Add cond_add/sub/mul for vector integer modes.

gcc/ChangeLog:

	* config/i386/sse.md (cond_<insn><mode>): New expander.
	(cond_mul<mode>): Ditto.

gcc/testsuite/ChangeLog:

	* gcc.target/i386/cond_op_addsubmul_d-1.c: New test.
	* gcc.target/i386/cond_op_addsubmul_d-2.c: New test.
	* gcc.target/i386/cond_op_addsubmul_q-1.c: New test.
	* gcc.target/i386/cond_op_addsubmul_q-2.c: New test.
	* gcc.target/i386/cond_op_addsubmul_w-1.c: New test.
	* gcc.target/i386/cond_op_addsubmul_w-2.c: New test.

diff --git a/gcc/config/i386/sse.md b/gcc/config/i386/sse.md
index 8bf1764..52b2b42 100644 (file)
--- a/gcc/config/i386/sse.md
+++ b/gcc/config/i386/sse.md
@@ -333,6 +333,14 @@
   [V16SI (V8SI  "TARGET_AVX512VL") (V4SI  "TARGET_AVX512VL")
    V8DI  (V4DI  "TARGET_AVX512VL") (V2DI  "TARGET_AVX512VL")])
 
+(define_mode_iterator VI1248_AVX512VLBW
+  [(V64QI "TARGET_AVX512BW") (V32QI "TARGET_AVX512VL && TARGET_AVX512BW")
+   (V16QI "TARGET_AVX512VL && TARGET_AVX512BW")
+   (V32HI "TARGET_AVX512BW") (V16HI "TARGET_AVX512VL && TARGET_AVX512BW")
+   (V8HI "TARGET_AVX512VL && TARGET_AVX512BW")
+   V16SI (V8SI "TARGET_AVX512VL") (V4SI "TARGET_AVX512VL")
+   V8DI (V4DI "TARGET_AVX512VL") (V2DI "TARGET_AVX512VL")])
+
 (define_mode_iterator VF_AVX512VL
   [V16SF (V8SF "TARGET_AVX512VL") (V4SF "TARGET_AVX512VL")
    V8DF (V4DF "TARGET_AVX512VL") (V2DF "TARGET_AVX512VL")])
@@ -11803,6 +11811,24 @@
   "TARGET_SSE2"
   "ix86_fixup_binary_operands_no_copy (<CODE>, <MODE>mode, operands);")
 
+(define_expand "cond_<insn><mode>"
+  [(set (match_operand:VI1248_AVX512VLBW 0 "register_operand")
+       (vec_merge:VI1248_AVX512VLBW
+         (plusminus:VI1248_AVX512VLBW
+           (match_operand:VI1248_AVX512VLBW 2 "nonimmediate_operand")
+           (match_operand:VI1248_AVX512VLBW 3 "nonimmediate_operand"))
+         (match_operand:VI1248_AVX512VLBW 4 "nonimm_or_0_operand")
+         (match_operand:<avx512fmaskmode> 1 "register_operand")))]
+  "TARGET_AVX512F"
+{
+  emit_insn (gen_<insn><mode>3_mask (operands[0],
+                                    operands[2],
+                                    operands[3],
+                                    operands[4],
+                                    operands[1]));
+  DONE;
+})
+
 (define_expand "<insn><mode>3_mask"
   [(set (match_operand:VI48_AVX512VL 0 "register_operand")
        (vec_merge:VI48_AVX512VL
@@ -11929,6 +11955,24 @@
   DONE;
 })
 
+(define_expand "cond_mul<mode>"
+  [(set (match_operand:VI2_AVX512VL 0 "register_operand")
+       (vec_merge:VI2_AVX512VL
+         (mult:VI2_AVX512VL
+           (match_operand:VI2_AVX512VL 2 "vector_operand")
+           (match_operand:VI2_AVX512VL 3 "vector_operand"))
+         (match_operand:VI2_AVX512VL 4 "nonimm_or_0_operand")
+         (match_operand:<avx512fmaskmode> 1 "register_operand")))]
+  "TARGET_AVX512BW"
+{
+  emit_insn (gen_mul<mode>3_mask (operands[0],
+                                 operands[2],
+                                 operands[3],
+                                 operands[4],
+                                 operands[1]));
+  DONE;
+})
+
 (define_expand "mul<mode>3<mask_name>"
   [(set (match_operand:VI2_AVX2 0 "register_operand")
        (mult:VI2_AVX2 (match_operand:VI2_AVX2 1 "vector_operand")
@@ -12363,6 +12407,24 @@
    (set_attr "prefix" "orig,vex")
    (set_attr "mode" "TI")])
 
