> It seems to me that you are right,
> but there is other cases; for example:
> X := T(0)*256 + T(1);
> compiled on a big endian architecture may well generate just one
> 16-bit word read...
Not in my word, not in gcc's world. That is just broken and wrong.
If you want/need the gcc doc to expound on this, write it up, and
we'll add it.
To unsubscribe from this list: send the line "unsubscribe linux-kernel" in
the body of a message to firstname.lastname@example.org
More majordomo info at http://vger.kernel.org/majordomo-info.html
Please read the FAQ at http://www.tux.org/lkml/