Re: [PATCH 8/9] Introduce class function_reader (v4)

[Bernd Schmidt <bschmidt at redhat dot com>]

On 12/01/2016 10:43 PM, David Malcolm wrote:
> > > +  rtx_insn *jump_insn = get_insn_by_uid (1);
> > > +  ASSERT_EQ (JUMP_INSN, GET_CODE (jump_insn));
> > > +  ASSERT_EQ (ret_rtx, JUMP_LABEL (jump_insn));
> > > +  // FIXME: ^^^ use ASSERT_RTX_PTR_EQ here ^^^
> >
> > Why is this a fixme and not just done that way (several instances)?
>
> ASSERT_RTX_PTR_EQ doesn't exist yet; there was some discussion about it
> in earlier versions of the patch, but I haven't written it.  It would
> be equivalent to ASSERT_EQ, checking pointer equality, but would dump
> the mismatching expected vs actual rtx on failure.
>
> Should I convert this to a TODO, or go ahead and implement
> ASSERT_RTX_PTR_EQ?

Missed this question. Just add ASSERT_RTX_PTR_EQ, that shouldn't be 
hard, should it?


Bernd