rs6000.c (map_to_integral_tree_type): New helper function.

[gcc]

2018-10-09  Will Schmidt <will_schmidt@vnet.ibm.com>

	* config/rs6000/rs6000.c (map_to_integral_tree_type): New helper
	function.
	(fold_mergeeo_helper): New helper function.
	(rs6000_gimple_fold_builtin): Add hooks for vec_mergee and vec_mergeo
	intrinsics.  Correct some whitespace indentation issues.

From-SVN: r265063

gcc/ChangeLog

@@ -1,3 +1,11 @@
+2018-10-11  Will Schmidt <will_schmidt@vnet.ibm.com>
+
+	* config/rs6000/rs6000.c (map_to_integral_tree_type): New helper
+	function.
+	(fold_mergeeo_helper): New helper function.
+	(rs6000_gimple_fold_builtin): Add hooks for vec_mergee and vec_mergeo
+	intrinsics.  Correct some whitespace indentation issues.
+
 2018-10-11  Wilco Dijkstra  <wdijkstr@arm.com>
 
 	PR target/87511

gcc/config/rs6000/rs6000.c

@@ -15238,6 +15238,25 @@ fold_compare_helper (gimple_stmt_iterator *gsi, tree_code code, gimple *stmt)
   gsi_replace (gsi, g, true);
 }
 
+/* Helper function to map V2DF and V4SF types to their
+ integral equivalents (V2DI and V4SI).  */
+tree map_to_integral_tree_type (tree input_tree_type)
+{
+  if (INTEGRAL_TYPE_P (TREE_TYPE (input_tree_type)))
+    return input_tree_type;
+  else
+    {
+      if (types_compatible_p (TREE_TYPE (input_tree_type),
+			      TREE_TYPE (V2DF_type_node)))
+	return V2DI_type_node;
+      else if (types_compatible_p (TREE_TYPE (input_tree_type),
+				   TREE_TYPE (V4SF_type_node)))
+	return V4SI_type_node;
+      else
+	gcc_unreachable ();
+    }
+}
+
 /* Helper function to handle the vector merge[hl] built-ins.  The
    implementation difference between h and l versions for this code are in
    the values used when building of the permute vector for high word versus
@@ -15260,19 +15279,7 @@ fold_mergehl_helper (gimple_stmt_iterator *gsi, gimple *stmt, int use_high)
      float types, the permute type needs to map to the V2 or V4 type that
      matches size.  */
   tree permute_type;
-  if (INTEGRAL_TYPE_P (TREE_TYPE (lhs_type)))
+  permute_type = map_to_integral_tree_type (lhs_type);
-    permute_type = lhs_type;
-  else
-    {
-      if (types_compatible_p (TREE_TYPE (lhs_type),
-			      TREE_TYPE (V2DF_type_node)))
-	permute_type = V2DI_type_node;
-      else if (types_compatible_p (TREE_TYPE (lhs_type),
-				   TREE_TYPE (V4SF_type_node)))
-	permute_type = V4SI_type_node;
-      else
-	gcc_unreachable ();
-    }
   tree_vector_builder elts (permute_type, VECTOR_CST_NELTS (arg0), 1);
 
   for (int i = 0; i < midpoint; i++)
@@ -15290,6 +15297,40 @@ fold_mergehl_helper (gimple_stmt_iterator *gsi, gimple *stmt, int use_high)
   gsi_replace (gsi, g, true);
 }
 
+/* Helper function to handle the vector merge[eo] built-ins.  */
+static void
+fold_mergeeo_helper (gimple_stmt_iterator *gsi, gimple *stmt, int use_odd)
+{
+  tree arg0 = gimple_call_arg (stmt, 0);
+  tree arg1 = gimple_call_arg (stmt, 1);
+  tree lhs = gimple_call_lhs (stmt);
+  tree lhs_type = TREE_TYPE (lhs);
+  int n_elts = TYPE_VECTOR_SUBPARTS (lhs_type);
+
+  /* The permute_type will match the lhs for integral types.  For double and
+     float types, the permute type needs to map to the V2 or V4 type that
+     matches size.  */
+  tree permute_type;
+  permute_type = map_to_integral_tree_type (lhs_type);
+
+  tree_vector_builder elts (permute_type, VECTOR_CST_NELTS (arg0), 1);
+
+ /* Build the permute vector.  */
+  for (int i = 0; i < n_elts / 2; i++)
+    {
+      elts.safe_push (build_int_cst (TREE_TYPE (permute_type),
+				     2*i + use_odd));
+      elts.safe_push (build_int_cst (TREE_TYPE (permute_type),
+				     2*i + use_odd + n_elts));
+    }
+
+  tree permute = elts.build ();
+
+  gimple *g = gimple_build_assign (lhs, VEC_PERM_EXPR, arg0, arg1, permute);
+  gimple_set_location (g, gimple_location (stmt));
+  gsi_replace (gsi, g, true);
+}
+
 /* Fold a machine-dependent built-in in GIMPLE.  (For folding into
    a constant, use rs6000_fold_builtin.)  */
 
@@ -15898,7 +15939,6 @@ rs6000_gimple_fold_builtin (gimple_stmt_iterator *gsi)
     case ALTIVEC_BUILTIN_VSPLTISW:
       {
 	int size;
-
 	if (fn_code == ALTIVEC_BUILTIN_VSPLTISB)
 	  size = 8;
 	else if (fn_code == ALTIVEC_BUILTIN_VSPLTISH)
@@ -15990,6 +16030,21 @@ rs6000_gimple_fold_builtin (gimple_stmt_iterator *gsi)
       fold_mergehl_helper (gsi, stmt, 0);
       return true;
 
+    /* Flavors of vec_mergee.  */
+    case P8V_BUILTIN_VMRGEW_V4SI:
+    case P8V_BUILTIN_VMRGEW_V2DI:
+    case P8V_BUILTIN_VMRGEW_V4SF:
+    case P8V_BUILTIN_VMRGEW_V2DF:
+      fold_mergeeo_helper (gsi, stmt, 0);
+      return true;
+    /* Flavors of vec_mergeo.  */
+    case P8V_BUILTIN_VMRGOW_V4SI:
+    case P8V_BUILTIN_VMRGOW_V2DI:
+    case P8V_BUILTIN_VMRGOW_V4SF:
+    case P8V_BUILTIN_VMRGOW_V2DF:
+      fold_mergeeo_helper (gsi, stmt, 1);
+      return true;
+
     /* d = vec_pack (a, b) */
     case P8V_BUILTIN_VPKUDUM:
     case ALTIVEC_BUILTIN_VPKUHUM: