Re: C structure padding

Dale R. Worley (drw@zermelo.mit.edu) writes:
> This leads up to a point that I've never seen satisfactorily resolved:
> Must a structure be padded when it is *not* part of an array? I've never
> seen anything in the Standard that makes it clear that such a structure
> must be padded, bit it seems clear that padding is necessary for common
> programming paradigms to work. Am I missing something?


> [It's pretty clear that all structures have to be treated the same, since
> if p and q are pointers to structures, you can write *p = *q regardless of
> whether they're pointing to simple structures or into an array. -John]


Right. Chapter and verse: section 3.5.2.1 (ANSI) / 6.5.2.1 (ISO),
Semantics, last paragraph, my emphasis added.


# There may also be unnamed padding at the end of a structure or union,
# as necessary to achieve the appropriate alignment *WERE* the structure
# or union to be an element of an array.


I may as well point out here the existence of comp.std.c, which is
specifically for topics about the C standard. If Dale had posted there,
he or she would no doubt have received half a dozen responses from people
familiar with the above text (followed, of course, by a minor flame war,
under the same Subject line, about some tangential topic or other...).
--
Mark Brader
SoftQuad Inc., Toronto
utzoo!sq!msb, msb@sq.com
--

Return to the comp.compilers page.
Search the comp.compilers archives again.