public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [PATCH V2] RISC-V: Support in-order floating-point reduction
@ 2023-07-20  8:51 Juzhe-Zhong
  2023-07-20  8:59 ` Kito Cheng
  0 siblings, 1 reply; 4+ messages in thread
From: Juzhe-Zhong @ 2023-07-20  8:51 UTC (permalink / raw)
  To: gcc-patches; +Cc: kito.cheng, kito.cheng, jeffreyalaw, rdapp.gcc, Juzhe-Zhong

This patch is depending on:
https://gcc.gnu.org/pipermail/gcc-patches/2023-July/624995.html

Consider this following case:
float foo (float *__restrict a, int n)
{
  float result = 1.0;
  for (int i = 0; i < n; i++)
   result += a[i];
  return result;
}

Compile with **NO** -ffast-math:

Before this patch:
<source>:4:21: missed: couldn't vectorize loop
<source>:1:7: missed: not vectorized: relevant phi not supported: result_14 = PHI <result_11(6), 1.0e+0(5)>

After this patch:
foo:
	lui	a5,%hi(.LC0)
	flw	fa0,%lo(.LC0)(a5)
	ble	a1,zero,.L4
.L3:
	vsetvli	a5,a1,e32,m1,ta,ma
	vle32.v	v1,0(a0)
	slli	a4,a5,2
	sub	a1,a1,a5
	vfmv.s.f	v2,fa0
	add	a0,a0,a4
	vfredosum.vs	v1,v1,v2     ----------> FOLD_LEFT_PLUS
	vfmv.f.s	fa0,v1
	bne	a1,zero,.L3
	ret
.L4:
	ret

gcc/ChangeLog:

	* config/riscv/autovec.md (fold_left_plus_<mode>): New pattern.
	(mask_len_fold_left_plus_<mode>): Ditto.
	* config/riscv/riscv-protos.h (enum insn_type): New enum.
	(enum reduction_type): Ditto.
	(expand_reduction): Add in-order reduction.
	* config/riscv/riscv-v.cc (emit_nonvlmax_fp_reduction_insn): New function.
	(expand_reduction): Add in-order reduction.

gcc/testsuite/ChangeLog:

	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c: New test.
	* gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c: New test.

---
 gcc/config/riscv/autovec.md                   | 39 ++++++++++++++
 gcc/config/riscv/riscv-protos.h               | 13 ++++-
 gcc/config/riscv/riscv-v.cc                   | 53 +++++++++++++++----
 .../riscv/rvv/autovec/reduc/reduc_strict-1.c  | 28 ++++++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-2.c  | 26 +++++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-3.c  | 18 +++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-4.c  | 24 +++++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-5.c  | 28 ++++++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-6.c  | 18 +++++++
 .../riscv/rvv/autovec/reduc/reduc_strict-7.c  | 21 ++++++++
 .../rvv/autovec/reduc/reduc_strict_run-1.c    | 29 ++++++++++
 .../rvv/autovec/reduc/reduc_strict_run-2.c    | 31 +++++++++++
 12 files changed, 317 insertions(+), 11 deletions(-)
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c

diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
index 00947207f3f..667a877d009 100644
--- a/gcc/config/riscv/autovec.md
+++ b/gcc/config/riscv/autovec.md
@@ -1687,3 +1687,42 @@
   riscv_vector::expand_reduction (SMIN, operands, f);
   DONE;
 })
+
+;; -------------------------------------------------------------------------
+;; ---- [FP] Left-to-right reductions
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - vfredosum.vs
+;; -------------------------------------------------------------------------
+
+;; Unpredicated in-order FP reductions.
+(define_expand "fold_left_plus_<mode>"
+  [(match_operand:<VEL> 0 "register_operand")
+   (match_operand:<VEL> 1 "register_operand")
+   (match_operand:VF 2 "register_operand")]
+  "TARGET_VECTOR"
+{
+  riscv_vector::expand_reduction (PLUS, operands,
+				  operands[1],
+				  riscv_vector::reduction_type::FOLD_LEFT);
+  DONE;
+})
+
+;; Predicated in-order FP reductions.
+(define_expand "mask_len_fold_left_plus_<mode>"
+  [(match_operand:<VEL> 0 "register_operand")
+   (match_operand:<VEL> 1 "register_operand")
+   (match_operand:VF 2 "register_operand")
+   (match_operand:<VM> 3 "vector_mask_operand")
+   (match_operand 4 "autovec_length_operand")
+   (match_operand 5 "const_0_operand")]
+  "TARGET_VECTOR"
+{
+  if (rtx_equal_p (operands[4], const0_rtx))
+    emit_move_insn (operands[0], operands[1]);
+  else
+    riscv_vector::expand_reduction (PLUS, operands,
+				    operands[1],
+				    riscv_vector::reduction_type::MASK_LEN_FOLD_LEFT);
+  DONE;
+})
diff --git a/gcc/config/riscv/riscv-protos.h b/gcc/config/riscv/riscv-protos.h
index 16fb8dabca0..c9520f689e2 100644
--- a/gcc/config/riscv/riscv-protos.h
+++ b/gcc/config/riscv/riscv-protos.h
@@ -199,6 +199,7 @@ enum insn_type
   RVV_GATHER_M_OP = 5,
   RVV_SCATTER_M_OP = 4,
   RVV_REDUCTION_OP = 3,
+  RVV_REDUCTION_TU_OP = RVV_REDUCTION_OP + 2,
 };
 enum vlmul_type
 {
@@ -247,7 +248,7 @@ void emit_vlmax_merge_insn (unsigned, int, rtx *);
 void emit_vlmax_cmp_insn (unsigned, rtx *);
 void emit_vlmax_cmp_mu_insn (unsigned, rtx *);
 void emit_vlmax_masked_mu_insn (unsigned, int, rtx *);
-void emit_scalar_move_insn (unsigned, rtx *);
+void emit_scalar_move_insn (unsigned, rtx *, rtx = 0);
 void emit_nonvlmax_integer_move_insn (unsigned, rtx *, rtx);
 enum vlmul_type get_vlmul (machine_mode);
 unsigned int get_ratio (machine_mode);
@@ -270,6 +271,13 @@ enum mask_policy
   MASK_AGNOSTIC = 1,
   MASK_ANY = 2,
 };
+
+enum class reduction_type
+{
+  UNORDERED,
+  FOLD_LEFT,
+  MASK_LEN_FOLD_LEFT,
+};
 enum tail_policy get_prefer_tail_policy ();
 enum mask_policy get_prefer_mask_policy ();
 rtx get_avl_type_rtx (enum avl_type);
@@ -282,7 +290,8 @@ bool has_vi_variant_p (rtx_code, rtx);
 void expand_vec_cmp (rtx, rtx_code, rtx, rtx);
 bool expand_vec_cmp_float (rtx, rtx_code, rtx, rtx, bool);
 void expand_cond_len_binop (rtx_code, rtx *);
-void expand_reduction (rtx_code, rtx *, rtx);
+void expand_reduction (rtx_code, rtx *, rtx,
+		       reduction_type = reduction_type::UNORDERED);
 #endif
 bool sew64_scalar_helper (rtx *, rtx *, rtx, machine_mode,
 			  bool, void (*)(rtx *, rtx));
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index 53088edf909..e338be151d3 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -1023,11 +1023,11 @@ emit_nonvlmax_fp_tu_insn (unsigned icode, int op_num, rtx *ops, rtx avl)
 /* Emit vmv.s.x instruction.  */
 
