This is the mail archive of the
gcc@gcc.gnu.org
mailing list for the GCC project.
Re: Incorrect C++ code or regression?
- From: "Giovanni Bajo" <rasky at develer dot com>
- To: <gcc at gcc dot gnu dot org>
- Cc: <devnull at kma dot eu dot org>
- Date: Mon, 24 Jan 2005 17:22:21 +0100
- Subject: Re: Incorrect C++ code or regression?
- References: <41F51D1D.7030800@kma.eu.org>
Grumble <devnull@kma.eu.org> wrote:
> I've tried to make a testcase from the original code.
> GCC 3.3.5 compiles with no warning, while GCC 3.4.3 rejects it.
> Can somebody tell me if it because the code is illegal, and GCC 3.4.3's
> stricter parser now rejects it, or is it some kind of regression?
http://gcc.gnu.org/gcc-3.4/changes.html
As a matter of fact, fixing applications to compile with 3.4 could be very
hard, and require at least some basic understanding of C++ and templates.
> template <class WHIRL>
> void
> WN_TREE_ITER<PRE_ORDER, WHIRL>::WN_TREE_next ()
> {
> if (_wn == 0) return;
> }
Bug here, use this->_wn instead.
--
Giovanni Bajo