Re: [C++ PATCH] Fix ICE with inline asm and MODIFY_EXPR/preinc/predec in output operand (PR c++/84961)

[Jakub Jelinek <jakub@redhat.com>]

On Tue, Mar 20, 2018 at 05:58:43PM -0400, Jason Merrill wrote:
> On Tue, Mar 20, 2018 at 5:04 PM, Jakub Jelinek <jakub@redhat.com> wrote:
> > --- gcc/cp/semantics.c.jj       2018-03-20 11:58:17.069356145 +0100
> > +++ gcc/cp/semantics.c  2018-03-20 21:56:43.745292245 +0100
> > @@ -1512,6 +1512,26 @@ finish_asm_stmt (int volatile_p, tree st
> >                       && C_TYPE_FIELDS_READONLY (TREE_TYPE (operand)))))
> >             cxx_readonly_error (operand, lv_asm);
> >
> > +         tree *op = &operand;
> > +         while (TREE_CODE (*op) == COMPOUND_EXPR)
> > +           op = &TREE_OPERAND (*op, 1);
> > +         switch (TREE_CODE (*op))
> > +           {
> > +           case PREINCREMENT_EXPR:
> > +           case PREDECREMENT_EXPR:
> > +           case MODIFY_EXPR:
> > +             if (TREE_SIDE_EFFECTS (TREE_OPERAND (*op, 0)))
> > +               *op = build2 (TREE_CODE (*op), TREE_TYPE (*op),
> > +                             cp_stabilize_reference (TREE_OPERAND (*op, 0)),
> > +                             TREE_OPERAND (*op, 1));
> > +             *op = build2 (COMPOUND_EXPR, TREE_TYPE (*op), *op,
> > +                           TREE_OPERAND (*op, 0));
> > +             op = &TREE_OPERAND (*op, 1);
> > +             break;
> > +           default:
> > +             break;
> > +           }
> 
> Hmm, it would be nice to share this with the similar patterns in
> unary_complex_lvalue and cp_build_modify_expr.

You mean just outline the
      if (TREE_SIDE_EFFECTS (TREE_OPERAND (lhs, 0)))
        lhs = build2 (TREE_CODE (lhs), TREE_TYPE (lhs),
                      cp_stabilize_reference (TREE_OPERAND (lhs, 0)),
                      TREE_OPERAND (lhs, 1));
      lhs = build2 (COMPOUND_EXPR, lhstype, lhs, TREE_OPERAND (lhs, 0));
into lhs = some_function (lhs); and use that in finish_asm_stmt,
unary_complex_lvalue and cp_build_modify_expr in these spots?
I really have no idea how to commonize anything else, e.g. the COMPOUND_EXPR
handling is substantially different between the 3 functions.

Any suggestion for the some_function name if you want that?

> Does COND_EXPR work without adjustment?

It seems to be:
int a, b;

void
foo (bool x)
{
  asm volatile ("" : "=r" (x ? a : b));
}

void
bar (bool x)
{
  asm volatile ("" : : "m" (x ? a : b));
}

I guess it should be added as another testcase.

	Jakub