 void
-emit_scalar_move_insn (unsigned icode, rtx *ops)
+emit_scalar_move_insn (unsigned icode, rtx *ops, rtx len)
 {
   machine_mode dest_mode = GET_MODE (ops[0]);
   machine_mode mask_mode = get_mask_mode (dest_mode).require ();
-  insn_expander<RVV_INSN_OPERANDS_MAX> e (riscv_vector::RVV_SCALAR_MOV_OP,
+  insn_expander<RVV_INSN_OPERANDS_MAX> e (RVV_SCALAR_MOV_OP,
 					  /* HAS_DEST_P */ true,
 					  /* FULLY_UNMASKED_P */ false,
 					  /* USE_REAL_MERGE_P */ true,
@@ -1038,7 +1038,7 @@ emit_scalar_move_insn (unsigned icode, rtx *ops)
 
   e.set_policy (TAIL_ANY);
   e.set_policy (MASK_ANY);
-  e.set_vl (CONST1_RTX (Pmode));
+  e.set_vl (len ? len : CONST1_RTX (Pmode));
   e.emit_insn ((enum insn_code) icode, ops);
 }
 
@@ -1196,6 +1196,26 @@ emit_vlmax_fp_reduction_insn (unsigned icode, int op_num, rtx *ops)
   e.emit_insn ((enum insn_code) icode, ops);
 }
 
+/* Emit reduction instruction.  */
+static void
+emit_nonvlmax_fp_reduction_insn (unsigned icode, int op_num, rtx *ops, rtx vl)
+{
+  machine_mode dest_mode = GET_MODE (ops[0]);
+  machine_mode mask_mode = get_mask_mode (GET_MODE (ops[1])).require ();
+  insn_expander<RVV_INSN_OPERANDS_MAX> e (op_num,
+					  /* HAS_DEST_P */ true,
+					  /* FULLY_UNMASKED_P */ false,
+					  /* USE_REAL_MERGE_P */ true,
+					  /* HAS_AVL_P */ true,
+					  /* VLMAX_P */ false, dest_mode,
+					  mask_mode);
+
+  e.set_policy (TAIL_ANY);
+  e.set_rounding_mode (FRM_DYN);
+  e.set_vl (vl);
+  e.emit_insn ((enum insn_code) icode, ops);
+}
+
 /* Emit merge instruction.  */
 
 static machine_mode
@@ -3343,9 +3363,10 @@ expand_cond_len_ternop (unsigned icode, rtx *ops)
 
 /* Expand reduction operations.  */
 void
-expand_reduction (rtx_code code, rtx *ops, rtx init)
+expand_reduction (rtx_code code, rtx *ops, rtx init, reduction_type type)
 {
-  machine_mode vmode = GET_MODE (ops[1]);
+  rtx vector = type == reduction_type::UNORDERED ? ops[1] : ops[2];
+  machine_mode vmode = GET_MODE (vector);
   machine_mode m1_mode = get_m1_mode (vmode).require ();
   machine_mode m1_mmode = get_mask_mode (m1_mode).require ();
 
