This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug tree-optimization/17528] [4.0 Regression] tree optimization incorrectly optimizes away entire function (incorrect aliasing detection?)
- From: "dnovillo at redhat dot com" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 16 Sep 2004 22:15:46 -0000
- Subject: [Bug tree-optimization/17528] [4.0 Regression] tree optimization incorrectly optimizes away entire function (incorrect aliasing detection?)
- References: <20040916215056.17528.rearnsha@gcc.gnu.org>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Additional Comments From dnovillo at redhat dot com 2004-09-16 22:15 -------
Subject: Re: [4.0 Regression] tree
optimization incorrectly optimizes away entire function (incorrect aliasing
detection?)
On Thu, 2004-09-16 at 18:10, pinskia at gcc dot gnu dot org wrote:
> ------- Additional Comments From pinskia at gcc dot gnu dot org 2004-09-16 22:10 -------
> hmm, DCE is removing a COND_EXPR and some BB for some reason:
> if (D.19406_5 != 0B) goto <L0>; else goto <L2>;
>
> I cannot figure out why yet.
>
One of the libstdc++ tests I'm looking at has similar symptoms. We take
the address of a variable that makes its way into some field in a
structure which we fail to relate in the alias analyzer.
Still not completely sure why the alias sets of the two objects don't
conflict.
Diego.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=17528