🚀 go-pugleaf

pds@lemming.webo.dg.com (Paul D. Smith) writes:
>I would
>expect that on at least some systems the compiler will happily allow
>you to assign 0x01381 (or some other non-aligned value) to an int*,
>say, and then just core with a bus error (or otherwise die) when you
...
>In fact I'd much rather it did that than quietly modify my pointer by
>dropping bits... it would be a *major* pain to find *that* bug! :-)

One common reason for dropping the additional accuracy is when the pointers are
actually of different format, so that the pointer-to-int cannot represent the
finer-grained pointing.  For example, the machine may be word-addressed, so
that (say) only every four eight-bit-bytes does the address go up by one (so
like the pointers to adjacent ints will have adjacent internal values).
Then a pointer-to-char may be larger than a pointer-to-int and contain the
pointer-to-int's information plus a number which is 0, 1, 2, or 3.  A cast from
pointer-to-char to pointer-to-int would simply have to throw away this second
number.