@@ -3353,16 +3374,30 @@ expand_reduction (rtx_code code, rtx *ops, rtx init)
   rtx m1_mask = gen_scalar_move_mask (m1_mmode);
   rtx m1_undef = RVV_VUNDEF (m1_mode);
   rtx scalar_move_ops[] = {m1_tmp, m1_mask, m1_undef, init};
-  emit_scalar_move_insn (code_for_pred_broadcast (m1_mode), scalar_move_ops);
+  rtx len = type == reduction_type::MASK_LEN_FOLD_LEFT ? ops[4] : NULL_RTX;
+  emit_scalar_move_insn (code_for_pred_broadcast (m1_mode), scalar_move_ops,
+			 len);
 
   rtx m1_tmp2 = gen_reg_rtx (m1_mode);
-  rtx reduc_ops[] = {m1_tmp2, ops[1], m1_tmp};
+  rtx reduc_ops[] = {m1_tmp2, vector, m1_tmp};
 
   if (FLOAT_MODE_P (vmode) && code == PLUS)
     {
       insn_code icode
-	= code_for_pred_reduc_plus (UNSPEC_UNORDERED, vmode, m1_mode);
-      emit_vlmax_fp_reduction_insn (icode, RVV_REDUCTION_OP, reduc_ops);
+	= code_for_pred_reduc_plus (type == reduction_type::UNORDERED
+				      ? UNSPEC_UNORDERED
+				      : UNSPEC_ORDERED,
+				    vmode, m1_mode);
+      if (type == reduction_type::MASK_LEN_FOLD_LEFT)
+	{
+	  rtx mask = ops[3];
+	  rtx mask_len_reduc_ops[]
+	    = {m1_tmp2, mask, RVV_VUNDEF (m1_mode), vector, m1_tmp};
+	  emit_nonvlmax_fp_reduction_insn (icode, RVV_REDUCTION_TU_OP,
+					   mask_len_reduc_ops, len);
+	}
+      else
+	emit_vlmax_fp_reduction_insn (icode, RVV_REDUCTION_OP, reduc_ops);
     }
   else
     {
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
new file mode 100644
index 00000000000..c293e9ae746
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
@@ -0,0 +1,28 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+#include <stdint-gcc.h>
+
+#define NUM_ELEMS(TYPE) ((int)(5 * (256 / sizeof (TYPE)) + 3))
+
+#define DEF_REDUC_PLUS(TYPE)			\
+  TYPE __attribute__ ((noinline, noclone))	\
+  reduc_plus_##TYPE (TYPE *a, TYPE *b)		\
+  {						\
+    TYPE r = 0, q = 3;				\
+    for (int i = 0; i < NUM_ELEMS (TYPE); i++)	\
+      {						\
+	r += a[i];				\
+	q -= b[i];				\
+      }						\
+    return r * q;				\
+  }
+
+#define TEST_ALL(T) \
+  T (_Float16) \
+  T (float) \
+  T (double)
+
+TEST_ALL (DEF_REDUC_PLUS)
+
+/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 6 } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
new file mode 100644
index 00000000000..2e1e7ab674d
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
@@ -0,0 +1,26 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+#define NUM_ELEMS(TYPE) ((int) (5 * (256 / sizeof (TYPE)) + 3))
+
+#define DEF_REDUC_PLUS(TYPE)					\
+void __attribute__ ((noinline, noclone))			\
+reduc_plus_##TYPE (TYPE (*restrict a)[NUM_ELEMS (TYPE)],	\
+		   TYPE *restrict r, int n)			\
+{								\
+  for (int i = 0; i < n; i++)					\
+    {								\
+      r[i] = 0;							\
+      for (int j = 0; j < NUM_ELEMS (TYPE); j++)		\
+        r[i] += a[i][j];					\
+    }								\
+}
+
+#define TEST_ALL(T) \
+  T (_Float16) \
+  T (float) \
+  T (double)
+
+TEST_ALL (DEF_REDUC_PLUS)
+
+/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 3 } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
new file mode 100644
index 00000000000..f559d40e60f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+double mat[100][2];
+
+double
+slp_reduc_plus (int n)
+{
+  double tmp = 0.0;
+  for (int i = 0; i < n; i++)
+    {
+      tmp = tmp + mat[i][0];
+      tmp = tmp + mat[i][1];
+    }
+  return tmp;
+}
+
+/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 1 } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
new file mode 100644
index 00000000000..428d371d9cf
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
@@ -0,0 +1,24 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+double mat[100][8];
+
+double
+slp_reduc_plus (int n)
+{
+  double tmp = 0.0;
+  for (int i = 0; i < n; i++)
+    {
+      tmp = tmp + mat[i][0];
+      tmp = tmp + mat[i][1];
+      tmp = tmp + mat[i][2];
+      tmp = tmp + mat[i][3];
+      tmp = tmp + mat[i][4];
+      tmp = tmp + mat[i][5];
+      tmp = tmp + mat[i][6];
+      tmp = tmp + mat[i][7];
+    }
+  return tmp;
+}
+
+/* { dg-final { scan-assembler {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
new file mode 100644
index 00000000000..24add2291f1
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
@@ -0,0 +1,28 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+double mat[100][12];
+
+double
+slp_reduc_plus (int n)
+{
+  double tmp = 0.0;
+  for (int i = 0; i < n; i++)
+    {
+      tmp = tmp + mat[i][0];
+      tmp = tmp + mat[i][1];
+      tmp = tmp + mat[i][2];
+      tmp = tmp + mat[i][3];
+      tmp = tmp + mat[i][4];
+      tmp = tmp + mat[i][5];
+      tmp = tmp + mat[i][6];
+      tmp = tmp + mat[i][7];
+      tmp = tmp + mat[i][8];
+      tmp = tmp + mat[i][9];
+      tmp = tmp + mat[i][10];
+      tmp = tmp + mat[i][11];
+    }
+  return tmp;
+}
+
+/* { dg-final { scan-assembler {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
new file mode 100644
index 00000000000..c1567b067ba
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model -fdump-tree-vect-details" } */
+
+float
+double_reduc (float (*i)[16])
+{
+  float l = 0;
+
+#pragma GCC unroll 0
+  for (int a = 0; a < 8; a++)
+    for (int b = 0; b < 100; b++)
+      l += i[b][a];
+  return l;
+}
+
+/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 1 } } */
+/* { dg-final { scan-tree-dump "Detected double reduction" "vect" } } */
+/* { dg-final { scan-tree-dump-not "OUTER LOOP VECTORIZED" "vect" } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
new file mode 100644
index 00000000000..f742a824bb2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
@@ -0,0 +1,21 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model -fdump-tree-vect-details" } */
+
+float
+double_reduc (float *i, float *j)
+{
+  float k = 0, l = 0;
+
+  for (int a = 0; a < 8; a++)
+    for (int b = 0; b < 100; b++)
+      {
+        k += i[b];
+        l += j[b];
+      }
+  return l * k;
+}
+
+/* { dg-final { scan-assembler-times {vle32\.v} 2 } } */
+/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 2 } } */
+/* { dg-final { scan-tree-dump "Detected double reduction" "vect" } } */
+/* { dg-final { scan-tree-dump-not "OUTER LOOP VECTORIZED" "vect" } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
new file mode 100644
index 00000000000..516be97e9eb
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
@@ -0,0 +1,29 @@
+/* { dg-do run { target { riscv_vector } } } */
+/* { dg-additional-options "--param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+#include "reduc_strict-1.c"
+
+#define TEST_REDUC_PLUS(TYPE)			\
+  {						\
+    TYPE a[NUM_ELEMS (TYPE)];			\
+    TYPE b[NUM_ELEMS (TYPE)];			\
+    TYPE r = 0, q = 3;				\
+    for (int i = 0; i < NUM_ELEMS (TYPE); i++)	\
+      {						\
+	a[i] = (i * 0.1) * (i & 1 ? 1 : -1);	\
+	b[i] = (i * 0.3) * (i & 1 ? 1 : -1);	\
+	r += a[i];				\
+	q -= b[i];				\
+	asm volatile ("" ::: "memory");		\
+      }						\
+    TYPE res = reduc_plus_##TYPE (a, b);	\
+    if (res != r * q)				\
+      __builtin_abort ();			\
+  }
+
+int __attribute__ ((optimize (1)))
+main ()
+{
+  TEST_ALL (TEST_REDUC_PLUS);
+  return 0;
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c
new file mode 100644
index 00000000000..0a4238d96f3
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c
@@ -0,0 +1,31 @@
+/* { dg-do run { target { riscv_vector } } } */
+/* { dg-additional-options "--param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
+
+#include "reduc_strict-2.c"
+
+#define NROWS 5
+
+#define TEST_REDUC_PLUS(TYPE)					\
+  {								\
+    TYPE a[NROWS][NUM_ELEMS (TYPE)];				\
+    TYPE r[NROWS];						\
+    TYPE expected[NROWS] = {};					\
+    for (int i = 0; i < NROWS; ++i)				\
+      for (int j = 0; j < NUM_ELEMS (TYPE); ++j)		\
+	{							\
+	  a[i][j] = (i * 0.1 + j * 0.6) * (j & 1 ? 1 : -1);	\
+	  expected[i] += a[i][j];				\
+	  asm volatile ("" ::: "memory");			\
+	}							\
+    reduc_plus_##TYPE (a, r, NROWS);				\
+    for (int i = 0; i < NROWS; ++i)				\
+      if (r[i] != expected[i])					\
+	__builtin_abort ();					\
+  }
+
+int __attribute__ ((optimize (1)))
+main ()
+{
+  TEST_ALL (TEST_REDUC_PLUS);
+  return 0;
+}
-- 
2.36.1


