From: Andrew Cox <andrew.cox@partner.samsung.com>
Date: Tue, 23 Dec 2014 19:35:16 +0000 (+0000)
Subject: Replaced shift and add sequence with single multiply in hash function
X-Git-Tag: dali_1.0.24~6^2
X-Git-Url: http://review.tizen.org/git/?a=commitdiff_plain;h=43c28815b34ba666d007c346c11753ed485496d0;p=platform%2Fcore%2Fuifw%2Fdali-core.git

Replaced shift and add sequence with single multiply in hash function

Archaic optimisation to think a sequence of instructions with
pipeline-stalling dependencies beats a single imul.
Compilers can still make the substitution with a shift and add if they
want to for a particular platform.

Change-Id: I313a5f98eae665f0fabd266fee09c078547be2ec
Signed-off-by: Andrew Cox <andrew.cox@partner.samsung.com>
---

diff --git a/dali/internal/common/dali-hash.cpp b/dali/internal/common/dali-hash.cpp
index f154b14..4101414 100644
--- a/dali/internal/common/dali-hash.cpp
+++ b/dali/internal/common/dali-hash.cpp
@@ -33,7 +33,7 @@ inline void HashShader( const char* string, unsigned long& hash )
 {
   while( int c = *string++ )
   {
-    hash = ((hash << 5) + hash) + c; /* hash * 33 + c */
+    hash = hash * 33 + c;
   }
 }
 } // un-named namespace