Re: Revised patch to remove integer output macros

[Richard Henderson <rth at redhat dot com>]

On Sat, Dec 15, 2001 at 04:31:06PM +0000, Richard Sandiford wrote:
> I guess the question is: should target independent code be allowed to
> access the string ops directly?

I think so.

> If target independent code _can_ access the ops directly, then I think
> assemble_integer should too.  Otherwise we'd have a situation where the
> string ops would be used in preference to the hook in some places and
> the hook would be used first in others.

Yes, and I don't necessarily consider that a bad thing.

For instance, if you print a function's address on ia64,
you'll get the fptr thing.  But when emitting dwarf2 debug
info, we *don't* want that.

> The avr port always uses ".word" for aligned word-sized objects, so it
> really ought to able use TARGET_ASM_ALIGNED_SI_OP.  The fact that
> certain labels need to wrapped in %st() seems to have more to do
> with output_addr_const() than the fact that we're writing an aligned
> word.  The same applies to arc, arm, ia64 & pa.

True, but often that special relocation can't be used at an
unaligned address anyway.

> I thought perhaps we need (yet another!) target hook to control how the
> object itself is written.
 
See how stormy16 does it.  I'd rather not confuse the issue
by fixing the others right now.


r~