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] RISC-V: Support in-order floating-point reduction
  2023-07-20  8:51 [PATCH V2] RISC-V: Support in-order floating-point reduction Juzhe-Zhong
@ 2023-07-20  8:59 ` Kito Cheng
  2023-07-20  9:00   ` Robin Dapp
  0 siblings, 1 reply; 4+ messages in thread
From: Kito Cheng @ 2023-07-20  8:59 UTC (permalink / raw)
  To: Juzhe-Zhong; +Cc: gcc-patches, kito.cheng, jeffreyalaw, rdapp.gcc

LGTM, but I would like make sure Robin is OK too

On Thu, Jul 20, 2023 at 4:51 PM Juzhe-Zhong <juzhe.zhong@rivai.ai> wrote:
>
> This patch is depending on:
> https://gcc.gnu.org/pipermail/gcc-patches/2023-July/624995.html
>
> Consider this following case:
> float foo (float *__restrict a, int n)
> {
>   float result = 1.0;
>   for (int i = 0; i < n; i++)
>    result += a[i];
>   return result;
> }
>
> Compile with **NO** -ffast-math:
>
> Before this patch:
> <source>:4:21: missed: couldn't vectorize loop
> <source>:1:7: missed: not vectorized: relevant phi not supported: result_14 = PHI <result_11(6), 1.0e+0(5)>
>
> After this patch:
> foo:
>         lui     a5,%hi(.LC0)
>         flw     fa0,%lo(.LC0)(a5)
>         ble     a1,zero,.L4
> .L3:
>         vsetvli a5,a1,e32,m1,ta,ma
>         vle32.v v1,0(a0)
>         slli    a4,a5,2
>         sub     a1,a1,a5
>         vfmv.s.f        v2,fa0
>         add     a0,a0,a4
>         vfredosum.vs    v1,v1,v2     ----------> FOLD_LEFT_PLUS
>         vfmv.f.s        fa0,v1
>         bne     a1,zero,.L3
>         ret
> .L4:
>         ret
>
> gcc/ChangeLog:
>
>         * config/riscv/autovec.md (fold_left_plus_<mode>): New pattern.
>         (mask_len_fold_left_plus_<mode>): Ditto.
>         * config/riscv/riscv-protos.h (enum insn_type): New enum.
>         (enum reduction_type): Ditto.
>         (expand_reduction): Add in-order reduction.
>         * config/riscv/riscv-v.cc (emit_nonvlmax_fp_reduction_insn): New function.
>         (expand_reduction): Add in-order reduction.
>
> gcc/testsuite/ChangeLog:
>
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c: New test.
>         * gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c: New test.
>
> ---
>  gcc/config/riscv/autovec.md                   | 39 ++++++++++++++
>  gcc/config/riscv/riscv-protos.h               | 13 ++++-
>  gcc/config/riscv/riscv-v.cc                   | 53 +++++++++++++++----
>  .../riscv/rvv/autovec/reduc/reduc_strict-1.c  | 28 ++++++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-2.c  | 26 +++++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-3.c  | 18 +++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-4.c  | 24 +++++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-5.c  | 28 ++++++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-6.c  | 18 +++++++
>  .../riscv/rvv/autovec/reduc/reduc_strict-7.c  | 21 ++++++++
>  .../rvv/autovec/reduc/reduc_strict_run-1.c    | 29 ++++++++++
>  .../rvv/autovec/reduc/reduc_strict_run-2.c    | 31 +++++++++++
>  12 files changed, 317 insertions(+), 11 deletions(-)
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
>  create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c
>
> diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
> index 00947207f3f..667a877d009 100644
> --- a/gcc/config/riscv/autovec.md
> +++ b/gcc/config/riscv/autovec.md
> @@ -1687,3 +1687,42 @@
>    riscv_vector::expand_reduction (SMIN, operands, f);
>    DONE;
>  })
> +
> +;; -------------------------------------------------------------------------
> +;; ---- [FP] Left-to-right reductions
> +;; -------------------------------------------------------------------------
> +;; Includes:
> +;; - vfredosum.vs
> +;; -------------------------------------------------------------------------
> +
> +;; Unpredicated in-order FP reductions.
> +(define_expand "fold_left_plus_<mode>"
> +  [(match_operand:<VEL> 0 "register_operand")
> +   (match_operand:<VEL> 1 "register_operand")
> +   (match_operand:VF 2 "register_operand")]
> +  "TARGET_VECTOR"
> +{
> +  riscv_vector::expand_reduction (PLUS, operands,
> +                                 operands[1],
> +                                 riscv_vector::reduction_type::FOLD_LEFT);
> +  DONE;
> +})
> +
> +;; Predicated in-order FP reductions.
> +(define_expand "mask_len_fold_left_plus_<mode>"
> +  [(match_operand:<VEL> 0 "register_operand")
> +   (match_operand:<VEL> 1 "register_operand")
> +   (match_operand:VF 2 "register_operand")
> +   (match_operand:<VM> 3 "vector_mask_operand")
> +   (match_operand 4 "autovec_length_operand")
> +   (match_operand 5 "const_0_operand")]
> +  "TARGET_VECTOR"
> +{
> +  if (rtx_equal_p (operands[4], const0_rtx))
> +    emit_move_insn (operands[0], operands[1]);
> +  else
> +    riscv_vector::expand_reduction (PLUS, operands,
> +                                   operands[1],
> +                                   riscv_vector::reduction_type::MASK_LEN_FOLD_LEFT);
> +  DONE;
> +})
> diff --git a/gcc/config/riscv/riscv-protos.h b/gcc/config/riscv/riscv-protos.h
> index 16fb8dabca0..c9520f689e2 100644
