Re:The first split vote about C compiler random test
by prodigal_erik on Thursday July 14, 2011 @03:57AM
(#2762125)
Thank you, I wasn't aware that C99 now mandates the twos-complement result; I think it had been implementation-defined in C89 (which is all I ever used seriously).
http://stackoverflow.com/questions/50605/signed-to-unsigned-...
Re:The first split vote about C compiler random test
by caf on Thursday July 14, 2011 @07:57PM
(#2765701)
C89 has this language in §3.1.2.5 Types:
A computation involving unsigned operands can
never overflow, because a result that cannot be
represented by the resulting unsigned integer type
is reduced modulo the number that is one greater
than the largest value that can be represented by
the resulting unsigned integer type.
Conversion of out-of-range values to signed types is, however, implementation-defined (and also allows an implementation-defined signal to be raised, which in practice makes it something to avoid in correct programs).