The current MMCSD_read & MMCSD_write function, which are designed to support reading any buffer size, are not designed for simplicity & readability.

The implementation looks like this:

• If #blocks are greater than max #blocks, split the transactions as:
  • Loop for multiple transactions with max #blocks.
  • Separate if condition to handle the leftover blocks (< max #blocks).
• Else, a single transaction for #blocks < max #blocks.

There are too many branches here which makes the code unnecessarily complicated & unreadable.

The implementation needs refactoring for simplicity, readability, and to show quality.

One simple implementation is to club everything into a single loop with conditionally assigning the #blocks, buf, cmd for the transaction like shown in the attached file.