> --- a/gcc/config/riscv/riscv-protos.h
> +++ b/gcc/config/riscv/riscv-protos.h
> @@ -199,6 +199,7 @@ enum insn_type
>    RVV_GATHER_M_OP = 5,
>    RVV_SCATTER_M_OP = 4,
>    RVV_REDUCTION_OP = 3,
> +  RVV_REDUCTION_TU_OP = RVV_REDUCTION_OP + 2,
>  };
>  enum vlmul_type
>  {
> @@ -247,7 +248,7 @@ void emit_vlmax_merge_insn (unsigned, int, rtx *);
>  void emit_vlmax_cmp_insn (unsigned, rtx *);
>  void emit_vlmax_cmp_mu_insn (unsigned, rtx *);
>  void emit_vlmax_masked_mu_insn (unsigned, int, rtx *);
> -void emit_scalar_move_insn (unsigned, rtx *);
> +void emit_scalar_move_insn (unsigned, rtx *, rtx = 0);
>  void emit_nonvlmax_integer_move_insn (unsigned, rtx *, rtx);
>  enum vlmul_type get_vlmul (machine_mode);
>  unsigned int get_ratio (machine_mode);
> @@ -270,6 +271,13 @@ enum mask_policy
>    MASK_AGNOSTIC = 1,
>    MASK_ANY = 2,
>  };
> +
> +enum class reduction_type
> +{
> +  UNORDERED,
> +  FOLD_LEFT,
> +  MASK_LEN_FOLD_LEFT,
> +};
>  enum tail_policy get_prefer_tail_policy ();
>  enum mask_policy get_prefer_mask_policy ();
>  rtx get_avl_type_rtx (enum avl_type);
> @@ -282,7 +290,8 @@ bool has_vi_variant_p (rtx_code, rtx);
>  void expand_vec_cmp (rtx, rtx_code, rtx, rtx);
>  bool expand_vec_cmp_float (rtx, rtx_code, rtx, rtx, bool);
>  void expand_cond_len_binop (rtx_code, rtx *);
> -void expand_reduction (rtx_code, rtx *, rtx);
> +void expand_reduction (rtx_code, rtx *, rtx,
> +                      reduction_type = reduction_type::UNORDERED);
>  #endif
>  bool sew64_scalar_helper (rtx *, rtx *, rtx, machine_mode,
>                           bool, void (*)(rtx *, rtx));
> diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
> index 53088edf909..e338be151d3 100644
> --- a/gcc/config/riscv/riscv-v.cc
> +++ b/gcc/config/riscv/riscv-v.cc
> @@ -1023,11 +1023,11 @@ emit_nonvlmax_fp_tu_insn (unsigned icode, int op_num, rtx *ops, rtx avl)
>  /* Emit vmv.s.x instruction.  */
>
>  void
> -emit_scalar_move_insn (unsigned icode, rtx *ops)
> +emit_scalar_move_insn (unsigned icode, rtx *ops, rtx len)
>  {
>    machine_mode dest_mode = GET_MODE (ops[0]);
>    machine_mode mask_mode = get_mask_mode (dest_mode).require ();
> -  insn_expander<RVV_INSN_OPERANDS_MAX> e (riscv_vector::RVV_SCALAR_MOV_OP,
> +  insn_expander<RVV_INSN_OPERANDS_MAX> e (RVV_SCALAR_MOV_OP,
>                                           /* HAS_DEST_P */ true,
>                                           /* FULLY_UNMASKED_P */ false,
>                                           /* USE_REAL_MERGE_P */ true,
> @@ -1038,7 +1038,7 @@ emit_scalar_move_insn (unsigned icode, rtx *ops)
>
>    e.set_policy (TAIL_ANY);
>    e.set_policy (MASK_ANY);
> -  e.set_vl (CONST1_RTX (Pmode));
> +  e.set_vl (len ? len : CONST1_RTX (Pmode));
>    e.emit_insn ((enum insn_code) icode, ops);
>  }
>
> @@ -1196,6 +1196,26 @@ emit_vlmax_fp_reduction_insn (unsigned icode, int op_num, rtx *ops)
>    e.emit_insn ((enum insn_code) icode, ops);
>  }
>
> +/* Emit reduction instruction.  */
> +static void
> +emit_nonvlmax_fp_reduction_insn (unsigned icode, int op_num, rtx *ops, rtx vl)
> +{
> +  machine_mode dest_mode = GET_MODE (ops[0]);
> +  machine_mode mask_mode = get_mask_mode (GET_MODE (ops[1])).require ();
> +  insn_expander<RVV_INSN_OPERANDS_MAX> e (op_num,
> +                                         /* HAS_DEST_P */ true,
> +                                         /* FULLY_UNMASKED_P */ false,
> +                                         /* USE_REAL_MERGE_P */ true,
> +                                         /* HAS_AVL_P */ true,
> +                                         /* VLMAX_P */ false, dest_mode,
> +                                         mask_mode);
> +
> +  e.set_policy (TAIL_ANY);
> +  e.set_rounding_mode (FRM_DYN);
> +  e.set_vl (vl);
> +  e.emit_insn ((enum insn_code) icode, ops);
> +}
> +
>  /* Emit merge instruction.  */
>
>  static machine_mode
> @@ -3343,9 +3363,10 @@ expand_cond_len_ternop (unsigned icode, rtx *ops)
>
>  /* Expand reduction operations.  */
>  void
> -expand_reduction (rtx_code code, rtx *ops, rtx init)
> +expand_reduction (rtx_code code, rtx *ops, rtx init, reduction_type type)
>  {
> -  machine_mode vmode = GET_MODE (ops[1]);
> +  rtx vector = type == reduction_type::UNORDERED ? ops[1] : ops[2];
> +  machine_mode vmode = GET_MODE (vector);
>    machine_mode m1_mode = get_m1_mode (vmode).require ();
>    machine_mode m1_mmode = get_mask_mode (m1_mode).require ();
>
> @@ -3353,16 +3374,30 @@ expand_reduction (rtx_code code, rtx *ops, rtx init)
>    rtx m1_mask = gen_scalar_move_mask (m1_mmode);
>    rtx m1_undef = RVV_VUNDEF (m1_mode);
>    rtx scalar_move_ops[] = {m1_tmp, m1_mask, m1_undef, init};
> -  emit_scalar_move_insn (code_for_pred_broadcast (m1_mode), scalar_move_ops);
> +  rtx len = type == reduction_type::MASK_LEN_FOLD_LEFT ? ops[4] : NULL_RTX;
> +  emit_scalar_move_insn (code_for_pred_broadcast (m1_mode), scalar_move_ops,
> +                        len);
>
>    rtx m1_tmp2 = gen_reg_rtx (m1_mode);
> -  rtx reduc_ops[] = {m1_tmp2, ops[1], m1_tmp};
