Decimal floating‐point fused multiply‐add with redundant internal encodings. Issue 4 (1st July 2016)