re PR tree-optimization/43280 (gcc4.5 -m32 -O2: misoptimizes sha256!)

2010-03-11  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>

	PR tree-optimization/43280
	* tree-ssa-math-opts.c (find_bswap_1): Modify symbolic number generation.
	Move calculation of size out of the if branch.
	(find_bswap): Modify compare number generation.

2010-03-11  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>

	* gcc.dg/optimize-bswapdi-1.c: Add OpenSSL bswap variant.
	* gcc.dg/pr43280.c: New testcase.

From-SVN: r157386

5 files changed, 68 insertions, 10 deletions

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index f10f808..5e07486 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,10 @@
+2010-03-11  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>
+
+	PR tree-optimization/43280
+	* tree-ssa-math-opts.c (find_bswap_1): Modify symbolic number generation.
+	Move calculation of size out of the if branch.
+	(find_bswap): Modify compare number generation.
+
 2010-03-11  Richard Guenther  <rguenther@suse.de>
 
 	PR lto/43200
diff --git a/gcc/testsuite/ChangeLog b/gcc/testsuite/ChangeLog
index bf7d44b..fc9f8e6 100644
--- a/gcc/testsuite/ChangeLog
+++ b/gcc/testsuite/ChangeLog
@@ -1,3 +1,8 @@
+2010-03-11  Andreas Krebbel  <Andreas.Krebbel@de.ibm.com>
+
+	* gcc.dg/optimize-bswapdi-1.c: Add OpenSSL bswap variant.
+	* gcc.dg/pr43280.c: New testcase.
+
 2010-03-11  Richard Guenther  <rguenther@suse.de>
 
 	PR lto/43200
diff --git a/gcc/testsuite/gcc.dg/optimize-bswapdi-1.c b/gcc/testsuite/gcc.dg/optimize-bswapdi-1.c
index 315ea6c..7d557f3 100644
--- a/gcc/testsuite/gcc.dg/optimize-bswapdi-1.c
+++ b/gcc/testsuite/gcc.dg/optimize-bswapdi-1.c
@@ -43,6 +43,19 @@ swap64_b (DItype u)
 	  | (((u) & 0x00000000000000ffull) << 56));
 }
 