> +  rtx reduc_ops[] = {m1_tmp2, vector, m1_tmp};
>
>    if (FLOAT_MODE_P (vmode) && code == PLUS)
>      {
>        insn_code icode
> -       = code_for_pred_reduc_plus (UNSPEC_UNORDERED, vmode, m1_mode);
> -      emit_vlmax_fp_reduction_insn (icode, RVV_REDUCTION_OP, reduc_ops);
> +       = code_for_pred_reduc_plus (type == reduction_type::UNORDERED
> +                                     ? UNSPEC_UNORDERED
> +                                     : UNSPEC_ORDERED,
> +                                   vmode, m1_mode);
> +      if (type == reduction_type::MASK_LEN_FOLD_LEFT)
> +       {
> +         rtx mask = ops[3];
> +         rtx mask_len_reduc_ops[]
> +           = {m1_tmp2, mask, RVV_VUNDEF (m1_mode), vector, m1_tmp};
> +         emit_nonvlmax_fp_reduction_insn (icode, RVV_REDUCTION_TU_OP,
> +                                          mask_len_reduc_ops, len);
> +       }
> +      else
> +       emit_vlmax_fp_reduction_insn (icode, RVV_REDUCTION_OP, reduc_ops);
>      }
>    else
>      {
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
> new file mode 100644
> index 00000000000..c293e9ae746
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-1.c
> @@ -0,0 +1,28 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +#include <stdint-gcc.h>
> +
> +#define NUM_ELEMS(TYPE) ((int)(5 * (256 / sizeof (TYPE)) + 3))
> +
> +#define DEF_REDUC_PLUS(TYPE)                   \
> +  TYPE __attribute__ ((noinline, noclone))     \
> +  reduc_plus_##TYPE (TYPE *a, TYPE *b)         \
> +  {                                            \
> +    TYPE r = 0, q = 3;                         \
> +    for (int i = 0; i < NUM_ELEMS (TYPE); i++) \
> +      {                                                \
> +       r += a[i];                              \
> +       q -= b[i];                              \
> +      }                                                \
> +    return r * q;                              \
> +  }
> +
> +#define TEST_ALL(T) \
> +  T (_Float16) \
> +  T (float) \
> +  T (double)
> +
> +TEST_ALL (DEF_REDUC_PLUS)
> +
> +/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 6 } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
> new file mode 100644
> index 00000000000..2e1e7ab674d
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-2.c
> @@ -0,0 +1,26 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +#define NUM_ELEMS(TYPE) ((int) (5 * (256 / sizeof (TYPE)) + 3))
> +
> +#define DEF_REDUC_PLUS(TYPE)                                   \
> +void __attribute__ ((noinline, noclone))                       \
> +reduc_plus_##TYPE (TYPE (*restrict a)[NUM_ELEMS (TYPE)],       \
> +                  TYPE *restrict r, int n)                     \
> +{                                                              \
> +  for (int i = 0; i < n; i++)                                  \
> +    {                                                          \
> +      r[i] = 0;                                                        \
> +      for (int j = 0; j < NUM_ELEMS (TYPE); j++)               \
> +        r[i] += a[i][j];                                       \
> +    }                                                          \
> +}
> +
> +#define TEST_ALL(T) \
> +  T (_Float16) \
> +  T (float) \
> +  T (double)
> +
> +TEST_ALL (DEF_REDUC_PLUS)
> +
> +/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 3 } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
> new file mode 100644
> index 00000000000..f559d40e60f
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-3.c
> @@ -0,0 +1,18 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +double mat[100][2];
> +
> +double
> +slp_reduc_plus (int n)
> +{
> +  double tmp = 0.0;
> +  for (int i = 0; i < n; i++)
> +    {
> +      tmp = tmp + mat[i][0];
> +      tmp = tmp + mat[i][1];
> +    }
> +  return tmp;
> +}
> +
> +/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 1 } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
> new file mode 100644
> index 00000000000..428d371d9cf
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-4.c
> @@ -0,0 +1,24 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +double mat[100][8];
> +
> +double
> +slp_reduc_plus (int n)
> +{
> +  double tmp = 0.0;
> +  for (int i = 0; i < n; i++)
> +    {
> +      tmp = tmp + mat[i][0];
> +      tmp = tmp + mat[i][1];
> +      tmp = tmp + mat[i][2];
> +      tmp = tmp + mat[i][3];
> +      tmp = tmp + mat[i][4];
> +      tmp = tmp + mat[i][5];
> +      tmp = tmp + mat[i][6];
> +      tmp = tmp + mat[i][7];
> +    }
> +  return tmp;
> +}
> +
> +/* { dg-final { scan-assembler {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
> new file mode 100644
> index 00000000000..24add2291f1
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-5.c
> @@ -0,0 +1,28 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +double mat[100][12];
> +
> +double
> +slp_reduc_plus (int n)
> +{
> +  double tmp = 0.0;
> +  for (int i = 0; i < n; i++)
> +    {
> +      tmp = tmp + mat[i][0];
> +      tmp = tmp + mat[i][1];
