arm.md (movsi): Don't split symbol refs here.

2012-07-10  Julian Brown  <julian@codesourcery.com>

	* config/arm/arm.md (movsi): Don't split symbol refs here.
	(define_split): New.

From-SVN: r189411

gcc/ChangeLog

+2012-07-10  Julian Brown  <julian@codesourcery.com>
+
+	* config/arm/arm.md (movsi): Don't split symbol refs here.
+	(define_split): New.
+
 2012-07-10  Andreas Schwab  <schwab@linux-m68k.org>
 
 	PR bootstrap/53913

gcc/config/arm/arm.md

@@ -5475,14 +5475,6 @@
 			       optimize && can_create_pseudo_p ());
           DONE;
         }
-
-      if (TARGET_USE_MOVT && !target_word_relocations
-	  && GET_CODE (operands[1]) == SYMBOL_REF
-	  && !flag_pic && !arm_tls_referenced_p (operands[1]))
-	{
-	  arm_emit_movpair (operands[0], operands[1]);
-	  DONE;
-	}
     }
   else /* TARGET_THUMB1...  */
     {
@@ -5591,6 +5583,24 @@
   "
 )
 
+;; Split symbol_refs at the later stage (after cprop), instead of generating
+;; movt/movw pair directly at expand.  Otherwise corresponding high_sum
+;; and lo_sum would be merged back into memory load at cprop.  However,
+;; if the default is to prefer movt/movw rather than a load from the constant
+;; pool, the performance is better.
+(define_split
+  [(set (match_operand:SI 0 "arm_general_register_operand" "")
+       (match_operand:SI 1 "general_operand" ""))]
+  "TARGET_32BIT
+   && TARGET_USE_MOVT && GET_CODE (operands[1]) == SYMBOL_REF
+   && !flag_pic && !target_word_relocations
+   && !arm_tls_referenced_p (operands[1])"
+  [(clobber (const_int 0))]
+{
+  arm_emit_movpair (operands[0], operands[1]);
+  DONE;
+})
+
 (define_insn "*thumb1_movsi_insn"
   [(set (match_operand:SI 0 "nonimmediate_operand" "=l,l,l,l,l,>,l, m,*l*h*k")
 	(match_operand:SI 1 "general_operand"      "l, I,J,K,>,l,mi,l,*l*h*k"))]