[libc-commits] [libc] [libc][math] Adding LIBC_MATH_ASSUME_ROUND_NEAREST_ONLY option (PR #201154)

[via libc-commits]

=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,
=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,
=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,
=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=,
=?utf-8?q?Ho=C3=A0ng_Minh_Thi=C3=AAn?=
Message-ID:
In-Reply-To: <llvm.org/llvm/llvm-project/pull/201154 at github.com>


================
@@ -190,6 +202,10 @@ template <size_t Bits> struct DyadicFloat {
         raise_except_if_required(FE_OVERFLOW | FE_INEXACT);
       }
 
+#ifdef LIBC_MATH_HAS_ASSUME_ROUND_NEAREST_ONLY
+      return FPBits::inf(sign).get_val();
+#endif // LIBC_MATH_HAS_ASSUME_ROUND_NEAREST_ONLY
----------------
lntue wrote:

put the `switch` below to the `#else` clause.

https://github.com/llvm/llvm-project/pull/201154