I have checked the official documentation and it can be stated that the decimal places are 17 and accuracy is guaranteed. After reading the source code of incrbyfloatCommand, I couldn't find a specific length representation for integer bits. How many verifications would you like to make?
- Can the incrbyfloat instruction support 17 bits of Floating-point arithmetic integer digits? My local test result is 17 bits. If it exceeds 18 bits, an overflow will occur?
- Does the incrbyfloat instruction cause accuracy issues when performing internal operations through long double?
- Is incrbyfloat suitable for accumulating and summarizing amounts, and what is the appropriate range?
Finally, thank you very much and look forward to hearing from community friends