better comments explaining why we can use expressions

Author: karolzwolak

library/core/src/num/f128.rs

@@ -645,6 +645,9 @@ impl f128 {
     #[unstable(feature = "f128", issue = "116909")]
     #[must_use = "this returns the result of the operation, without modifying the original"]
     pub const fn to_degrees(self) -> Self {
+        // The division here is correctly rounded with respect to the true value of 180/π.
+        // Although π is irrational and already rounded, the double rounding happens
+        // to produce correct result for f128.
         const PIS_IN_180: f128 = 180.0 / consts::PI;
         self * PIS_IN_180
     }

library/core/src/num/f32.rs

@@ -872,6 +872,9 @@ impl f32 {
     #[rustc_const_stable(feature = "const_float_methods", since = "1.85.0")]
     #[inline]
     pub const fn to_radians(self) -> f32 {
+        // The division here is correctly rounded with respect to the true value of π/180.
+        // Although π is irrational and already rounded, the double rounding happens
+        // to produce correct result for f32.
         const RADS_PER_DEG: f32 = consts::PI / 180.0;
         self * RADS_PER_DEG
     }

library/core/src/num/f64.rs

@@ -869,9 +869,9 @@ impl f64 {
     #[rustc_const_stable(feature = "const_float_methods", since = "1.85.0")]
     #[inline]
     pub const fn to_degrees(self) -> f64 {
-        // The division here is correctly rounded with respect to the true
-        // value of 180/π. (This differs from f32, where a constant must be
-        // used to ensure a correctly rounded result.)
+        // The division here is correctly rounded with respect to the true value of 180/π.
+        // Although π is irrational and already rounded, the double rounding happens
+        // to produce correct result for f64.
         const PIS_IN_180: f64 = 180.0 / consts::PI;
         self * PIS_IN_180
     }
@@ -891,6 +891,9 @@ impl f64 {
     #[rustc_const_stable(feature = "const_float_methods", since = "1.85.0")]
     #[inline]
     pub const fn to_radians(self) -> f64 {
+        // The division here is correctly rounded with respect to the true value of π/180.
+        // Although π is irrational and already rounded, the double rounding happens
+        // to produce correct result for f64.
         const RADS_PER_DEG: f64 = consts::PI / 180.0;
         self * RADS_PER_DEG
     }