Clarify in the docs that mul_add is not always faster.

Fixes https://github.com/rust-lang/rust/issues/49842.

Other resources:

- https://users.rust-lang.org/t/why-does-the-mul-add-method-produce-a-more-accurate-result-with-better-performance/1626
- https://en.wikipedia.org/wiki/Multiply%E2%80%93accumulate_operation

src/libstd/f32.rs

@@ -195,8 +195,10 @@ impl f32 {
     }
 
     /// Fused multiply-add. Computes `(self * a) + b` with only one rounding
-    /// error. This produces a more accurate result with better performance than
-    /// a separate multiplication operation followed by an add.
+    /// error, yielding a more accurate result than an unfused multiply-add.
+    ///
+    /// Using `mul_add` can be more performant than an unfused multiply-add if
+    /// the target architecture has a dedicated `fma` CPU instruction.
     ///
     /// ```
     /// use std::f32;

src/libstd/f64.rs

@@ -173,8 +173,10 @@ impl f64 {
     }
 
     /// Fused multiply-add. Computes `(self * a) + b` with only one rounding
-    /// error. This produces a more accurate result with better performance than
-    /// a separate multiplication operation followed by an add.
+    /// error, yielding a more accurate result than an unfused multiply-add.
+    ///
+    /// Using `mul_add` can be more performant than an unfused multiply-add if
+    /// the target architecture has a dedicated `fma` CPU instruction.
     ///
     /// ```
     /// let m = 10.0_f64;