+(define_expand "cond_mul<mode>"
+  [(set (match_operand:VI8_AVX512VL 0 "register_operand")
+       (vec_merge:VI8_AVX512VL
+         (mult:VI8_AVX512VL
+           (match_operand:VI8_AVX512VL 2 "vector_operand")
+           (match_operand:VI8_AVX512VL 3 "vector_operand"))
+         (match_operand:VI8_AVX512VL 4 "nonimm_or_0_operand")
+         (match_operand:<avx512fmaskmode> 1 "register_operand")))]
+  "TARGET_AVX512DQ"
+{
+  emit_insn (gen_avx512dq_mul<mode>3_mask (operands[0],
+                                          operands[2],
+                                          operands[3],
+                                          operands[4],
+                                          operands[1]));
+  DONE;
+})
+
 (define_insn "avx512dq_mul<mode>3<mask_name>"
   [(set (match_operand:VI8_AVX512VL 0 "register_operand" "=v")
        (mult:VI8_AVX512VL
@@ -12375,6 +12437,24 @@
    (set_attr "prefix" "evex")
    (set_attr "mode" "<sseinsnmode>")])
 
+(define_expand "cond_mul<mode>"
+  [(set (match_operand:VI4_AVX512VL 0 "register_operand")
+       (vec_merge:VI4_AVX512VL
+         (mult:VI4_AVX512VL
+           (match_operand:VI4_AVX512VL 2 "vector_operand")
+           (match_operand:VI4_AVX512VL 3 "vector_operand"))
+         (match_operand:VI4_AVX512VL 4 "nonimm_or_0_operand")
+         (match_operand:<avx512fmaskmode> 1 "register_operand")))]
+  "TARGET_AVX512F"
+{
+  emit_insn (gen_mul<mode>3_mask (operands[0],
+                                 operands[2],
+                                 operands[3],
+                                 operands[4],
+                                 operands[1]));
+  DONE;
+})
+
 (define_expand "mul<mode>3<mask_name>"
   [(set (match_operand:VI4_AVX512F 0 "register_operand")
        (mult:VI4_AVX512F
@@ -14043,14 +14123,6 @@
              ]
              (const_string "<sseinsnmode>")))])
 
-(define_mode_iterator VI1248_AVX512VLBW
-  [(V64QI "TARGET_AVX512BW") (V32QI "TARGET_AVX512VL && TARGET_AVX512BW")
-   (V16QI "TARGET_AVX512VL && TARGET_AVX512BW")
-   (V32HI "TARGET_AVX512BW") (V16HI "TARGET_AVX512VL && TARGET_AVX512BW")
-   (V8HI "TARGET_AVX512VL && TARGET_AVX512BW")
-   V16SI (V8SI "TARGET_AVX512VL") (V4SI "TARGET_AVX512VL")
-   V8DI (V4DI "TARGET_AVX512VL") (V2DI "TARGET_AVX512VL")])
-
 (define_mode_iterator AVX512ZEXTMASK
   [(DI "TARGET_AVX512BW") (SI "TARGET_AVX512BW") HI])
 

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-1.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-1.c
new file mode 100644 (file)
index 0000000..c506fa5
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-1.c
@@ -0,0 +1,32 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -march=skylake-avx512 -fdump-tree-optimized" } */
+/* { dg-final { scan-tree-dump ".COND_ADD" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_SUB" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_MUL" "optimized" } } */
+#ifndef NUM
+#define NUM 800
+#endif
+#ifndef TYPE
+#define TYPE int
+#endif
+
+TYPE a[NUM], b[NUM], c[NUM], d[NUM], e[NUM], j[NUM];
+#define MIN(X,Y) ((X) < (Y) ? (X) : (Y))
+#define MAX(X,Y) ((X) < (Y) ? (Y) : (X))
+
+#define BIN(OPNAME, OP)                                \
+  void                                         \
+  __attribute__ ((noipa,optimize ("O3")))      \
+  foo_##OPNAME ()                              \
+  {                                            \
+    for (int i = 0; i != NUM; i++)             \
+      if (b[i] < c[i])                         \
+       a[i] = d[i] OP e[i];                    \
+      else                                     \
+       a[i] = MAX(d[i], e[i]);                 \
+  }
+
+
+BIN (add, +);
+BIN (sub, -);
+BIN (mul, *);

