.isneg() && (rhs < 0))
-    {
-        return lhs + sig_rhs;
-    }
-
     const bool abs_lhs_bigger {abs(lhs) > sig_rhs};
 
     exp_type exp_rhs {0};
     detail::normalize(sig_rhs, exp_rhs);
     auto final_sig_rhs {static_cast<decimal32_fast::significand_type>(detail::make_positive_unsigned(sig_rhs))};
 
-    return detail::sub_impl<decimal32_fast>(
+    return detail::new_sub_impl<decimal32_fast>(
             lhs.significand_, lhs.biased_exponent(), lhs.sign_,
             final_sig_rhs, exp_rhs, (rhs < 0),
             abs_lhs_bigger);
@@ -881,19 +869,14 @@ constexpr auto operator-(Integer lhs, decimal32_fast rhs) noexcept
     }
     #endif
 
-    if (lhs >= 0 && rhs.isneg())
-    {
-        return lhs + (-rhs);
-    }
-
     auto sig_lhs {static_cast<promoted_significand_type>(detail::make_positive_unsigned(lhs))};
     const bool abs_lhs_bigger {sig_lhs > abs(rhs)};
 
     exp_type exp_lhs {0};
     detail::normalize(sig_lhs, exp_lhs);
     auto final_sig_lhs {static_cast<decimal32_fast::significand_type>(detail::make_positive_unsigned(sig_lhs))};
 
-    return detail::sub_impl<decimal32_fast>(
+    return detail::new_sub_impl<decimal32_fast>(
             final_sig_lhs, exp_lhs, (lhs < 0),
             rhs.significand_, rhs.biased_exponent(), rhs.sign_,
             abs_lhs_bigger