math.bits: optimizations, tweaks to doc comments

[JalonSolov]

• optimized trailing_zeros_* routines to use existing ntz_8_tab table
• switched from parallel-bit-summing to calling ones_count_default
• switched to direct byte-lane swaps in reverse_bytes_* routines
• added a fast path to div_64_default if hi == 0
• removed need for de Bruijn/popcount mask constants at top
• added direct_array_access attribute where table paths used

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 201dfabf2a

ℹ️ About Codex in GitHub

Codex has been enabled to automatically review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

When you sign up for Codex through ChatGPT, Codex can also answer questions or update the PR, like "@codex address that feedback".