Re: [PATCH] lower-bitint: Force some arrays corresponding to large/huge _BitInt SSA_NAMEs to BLKmode

[Jakub Jelinek <jakub@redhat.com>]

On Thu, Jan 18, 2024 at 01:57:49PM +0100, Richard Biener wrote:
> > -     RTL expansion expectes TARGET_MEM_REF to always address actual memory.  */
> > +     RTL expansion expectes TARGET_MEM_REF to always address actual memory.
> > +     Also, force to stack non-BLKmode vars accessed through VIEW_CONVERT_EXPR
> > +     to BLKmode BITINT_TYPEs.  */
> >    else if (TREE_CODE (t) == TARGET_MEM_REF
> >  	   || (TREE_CODE (t) == MEM_REF
> >  	       && TYPE_SIZE (TREE_TYPE (t))
> > -	       && POLY_INT_CST_P (TYPE_SIZE (TREE_TYPE (t)))))
> > +	       && POLY_INT_CST_P (TYPE_SIZE (TREE_TYPE (t))))
> > +	   || (TREE_CODE (t) == VIEW_CONVERT_EXPR
> > +	       && TREE_CODE (TREE_TYPE (t)) == BITINT_TYPE
> > +	       && TYPE_MODE (TREE_TYPE (t)) == BLKmode))
> 
> I'm still not getting what's special about BITINT_TYPE here so
> shouldn't that apply to all BLKmode V_C_E?  But sure we can for
> now just handle BITINT_TYPE.
> 
> That hunk looks OK to me.

The == BITINT_TYPE check is non-essential, was just trying to keep existing
behavior otherwise.  I can certainly drop that.

> > --- gcc/expr.cc.jj	2024-01-12 10:07:58.194851657 +0100
> > +++ gcc/expr.cc	2024-01-18 13:38:19.677556646 +0100
> > @@ -12382,6 +12382,17 @@ expand_expr_real_1 (tree exp, rtx target
> >  	  }
> >        }
> >  
> > +      /* Ensure non-BLKmode array VAR_DECLs VCEd to BLKmode BITINT_TYPE
> > +	 aren't promoted to registers.  */
> > +      if (op0 == NULL_RTX
> > +	  && mode == BLKmode
> > +	  && TREE_CODE (type) == BITINT_TYPE
> > +	  && VAR_P (treeop0)
> > +	  && DECL_MODE (treeop0) != BLKmode
> > +	  && DECL_RTL_SET_P (treeop0)
> > +	  && MEM_P (DECL_RTL (treeop0)))
> > +	op0 = adjust_address (DECL_RTL (treeop0), BLKmode, 0);
> > +
> >        if (!op0)
> >  	op0 = expand_expr_real (treeop0, NULL_RTX, VOIDmode, modifier,
> >  				NULL, inner_reference_p);
> 
> So we're now sure we have MEM_P (op0) after expand_expr_real,
> even without this change, right?  What's wrong with the
> suggestion to use

I wasn't sure if VAR_P (treeop0) && MEM_P (DECL_RTL (treeop0)) implies that
expand_expr_real will return a MEM, but I'm not able to find a path in which
it would return something different, so maybe ok.

>   if (mode == GET_MODE (op0) || (mode == BLKmode && MEM_P (op0))
> 
> thus not run into any of the special-casing?  We're doing just

It is true the later code will then do:
> 
>           op0 = adjust_address (op0, mode, 0);

so perhaps it is ok as you wrote it (though perhaps adding it as a separate
else if would allow a separate comment).

	Jakub