This is the mail archive of the
gcc-bugs@gcc.gnu.org
mailing list for the GCC project.
[Bug tree-optimization/28868] [4.0/4.1/4.2 Regression] Not elimintating the PHIs which have the same agruments
- From: "dberlin at gcc dot gnu dot org" <gcc-bugzilla at gcc dot gnu dot org>
- To: gcc-bugs at gcc dot gnu dot org
- Date: 29 Aug 2006 14:59:34 -0000
- Subject: [Bug tree-optimization/28868] [4.0/4.1/4.2 Regression] Not elimintating the PHIs which have the same agruments
- References: <bug-28868-6528@http.gcc.gnu.org/bugzilla/>
- Reply-to: gcc-bugzilla at gcc dot gnu dot org
------- Comment #6 from dberlin at gcc dot gnu dot org 2006-08-29 14:59 -------
SCCVN comes up with
Value numbers:
d_2 = c_1
c_7 = b_6
d_8 = b_6
c_10 = a_9
d_11 = a_9
where
# PRED: 3 (fallthru,exec) 4 (fallthru,exec)
# dD.1526_2 = PHI <dD.1526_11(3), dD.1526_8(4)>;
# cD.1525_1 = PHI <cD.1525_10(3), cD.1525_7(4)>;
...
As a result, once integrated into PRE/FRE, it will eliminate uses of the d_2
phi with the c_1 phi, which is what you want.
--
http://gcc.gnu.org/bugzilla/show_bug.cgi?id=28868