docs: clarify that float Rem formula uses exact arithmetic

[mango766]

Summary

Fixes #133758.

The documentation for <{f16,f32,f64,f128} as Rem>::rem previously described the remainder as being "computed as: x - (x / y).trunc() * y", which reads as a literal Rust expression. In reality, the operation is mathematically defined as exact x - trunc(x/y) * y (computed in infinite precision, then rounded to the result type), equivalent to C's fmod. Evaluating the expression literally in Rust produces different results for many inputs due to intermediate floating-point rounding.

For example, as noted in the issue:

let (x, y) = (11f64, 1.1f64);
// These are NOT equal:
assert_eq!(x - (x / y).trunc() * y, x % y);  // panics: left=0.0, right≈1.1

This PR:

• Rewrites the formula using mathematical notation (trunc(x / y)) rather than Rust method syntax, making it clear this is not meant as literal code
• Explicitly states the computation uses infinite precision arithmetic
• Adds a note explaining why the literal Rust expression differs
• Simplifies the example to avoid suggesting the expressions are equivalent

This aligns with the consensus in #133758 and the IEEE 754 / C fmod semantics documented in RFC 3514.

[rustbot]

r? @joboet

rustbot has assigned @joboet.
They will have a look at your PR within the next two weeks and either review your PR or reassign to another reviewer.

Use r? to explicitly pick a reviewer

Why was this reviewer chosen?

The reviewer was selected based on:

• Owners of files modified in this PR: @scottmcm, libs
• @scottmcm, libs expanded to 8 candidates
• Random selection from Mark-Simulacrum, joboet, scottmcm

[rustbot]

⚠️ Warning ⚠️

• There are issue links (such as #123) in the commit messages of the following commits.
  Please move them to the PR description, to avoid spamming the issues with references to the commit, and so this bot can automatically canonicalize them to avoid issues with subtree.
  • 56113f9