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-2.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-2.c
new file mode 100644 (file)
index 0000000..490f4af
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_d-2.c
@@ -0,0 +1,76 @@
+/* { dg-do run } */
+/* { dg-options "-O2 -mavx512vl -mprefer-vector-width=256" } */
+#define AVX512VL
+#ifndef CHECK
+#define CHECK "avx512f-helper.h"
+#endif
+
+#include CHECK
+
+#include "cond_op_addsubmul_d-1.c"
+#define BINO2(OPNAME, OP)                      \
+  void                                         \
+  __attribute__ ((noipa,optimize ("O2")))      \
+  foo_o2_##OPNAME ()                           \
+  {                                            \
+    for (int i = 0; i != NUM; i++)             \
+      if (b[i] < c[i])                         \
+       j[i] = d[i] OP e[i];                    \
+      else                                     \
+       j[i] = MAX(d[i], e[i]);                 \
+  }
+
+BINO2 (add, +);
+BINO2 (sub, -);
+BINO2 (mul, *);
+
+static void
+test_256 (void)
+{
+  int sign = -1;
+  for (int i = 0; i != NUM; i++)
+    {
+      a[i] = 0;
+      d[i] = i * 2;
+      e[i] = i * i * 3 - i * 9 + 153;
+      b[i] = i * 83;
+      c[i] = b[i] + sign;
+      sign *= -1;
+      j[i] = 1;
+    }
+  foo_add ();
+  foo_o2_add ();
+  for (int i = 0; i != NUM; i++)
+    {
+      if (a[i] != j[i])
+       abort ();
+      a[i] = 0;
+      b[i] = 1;
+    }
+
+  foo_sub ();
+  foo_o2_sub ();
+  for (int i = 0; i != NUM; i++)
+    {
+      if (a[i] != j[i])
+       abort ();
+      a[i] = 0;
+      j[i] = 1;
+    }
+
+  foo_mul ();
+  foo_o2_mul ();
+  for (int i = 0; i != NUM; i++)
+    {
+      if (a[i] != j[i])
+       abort ();
+      a[i] = 0;
+      j[i] = 1;
+    }
+}
+
+static void
+test_128 ()
+{
+  
+}

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-1.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-1.c
new file mode 100644 (file)
index 0000000..5e25350
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-1.c
@@ -0,0 +1,7 @@
+/* { dg-do compile { target { ! ia32 } } } */
+/* { dg-options "-O2 -march=skylake-avx512 -DTYPE=long -fdump-tree-optimized" } */
+/* { dg-final { scan-tree-dump ".COND_ADD" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_SUB" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_MUL" "optimized" } } */
+#define AVX512DQ
+#include "cond_op_addsubmul_d-1.c"

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-2.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-2.c
new file mode 100644 (file)
index 0000000..09a87de
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_q-2.c
@@ -0,0 +1,4 @@
+/* { dg-do run { target { ! ia32 } } } */
+/* { dg-options "-O2 -mavx512vl -mprefer-vector-width=256 -mavx512dq -DTYPE=long" } */
+
+#include "cond_op_addsubmul_d-2.c"

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-1.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-1.c
new file mode 100644 (file)
index 0000000..80d78d9
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-1.c
@@ -0,0 +1,6 @@
+/* { dg-do compile } */
+/* { dg-options "-O2 -march=skylake-avx512 -DTYPE=short -fdump-tree-optimized" } */
+/* { dg-final { scan-tree-dump ".COND_ADD" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_SUB" "optimized" } } */
+/* { dg-final { scan-tree-dump ".COND_MUL" "optimized" } } */
+#include "cond_op_addsubmul_d-1.c"

diff --git a/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-2.c b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-2.c
new file mode 100644 (file)
index 0000000..fdcdb34
--- /dev/null
+++ b/gcc/testsuite/gcc.target/i386/cond_op_addsubmul_w-2.c
@@ -0,0 +1,5 @@
+/* { dg-do run } */
+/* { dg-options "-O2 -mavx512vl -mprefer-vector-width=256 -mavx512bw -DTYPE=short" } */
+
+#define AVX512BW
+#include "cond_op_addsubmul_d-2.c"