This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug middle-end/39885] Missed fre optimization
- From: "rguenth at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 24 Apr 2009 15:30:20 -0000
- Subject: [Bug middle-end/39885] Missed fre optimization
- References: <bug-39885-511@http.gcc.gnu.org/bugzilla/>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Comment #7 from rguenth at gcc dot gnu dot org 2009-04-24 15:30 -------
But in the end the different types are really the problem. Several issues
in the value-numberer though:
- throwing stmts are not value numbered
- pointers are needlessly value numbered with their types
- reference ops all have types, but during valueization the types are not
valueized
- component refs refer to field decls, thus reference ops are not only
based on offset/size of the access.
I am testing if value-numbering throwing stmts will pass Java testing. That
helps functions besides foo at least.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=39885