> +      tmp = tmp + mat[i][2];
> +      tmp = tmp + mat[i][3];
> +      tmp = tmp + mat[i][4];
> +      tmp = tmp + mat[i][5];
> +      tmp = tmp + mat[i][6];
> +      tmp = tmp + mat[i][7];
> +      tmp = tmp + mat[i][8];
> +      tmp = tmp + mat[i][9];
> +      tmp = tmp + mat[i][10];
> +      tmp = tmp + mat[i][11];
> +    }
> +  return tmp;
> +}
> +
> +/* { dg-final { scan-assembler {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
> new file mode 100644
> index 00000000000..c1567b067ba
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-6.c
> @@ -0,0 +1,18 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model -fdump-tree-vect-details" } */
> +
> +float
> +double_reduc (float (*i)[16])
> +{
> +  float l = 0;
> +
> +#pragma GCC unroll 0
> +  for (int a = 0; a < 8; a++)
> +    for (int b = 0; b < 100; b++)
> +      l += i[b][a];
> +  return l;
> +}
> +
> +/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 1 } } */
> +/* { dg-final { scan-tree-dump "Detected double reduction" "vect" } } */
> +/* { dg-final { scan-tree-dump-not "OUTER LOOP VECTORIZED" "vect" } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
> new file mode 100644
> index 00000000000..f742a824bb2
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict-7.c
> @@ -0,0 +1,21 @@
> +/* { dg-do compile } */
> +/* { dg-additional-options "-march=rv32gcv_zvfh -mabi=ilp32d --param=riscv-autovec-preference=scalable -fno-vect-cost-model -fdump-tree-vect-details" } */
> +
> +float
> +double_reduc (float *i, float *j)
> +{
> +  float k = 0, l = 0;
> +
> +  for (int a = 0; a < 8; a++)
> +    for (int b = 0; b < 100; b++)
> +      {
> +        k += i[b];
> +        l += j[b];
> +      }
> +  return l * k;
> +}
> +
> +/* { dg-final { scan-assembler-times {vle32\.v} 2 } } */
> +/* { dg-final { scan-assembler-times {vfredosum\.vs\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 2 } } */
> +/* { dg-final { scan-tree-dump "Detected double reduction" "vect" } } */
> +/* { dg-final { scan-tree-dump-not "OUTER LOOP VECTORIZED" "vect" } } */
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
> new file mode 100644
> index 00000000000..516be97e9eb
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-1.c
> @@ -0,0 +1,29 @@
> +/* { dg-do run { target { riscv_vector } } } */
> +/* { dg-additional-options "--param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +#include "reduc_strict-1.c"
> +
> +#define TEST_REDUC_PLUS(TYPE)                  \
> +  {                                            \
> +    TYPE a[NUM_ELEMS (TYPE)];                  \
> +    TYPE b[NUM_ELEMS (TYPE)];                  \
> +    TYPE r = 0, q = 3;                         \
> +    for (int i = 0; i < NUM_ELEMS (TYPE); i++) \
> +      {                                                \
> +       a[i] = (i * 0.1) * (i & 1 ? 1 : -1);    \
> +       b[i] = (i * 0.3) * (i & 1 ? 1 : -1);    \
> +       r += a[i];                              \
> +       q -= b[i];                              \
> +       asm volatile ("" ::: "memory");         \
> +      }                                                \
> +    TYPE res = reduc_plus_##TYPE (a, b);       \
> +    if (res != r * q)                          \
> +      __builtin_abort ();                      \
> +  }
> +
> +int __attribute__ ((optimize (1)))
> +main ()
> +{
> +  TEST_ALL (TEST_REDUC_PLUS);
> +  return 0;
> +}
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c
> new file mode 100644
> index 00000000000..0a4238d96f3
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/reduc/reduc_strict_run-2.c
> @@ -0,0 +1,31 @@
> +/* { dg-do run { target { riscv_vector } } } */
> +/* { dg-additional-options "--param=riscv-autovec-preference=scalable -fno-vect-cost-model" } */
> +
> +#include "reduc_strict-2.c"
> +
> +#define NROWS 5
> +
> +#define TEST_REDUC_PLUS(TYPE)                                  \
> +  {                                                            \
> +    TYPE a[NROWS][NUM_ELEMS (TYPE)];                           \
> +    TYPE r[NROWS];                                             \
> +    TYPE expected[NROWS] = {};                                 \
> +    for (int i = 0; i < NROWS; ++i)                            \
> +      for (int j = 0; j < NUM_ELEMS (TYPE); ++j)               \
> +       {                                                       \
> +         a[i][j] = (i * 0.1 + j * 0.6) * (j & 1 ? 1 : -1);     \
> +         expected[i] += a[i][j];                               \
> +         asm volatile ("" ::: "memory");                       \
> +       }                                                       \
> +    reduc_plus_##TYPE (a, r, NROWS);                           \
> +    for (int i = 0; i < NROWS; ++i)                            \
> +      if (r[i] != expected[i])                                 \
> +       __builtin_abort ();                                     \
> +  }
> +
> +int __attribute__ ((optimize (1)))
> +main ()
> +{
> +  TEST_ALL (TEST_REDUC_PLUS);
> +  return 0;
> +}
> --
> 2.36.1
>

