Re: Use of unaligned load/stores by compilers

Sazal <sazal@aol.com> wrote:
> I was wondering if anyone can point out the different capacities in
>which compilers use unaligned load/store instructions. So far I have
>seen them used in structure copy blocks. Can a architecture whose
>code is solely generated by compilers live without unaligned
>load/stores. Any references will be greatly apppreciated.




>[Any architecture can live without unaligned loads and stores. The
>IBM 360 didn't have them. (The 370 added them for reasons I don't
>totally grasp.) Most RISC chips don't have them.


>Fortran lets you use equivalence statements to misalign double precision
>floats, other than that I don't know of any languages where misaligned
>data will happen without extraordinary programmer effort. -John]


It depends on what you mean by extraordinary effort. Of course, one
cannot do unaligned loads/stores if the architecture automatically
blocks the process, although one can simulate it without what I would
call extraordinary effort. And, while it is technically not legal in
C, most implementations do not check pointer casts to see if they are
properly aligned. This is certainly not extraordinary effort.
--
Herman Rubin, Dept. of Statistics, Purdue Univ., West Lafayette IN47907-1399
hrubin@stat.purdue.edu Phone: (765)494-6054 FAX: (765)494-0558
--

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