Replace the ad-hoc hashing in GVN with the new hashing infrastructure.

This implicitly fixes a nasty bug in the GVN hashing (that thankfully
could only manifest as a performance bug): actually include the opcode
in the hash. The old code started the hash off with the opcode, but then
overwrote it with the type pointer.

Since this is likely to be pretty hot (GVN being already pretty
expensive) I've included a micro-optimization to just not bother with
the varargs hashing if they aren't present. I can't measure any change
in GVN performance due to this, even with a big test case like Duncan's
sqlite one. Everything I see is in the noise floor. That said, this
closes a loop hole for a potential scaling problem due to collisions if
the opcode were the differentiating aspect of the expression.

git-svn-id: https://llvm.org/svn/llvm-project/llvm/trunk@152025 91177308-0d34-0410-b5e6-96231b3b80d8

diff --git a/lib/Transforms/Scalar/GVN.cpp b/lib/Transforms/Scalar/GVN.cpp
index dc13edff2555cda456d51d4e85c914fcca5badd0..bc32b7f44181d68a1a1635272d0e35616562991c 100644 (file)
--- a/lib/Transforms/Scalar/GVN.cpp
+++ b/lib/Transforms/Scalar/GVN.cpp
@@ -36,6 +36,7 @@
 #include "llvm/Transforms/Utils/SSAUpdater.h"
 #include "llvm/ADT/DenseMap.h"
 #include "llvm/ADT/DepthFirstIterator.h"
+#include "llvm/ADT/Hashing.h"
 #include "llvm/ADT/SmallPtrSet.h"
 #include "llvm/ADT/Statistic.h"
 #include "llvm/Support/Allocator.h"
@@ -84,6 +85,16 @@ namespace {
         return false;
       return true;
     }
+
+    friend hash_code hash_value(const Expression &Value) {
+      // Optimize for the common case.
+      if (Value.varargs.empty())
+        return hash_combine(Value.opcode, Value.type);
+
+      return hash_combine(Value.opcode, Value.type,
+                          hash_combine_range(Value.varargs.begin(),
+                                             Value.varargs.end()));
+    }
   };
 
   class ValueTable {
@@ -130,16 +141,8 @@ template <> struct DenseMapInfo<Expression> {
   }
 
   static unsigned getHashValue(const Expression e) {
-    unsigned hash = e.opcode;
-
-    hash = ((unsigned)((uintptr_t)e.type >> 4) ^
-            (unsigned)((uintptr_t)e.type >> 9));
-
-    for (SmallVector<uint32_t, 4>::const_iterator I = e.varargs.begin(),
-         E = e.varargs.end(); I != E; ++I)
-      hash = *I + hash * 37;
-    
-    return hash;
+    using llvm::hash_value;
+    return static_cast<unsigned>(hash_value(e));
   }
   static bool isEqual(const Expression &LHS, const Expression &RHS) {
     return LHS == RHS;