^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] RISC-V: Support in-order floating-point reduction
  2023-07-20  8:59 ` Kito Cheng
@ 2023-07-20  9:00   ` Robin Dapp
  2023-07-24  8:38     ` Lehua Ding
  0 siblings, 1 reply; 4+ messages in thread
From: Robin Dapp @ 2023-07-20  9:00 UTC (permalink / raw)
  To: Kito Cheng, Juzhe-Zhong; +Cc: rdapp.gcc, gcc-patches, kito.cheng, jeffreyalaw

> LGTM, but I would like make sure Robin is OK too

Yes, LGTM as well.

Regards
 Robin


^ permalink raw reply	[flat|nested] 4+ messages in thread

* Re: [PATCH V2] RISC-V: Support in-order floating-point reduction
  2023-07-20  9:00   ` Robin Dapp
@ 2023-07-24  8:38     ` Lehua Ding
  0 siblings, 0 replies; 4+ messages in thread
From: Lehua Ding @ 2023-07-24  8:38 UTC (permalink / raw)
  To: Kito Cheng, Robin Dapp, gcc-patches; +Cc: jeffreyalaw, Juzhe-Zhong

[-- Attachment #1: Type: text/plain, Size: 62 bytes --]

Committed to the trunk, thanks Kito and Robin.

Best,
Lehua

^ permalink raw reply	[flat|nested] 4+ messages in thread

end of thread, other threads:[~2023-07-24  8:38 UTC | newest]

Thread overview: 4+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-07-20  8:51 [PATCH V2] RISC-V: Support in-order floating-point reduction Juzhe-Zhong
2023-07-20  8:59 ` Kito Cheng
2023-07-20  9:00   ` Robin Dapp
2023-07-24  8:38     ` Lehua Ding

This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for read-only IMAP folder(s) and NNTP newsgroup(s).