Integers larger than or equal to 2^62 are truncated to 31-bit signed int on 64-bit

[Per Östlund]

Model:

model m
  Integer i = 4611686018427387904 //2^62;
end m;

Message from OMC:

Warning: OpenModelica (64-bit) only supports 63-bit signed integers! Truncating integer: 4611686018427387904 to 1073741823

i should be truncated to 2^{62-1, not 2}30-1.

[Martin Sjölund]

The problem here is that the RML runtime mk_icon only accepts "int" as input; not long. External functions cannot create long enough integer constants :)

Edit: Used to be this way. Apparently it's changed