builtins.c (expand_builtin_acc_on_device): Delete.

	gcc/
	* builtins.c (expand_builtin_acc_on_device): Delete.
	(expand_builtin): Don't call it.
	(fold_builtin_1): Fold acc_on_device.

	libgomp/
	* oacc-init.c (acc_on_device): Force optimization level.

From-SVN: r228267

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 9805eff3e0099237ed0a713482e6790f7ec8bdf3..7e34d4cfc161d3a8157a41a03f9be6c9f67d1c3d 100644 (file)
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,9 @@
+2015-09-29  Nathan Sidwell  <nathan@codesourcery.com>
+
+       * builtins.c (expand_builtin_acc_on_device): Delete.
+       (expand_builtin): Don't call it.
+       (fold_builtin_1): Fold acc_on_device.
+
 2015-09-29  H.J. Lu  <hongjiu.lu@intel.com>
 
        * config/i386/i386.c (ix86_function_arg): Fix typo in comments.

diff --git a/gcc/builtins.c b/gcc/builtins.c
index aeec1708759c4737542ba3c0d1d409a8881f9988..15928109247a97534b74d0c9478c89381cf4ad1d 100644 (file)
--- a/gcc/builtins.c
+++ b/gcc/builtins.c
@@ -5859,46 +5859,6 @@ expand_stack_save (void)
 }
 
 
-/* Expand OpenACC acc_on_device.
-
-   This has to happen late (that is, not in early folding; expand_builtin_*,
-   rather than fold_builtin_*), as we have to act differently for host and
-   acceleration device (ACCEL_COMPILER conditional).  */
-
-static rtx
-expand_builtin_acc_on_device (tree exp, rtx target)
-{
-  if (!validate_arglist (exp, INTEGER_TYPE, VOID_TYPE))
-    return NULL_RTX;
-
-  tree arg = CALL_EXPR_ARG (exp, 0);
-
-  /* Return (arg == v1 || arg == v2) ? 1 : 0.  */
-  machine_mode v_mode = TYPE_MODE (TREE_TYPE (arg));
-  rtx v = expand_normal (arg), v1, v2;
-#ifdef ACCEL_COMPILER
-  v1 = GEN_INT (GOMP_DEVICE_NOT_HOST);
-  v2 = GEN_INT (ACCEL_COMPILER_acc_device);
-#else
-  v1 = GEN_INT (GOMP_DEVICE_NONE);
-  v2 = GEN_INT (GOMP_DEVICE_HOST);
-#endif
-  machine_mode target_mode = TYPE_MODE (integer_type_node);
-  if (!target || !register_operand (target, target_mode))
-    target = gen_reg_rtx (target_mode);
-  emit_move_insn (target, const1_rtx);
-  rtx_code_label *done_label = gen_label_rtx ();
-  do_compare_rtx_and_jump (v, v1, EQ, false, v_mode, NULL_RTX,
-                          NULL, done_label, PROB_EVEN);
-  do_compare_rtx_and_jump (v, v2, EQ, false, v_mode, NULL_RTX,
-                          NULL, done_label, PROB_EVEN);
-  emit_move_insn (target, const0_rtx);
-  emit_label (done_label);
-
-  return target;
-}
-
-
 /* Expand an expression EXP that calls a built-in function,
    with result going to TARGET if that's convenient
    (and in mode MODE if that's convenient).
@@ -7036,9 +6996,8 @@ expand_builtin (tree exp, rtx target, rtx subtarget, machine_mode mode,
       break;
 
     case BUILT_IN_ACC_ON_DEVICE:
-      target = expand_builtin_acc_on_device (exp, target);
-      if (target)
-       return target;
+      /* Do library call, if we failed to expand the builtin when
+        folding.  */
       break;
 
     default:   /* just do library call, if unknown builtin */
@@ -10271,6 +10230,27 @@ fold_builtin_1 (location_t loc, tree fndecl, tree arg0)
        return build_empty_stmt (loc);
       break;
 
+    case BUILT_IN_ACC_ON_DEVICE:
+      /* Don't fold on_device until we know which compiler is active.  */
+      if (symtab->state == EXPANSION)
+       {
+         unsigned val_host = GOMP_DEVICE_HOST;
+         unsigned val_dev = GOMP_DEVICE_NONE;
+
+#ifdef ACCEL_COMPILER
+         val_host = GOMP_DEVICE_NOT_HOST;
+         val_dev = ACCEL_COMPILER_acc_device;
+#endif
+         tree host = build2 (EQ_EXPR, boolean_type_node, arg0,
+                             build_int_cst (integer_type_node, val_host));
+         tree dev = build2 (EQ_EXPR, boolean_type_node, arg0,
+                            build_int_cst (integer_type_node, val_dev));
+
+         tree result = build2 (TRUTH_OR_EXPR, boolean_type_node, host, dev);
+         return fold_convert (integer_type_node, result);
+       }
+      break;
+
     default:
       break;
     }

diff --git a/libgomp/ChangeLog b/libgomp/ChangeLog
index 88a25fcb8bf915878bb38f2b552640e4855a9792..8fbd6c59dee1a0a7e10fce3146e778b93cc62e52 100644 (file)
--- a/libgomp/ChangeLog
+++ b/libgomp/ChangeLog
@@ -1,3 +1,7 @@
+2015-09-29  Nathan Sidwell  <nathan@codesourcery.com>
+
+       * oacc-init.c (acc_on_device): Force optimization level.
+
 2015-09-29  Nathan Sidwell  <nathan@codesourcery.com>
 
        * plugin/plugin-nvptx.c (ARRAYSIZE): Delete.

diff --git a/libgomp/oacc-init.c b/libgomp/oacc-init.c
index 2258a72bbe58fafc97e1023a37901aba73daadc9..28b9e7a90f1b8a7d838bac684ba3c6b9340299d5 100644 (file)
--- a/libgomp/oacc-init.c
+++ b/libgomp/oacc-init.c
@@ -620,10 +620,12 @@ acc_set_device_num (int ord, acc_device_t d)
 
 ialias (acc_set_device_num)
 
-int
+/* Compile on_device with optimization, so that the compiler expands
+   this, rather than generating infinitely recursive code.  */
+
+int __attribute__ ((__optimize__ ("O2")))
 acc_on_device (acc_device_t dev)
 {
-  /* Just rely on the compiler builtin.  */
   return __builtin_acc_on_device (dev);
 }