predicates.md (general_vector_operand): New predicate.

	* config/i386/predicates.md (general_vector_operand): New predicate.
	* config/i386/i386.c (const_vector_equal_evenodd_p): New function.
	(ix86_expand_mul_widen_evenodd): Force op1 resp. op2 into register
	if they aren't nonimmediate operands.  If their original values
	satisfy const_vector_equal_evenodd_p, don't shift them.
	* config/i386/sse.md (mul<mode>3): Use general_vector_operand
	predicates.  For the SSE4.1 case force operands[{1,2}] into registers
	if not nonimmediate_operand.
	(vec_widen_smult_even_v4si): Use nonimmediate_operand predicates
	instead of register_operand.
	(vec_widen_<s>mult_odd_<mode>): Use general_vector_operand predicates.

From-SVN: r198387

4 files changed, 59 insertions, 12 deletions

diff --git a/gcc/ChangeLog b/gcc/ChangeLog
index 7a15745..85dfc35 100644
--- a/gcc/ChangeLog
+++ b/gcc/ChangeLog
@@ -1,3 +1,17 @@
+2013-04-29  Jakub Jelinek  <jakub@redhat.com>
+
+	* config/i386/predicates.md (general_vector_operand): New predicate.
+	* config/i386/i386.c (const_vector_equal_evenodd_p): New function.
+	(ix86_expand_mul_widen_evenodd): Force op1 resp. op2 into register
+	if they aren't nonimmediate operands.  If their original values
+	satisfy const_vector_equal_evenodd_p, don't shift them.
+	* config/i386/sse.md (mul<mode>3): Use general_vector_operand
+	predicates.  For the SSE4.1 case force operands[{1,2}] into registers
+	if not nonimmediate_operand.
+	(vec_widen_smult_even_v4si): Use nonimmediate_operand predicates
+	instead of register_operand.
+	(vec_widen_<s>mult_odd_<mode>): Use general_vector_operand predicates.
+
 2013-04-28  Eric Botcazou  <ebotcazou@gcc.gnu.org>
 
 	* stor-layout.c (finalize_size_functions): Allocate a structure and
diff --git a/gcc/config/i386/i386.c b/gcc/config/i386/i386.c
index 03191a6..06aeecf 100644
--- a/gcc/config/i386/i386.c
+++ b/gcc/config/i386/i386.c
@@ -40827,6 +40827,24 @@ ix86_expand_vecop_qihi (enum rtx_code code, rtx dest, rtx op1, rtx op2)
 		       gen_rtx_fmt_ee (code, qimode, op1, op2));
 }
 
+/* Helper function of ix86_expand_mul_widen_evenodd.  Return true
+   if op is CONST_VECTOR with all odd elements equal to their
+   preceeding element.  */
+
+static bool
+const_vector_equal_evenodd_p (rtx op)
+{
+  enum machine_mode mode = GET_MODE (op);
+  int i, nunits = GET_MODE_NUNITS (mode);
+  if (GET_CODE (op) != CONST_VECTOR
+      || nunits != CONST_VECTOR_NUNITS (op))
+    return false;
+  for (i = 0; i < nunits; i += 2)
+    if (CONST_VECTOR_ELT (op, i) != CONST_VECTOR_ELT (op, i + 1))
+      return false;
+  return true;
+}
+
 void
 ix86_expand_mul_widen_evenodd (rtx dest, rtx op1, rtx op2,
 			       bool uns_p, bool odd_p)
@@ -40834,6 +40852,12 @@ ix86_expand_mul_widen_evenodd (rtx dest, rtx op1, rtx op2,
   enum machine_mode mode = GET_MODE (op1);
   enum machine_mode wmode = GET_MODE (dest);
   rtx x;
+  rtx orig_op1 = op1, orig_op2 = op2;
+
+  if (!nonimmediate_operand (op1, mode))
+    op1 = force_reg (mode, op1);
+  if (!nonimmediate_operand (op2, mode))
+    op2 = force_reg (mode, op2);
 
