Returns 'Error: value out of range'
I guess it's because some changes made recently to avoid infinite loops when preprocessing or something like that, but imho such changes shouldn't stop any programmer from making such calculations as above. Is it a bug or will it stay like this?

Read the WHATSNEW.TXT - it's explained there.

For the reason look also here: http://board.flatassembler.net/topic.php?t=3844
The 65-bit storage would need many changes in expression calculator and symbols storage, so for now I've just added the overflow checking to avoid situations when multiplying two positive numbers gives negative result.

For that particular calculation you can use this: Since the lower 32 bits of any 32x32 multiply are not affected by pos/neg number signs.

Ok, big thanks for the answer revolution. I'll use this method