On Fri, Mar 3, 2023, at 16:11, Dean Rasheed wrote: > Attachments: > * make-result-using-vars-buf-v2.patch
One suggestion: maybe add a comment explaining why the allocated buffer which size is based on strlen(cp) for the decimal digit values, is guaranteed to be large enough also for the result's digit buffer? I.e. some kind of proof why (NUMERIC_HDRSZ + strlen(cp) + DEC_DIGITS * 2) >= ((ndigits + 1) * sizeof(NumericDigit))) holds true in general. /Joel