   /* We only play even/odd games with vectors of SImode.  */
   gcc_assert (mode == V4SImode || mode == V8SImode);
@@ -40852,10 +40876,12 @@ ix86_expand_mul_widen_evenodd (rtx dest, rtx op1, rtx op2,
 	}
 
       x = GEN_INT (GET_MODE_UNIT_BITSIZE (mode));
-      op1 = expand_binop (wmode, lshr_optab, gen_lowpart (wmode, op1),
-			  x, NULL, 1, OPTAB_DIRECT);
-      op2 = expand_binop (wmode, lshr_optab, gen_lowpart (wmode, op2),
-			  x, NULL, 1, OPTAB_DIRECT);
+      if (!const_vector_equal_evenodd_p (orig_op1))
+	op1 = expand_binop (wmode, lshr_optab, gen_lowpart (wmode, op1),
+			    x, NULL, 1, OPTAB_DIRECT);
+      if (!const_vector_equal_evenodd_p (orig_op2))
+	op2 = expand_binop (wmode, lshr_optab, gen_lowpart (wmode, op2),
+			    x, NULL, 1, OPTAB_DIRECT);
       op1 = gen_lowpart (mode, op1);
       op2 = gen_lowpart (mode, op2);
     }
diff --git a/gcc/config/i386/predicates.md b/gcc/config/i386/predicates.md
index baa0913..90ab8da 100644
--- a/gcc/config/i386/predicates.md
+++ b/gcc/config/i386/predicates.md
@@ -1303,3 +1303,8 @@
   HOST_WIDE_INT low = val & 0xff;
   return val == ((low << 8) | low);
 })
+
+;; Return true if OP is nonimmediate_operand or CONST_VECTOR.
+(define_predicate "general_vector_operand"
+  (ior (match_operand 0 "nonimmediate_operand")
+       (match_code "const_vector")))
diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md
index 4597385..f630303e 100644
--- a/gcc/config/i386/sse.md
+++ b/gcc/config/i386/sse.md
@@ -5631,14 +5631,16 @@
 (define_expand "mul<mode>3"
   [(set (match_operand:VI4_AVX2 0 "register_operand")
 	(mult:VI4_AVX2
-	  (match_operand:VI4_AVX2 1 "nonimmediate_operand")
-	  (match_operand:VI4_AVX2 2 "nonimmediate_operand")))]
+	  (match_operand:VI4_AVX2 1 "general_vector_operand")
+	  (match_operand:VI4_AVX2 2 "general_vector_operand")))]
   "TARGET_SSE2"
 {
   if (TARGET_SSE4_1)
     {
-      if (CONSTANT_P (operands[2]))
-	operands[2] = force_const_mem (<MODE>mode, operands[2]);
+      if (!nonimmediate_operand (operands[1], <MODE>mode))
+	operands[1] = force_reg (<MODE>mode, operands[1]);
+      if (!nonimmediate_operand (operands[2], <MODE>mode))
+	operands[2] = force_reg (<MODE>mode, operands[2]);
       ix86_fixup_binary_operands_no_copy (MULT, <MODE>mode, operands);
     }
   else
@@ -5702,8 +5704,8 @@
 ;; named patterns, but signed V4SI needs special help for plain SSE2.
 (define_expand "vec_widen_smult_even_v4si"
   [(match_operand:V2DI 0 "register_operand")
-   (match_operand:V4SI 1 "register_operand")
-   (match_operand:V4SI 2 "register_operand")]
+   (match_operand:V4SI 1 "nonimmediate_operand")
+   (match_operand:V4SI 2 "nonimmediate_operand")]
   "TARGET_SSE2"
 {
   ix86_expand_mul_widen_evenodd (operands[0], operands[1], operands[2],
@@ -5714,8 +5716,8 @@
 (define_expand "vec_widen_<s>mult_odd_<mode>"
   [(match_operand:<sseunpackmode> 0 "register_operand")
    (any_extend:<sseunpackmode>
-     (match_operand:VI4_AVX2 1 "register_operand"))
-   (match_operand:VI4_AVX2 2 "register_operand")]
+     (match_operand:VI4_AVX2 1 "general_vector_operand"))
+   (match_operand:VI4_AVX2 2 "general_vector_operand")]
   "TARGET_SSE2"
 {
   ix86_expand_mul_widen_evenodd (operands[0], operands[1], operands[2],