+/* The OpenSSL variant.  */
 
-/* { dg-final { scan-tree-dump-times "64 bit bswap implementation found at" 2 "bswap" } } */
+uint64_t
+swap64_c (uint64_t x)
+{
+  uint32_t a = x >> 32;
+  uint32_t b = (uint32_t) x;
+  return ((uint64_t) ((((((b)) >> (8)) | (((b)) << (32 - (8)))) & 0xff00ff00L)
+		      | (((((b)) << (8)) | (((b)) >> (32 - (8)))) & 0x00ff00ffL)) << 32)
+          | (uint64_t) ((((((a)) >> (8)) | (((a)) << (32 - (8)))) & 0xff00ff00L)
+			| (((((a)) << (8)) | (((a)) >> (32 - (8)))) & 0x00ff00ffL));
+}
+
+
+/* { dg-final { scan-tree-dump-times "64 bit bswap implementation found at" 3 "bswap" } } */
 /* { dg-final { cleanup-tree-dump "bswap" } } */
diff --git a/gcc/testsuite/gcc.dg/pr43280.c b/gcc/testsuite/gcc.dg/pr43280.c
new file mode 100644
index 0000000..dfb369b
--- /dev/null
+++ b/gcc/testsuite/gcc.dg/pr43280.c
@@ -0,0 +1,30 @@
+/* { dg-do run } */
+/* { dg-require-effective-target stdint_types } */
+/* { dg-options "-O2" } */
+
+#include <stdint.h>
+
+extern void abort (void);
+
+uint64_t __attribute__((noinline))
+byteswap64(uint64_t x)
+{
+  uint32_t a = x >> 32;
+  uint32_t b = (uint32_t) x;
+  return ((uint64_t) ((((((b)) >> (8)) | (((b)) << (32 - (8)))) & 0xff00ff00L)
+		      | (((((b)) << (8)) | (((b)) >> (32 - (8)))) & 0x00ff00ffL)) << 32)
+          | (uint64_t) ((((((a)) >> (8)) | (((a)) << (32 - (8)))) & 0xff00ff00L)
+			| (((((a)) << (8)) | (((a)) >> (32 - (8)))) & 0x00ff00ffL));
+}
+
+int
+main ()
+{
+  uint64_t in = (uint64_t)0x01020304 << 32 | 0x05060708;
+  uint64_t cmp = (uint64_t)0x08070605 << 32 | 0x04030201;
+
+  if (cmp != byteswap64 (in))
+    abort ();
+
+  return 0;
+}
diff --git a/gcc/tree-ssa-math-opts.c b/gcc/tree-ssa-math-opts.c
index c46a57f..d38df8d 100644
--- a/gcc/tree-ssa-math-opts.c
+++ b/gcc/tree-ssa-math-opts.c
@@ -940,15 +940,18 @@ find_bswap_1 (gimple stmt, struct symbolic_number *n, int limit)
 	{
 	  /* Set up the symbolic number N by setting each byte to a
 	     value between 1 and the byte size of rhs1.  The highest
-	     order byte is set to 1 and the lowest order byte to
-	     n.size.  */
+	     order byte is set to n->size and the lowest order
+	     byte to 1.  */
 	  n->size = TYPE_PRECISION (TREE_TYPE (rhs1));
 	  if (n->size % BITS_PER_UNIT != 0)
 	    return NULL_TREE;
 	  n->size /= BITS_PER_UNIT;
 	  n->n = (sizeof (HOST_WIDEST_INT) < 8 ? 0 :
-		  (unsigned HOST_WIDEST_INT)0x01020304 << 32 | 0x05060708);
-	  n->n >>= (sizeof (HOST_WIDEST_INT) - n->size) * BITS_PER_UNIT;
+		  (unsigned HOST_WIDEST_INT)0x08070605 << 32 | 0x04030201);
+
+	  if (n->size < (int)sizeof (HOST_WIDEST_INT))
+	    n->n &= ((unsigned HOST_WIDEST_INT)1 <<
+		     (n->size * BITS_PER_UNIT)) - 1;
 
 	  source_expr1 = rhs1;
 	}
@@ -988,9 +991,9 @@ find_bswap_1 (gimple stmt, struct symbolic_number *n, int limit)
 	      {
 		/* If STMT casts to a smaller type mask out the bits not
 		   belonging to the target type.  */
-		n->size = type_size / BITS_PER_UNIT;
 		n->n &= ((unsigned HOST_WIDEST_INT)1 << type_size) - 1;
 	      }
+	    n->size = type_size / BITS_PER_UNIT;
 	  }
 	  break;
 	default:
@@ -1051,11 +1054,11 @@ static tree
 find_bswap (gimple stmt)
 {
 /* The number which the find_bswap result should match in order to
-   have a full byte swap.  The insignificant bytes are masked out
-   before using it.  */
+   have a full byte swap.  The number is shifted to the left according
+   to the size of the symbolic number before using it.  */
   unsigned HOST_WIDEST_INT cmp =
     sizeof (HOST_WIDEST_INT) < 8 ? 0 :
-    (unsigned HOST_WIDEST_INT)0x08070605 << 32 | 0x04030201;
+    (unsigned HOST_WIDEST_INT)0x01020304 << 32 | 0x05060708;
 
   struct symbolic_number n;
   tree source_expr;
@@ -1079,7 +1082,7 @@ find_bswap (gimple stmt)
 	((unsigned HOST_WIDEST_INT)1 << (n.size * BITS_PER_UNIT)) - 1;
 
       n.n &= mask;
-      cmp &= mask;
+      cmp >>= (sizeof (HOST_WIDEST_INT) - n.size) * BITS_PER_UNIT;
     }
 
   /* A complete byte swap should make the symbolic number to start