tree-ssa-loop-unswitch.c: Include "gimple-iterator.h" and "cfghooks.h"...

gcc/

	* tree-ssa-loop-unswitch.c: Include "gimple-iterator.h" and
	"cfghooks.h", add prototypes for introduced new functions.
	(tree_ssa_unswitch_loops): Use from innermost loop iterator, move all
	checks on ability of loop unswitching to tree_unswitch_single_loop;
	invoke tree_unswitch_single_loop or tree_unswitch_outer_loop depending
	on innermost loop check.
	(tree_unswitch_single_loop): Add all required checks on ability of
	loop unswitching under zero recursive level guard.
	(tree_unswitch_outer_loop): New function.
	(find_loop_guard): Likewise.
	(empty_bb_without_guard_p): Likewise.
	(used_outside_loop_p): Likewise.
	(get_vop_from_header): Likewise.
	(hoist_guard): Likewise.
	(check_exit_phi): Likewise.

gcc/testsuite/

	* gcc.dg/loop-unswitch-2.c: New test.
	* gcc.dg/loop-unswitch-3.c: Likewise.
	* gcc.dg/loop-unswitch-4.c: Likewise.

From-SVN: r228599

gcc/ChangeLog

+2015-10-08  Yuri Rumyantsev  <ysrumyan@gmail.com>
+
+	* tree-ssa-loop-unswitch.c: Include "gimple-iterator.h" and
+	"cfghooks.h", add prototypes for introduced new functions.
+	(tree_ssa_unswitch_loops): Use from innermost loop iterator, move all
+	checks on ability of loop unswitching to tree_unswitch_single_loop;
+	invoke tree_unswitch_single_loop or tree_unswitch_outer_loop depending
+	on innermost loop check.
+	(tree_unswitch_single_loop): Add all required checks on ability of
+	loop unswitching under zero recursive level guard.
+	(tree_unswitch_outer_loop): New function.
+	(find_loop_guard): Likewise.
+	(empty_bb_without_guard_p): Likewise.
+	(used_outside_loop_p): Likewise.
+	(get_vop_from_header): Likewise.
+	(hoist_guard): Likewise.
+	(check_exit_phi): Likewise.
+
 2015-10-08  Marek Polacek  <polacek@redhat.com>
 
 	* tree-ssa-reassoc.c (dump_ops_vector): Print newline after each

gcc/testsuite/ChangeLog

+2015-10-08  Yuri Rumyantsev  <ysrumyan@gmail.com>
+
+	* gcc.dg/loop-unswitch-2.c: New test.
+	* gcc.dg/loop-unswitch-3.c: Likewise.
+	* gcc.dg/loop-unswitch-4.c: Likewise.
+
 2015-10-08  Tom de Vries  <tom@codesourcery.com>
 
 	* gcc.dg/dse.c: Only dump in dse1 pass.

gcc/testsuite/gcc.dg/loop-unswitch-2.c

+/* { dg-do compile } */
+/* { dg-options "-O2 -funswitch-loops -fdump-tree-unswitch-details" } */
+
+void foo (float **a, float **b, float *c, int n, int m, int l)
+{
+  int i,j,k;
+  float s;
+  for (i=0; i<l; i++)
+    for (j=0; j<n; j++)
+      for (k=0; k<m; k++)
+	c[i] += a[i][k] * b[k][j];
+}
+
+/* { dg-final { scan-tree-dump-times "guard hoisted" 2 "unswitch" } } */
+

gcc/testsuite/gcc.dg/loop-unswitch-3.c

+/* { dg-do compile } */
+/* { dg-options "-O2 -funswitch-loops -fdump-tree-unswitch-details" } */
+
+#include <stdlib.h>
+#define N 32
+float *foo(int ustride, int size, float *src)
+{
+   float *buffer, *p;
+   int i, k;
+
+   if (!src)
+    return NULL;
+
+   buffer = (float *) malloc(N * size * sizeof(float));
+
+   if(buffer)
+      for(i=0, p=buffer; i<N; i++, src+=ustride)
+	for(k=0; k<size; k++)
+	  *p++ = src[k];
+
+   return buffer;
+}
+
+/* { dg-final { scan-tree-dump-times "guard hoisted" 1 "unswitch" } } */
+
+

gcc/testsuite/gcc.dg/loop-unswitch-4.c

+/* { dg-do run } */
+/* { dg-options "-O2 -funswitch-loops" } */
+
+#include <stdlib.h>
+__attribute__ ((noinline))
+void foo (float **a, float **b, float *c, int n, int m, int l)
+{
+  int i,j,k;
+  float s;
+  for (i=0; i<l; i++)
+    for (j=0; j<n; j++)
+      for (k=0; k<m; k++)
+	c[i] += a[i][k] * b[k][j];
+}
+
+int main()
+{
+  const int N = 32;
+  float **ar1, **ar2;
+  float *res;
+  int i, j;
+  ar1 = (float **)malloc (N * sizeof (float*));
+  ar2 = (float **)malloc (N * sizeof (float*));
+  res = (float *)malloc( N * sizeof (float));
+  for (i=0; i<N; i++)
+    {
+      ar1[i] = (float*)malloc (N * sizeof (float));
+      ar2[i] = (float*)malloc (N * sizeof (float));
+    }
+  for (i=0; i<N; i++)
+    {
+      for (j=0; j<N; j++)
+	{
+	  ar1[i][j] = 2.0f;
+	  ar2[i][j] = 1.5f;
+	}
+      res[i] = 0.0f;
+    }
+  foo (ar1, ar2, res, N, N, N);
+  for (i=0; i<N; i++)
+    if (res[i] != 3072.0f)
+      abort();
+  for (i=0; i<N; i++)
+    res[i] = 0.0f;
+  foo (ar1, ar2, res, N, 0, N);
+  for (i=0; i<N; i++)
+    if (res[i] != 0.0f)
+      abort();
+ 
+  return 0;
+}
+