http://bugs.winehq.org/show_bug.cgi?id=6677
------- Additional Comments From the3dfxdude@gmail.com 2006-13-11 14:44 ------- I'm not saying that you're wrong. I'm saying that size checking is unneccesary in pf_integer_conv as is. All that function does is return a string representation of a 64-bit integer. It doesn't output fields, precision, or signs, etc. If you want to see the maximum output length of the function, do something like this:
printf("%I64u\n", (LONGLONG)-1);
If what was converted for I64u was within 40, then that is all that is necessary.
Octal is slightly different, but it's still within bounds. I know all this because I wrote the function and I checked to see if the function's output fit in the buffer in all cases.