The oracle will not register nonsense/useless relations, (basically X op
X). Symbolically, x == x is implied, and x != x, x< x, etc are all
nonsense.
Now that we are using class value_relation in a couple of other places,
it shouldn't either.
Bootstrapped on x86_64-pc-linux-gnu with no regressions. Pushed.
Andrew
From fca529517484bf19098ca9efa77e95534086abdc Mon Sep 17 00:00:00 2001
From: Andrew MacLeod <amacl...@redhat.com>
Date: Fri, 14 Oct 2022 14:31:02 -0400
Subject: [PATCH 1/4] Don't set useless relations.
The oracle will not register nonssense/useless relations, class
value_relation shouldn't either.
* value-relation.cc (value_relation::dump): Change message.
* value-relation.h (value_relation::set_relation): If op1 is the
same as op2 do not create a relation.
---
gcc/value-relation.cc | 2 +-
gcc/value-relation.h | 7 +++++++
2 files changed, 8 insertions(+), 1 deletion(-)
diff --git a/gcc/value-relation.cc b/gcc/value-relation.cc
index 50fc190a36b..3fb7b96c9e0 100644
--- a/gcc/value-relation.cc
+++ b/gcc/value-relation.cc
@@ -891,7 +891,7 @@ value_relation::dump (FILE *f) const
{
if (!name1 || !name2)
{
- fprintf (f, "uninitialized");
+ fprintf (f, "no relation registered");
return;
}
fputc ('(', f);
diff --git a/gcc/value-relation.h b/gcc/value-relation.h
index a3bbe1e8157..fa9097a8069 100644
--- a/gcc/value-relation.h
+++ b/gcc/value-relation.h
@@ -349,6 +349,13 @@ value_relation::set_relation (relation_kind r, tree n1, tree n2)
{
gcc_checking_assert (TREE_CODE (n1) == SSA_NAME
&& TREE_CODE (n2) == SSA_NAME);
+ if (n1 == n2)
+ {
+ related = VREL_VARYING;
+ name1 = NULL_TREE;
+ name2 = NULL_TREE;
+ return;
+ }
related = r;
name1 = n1;
name2 = n2;
--
2.37.3
