public inbox for gcc-patches@gcc.gnu.org
 help / color / mirror / Atom feed
* [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode
@ 2023-09-02  8:53 pan2.li
  2023-09-02 15:40 ` Kito Cheng
  0 siblings, 1 reply; 3+ messages in thread
From: pan2.li @ 2023-09-02  8:53 UTC (permalink / raw)
  To: gcc-patches; +Cc: juzhe.zhong, pan2.li, yanzhang.wang, kito.cheng

From: Pan Li <pan2.li@intel.com>

This patch would like to allow the VLS mode autovec for the
floating-point binary operation MAX/MIN.

Given below code example:

test (float *out, float *in1, float *in2)
{
  for (int i = 0; i < 128; i++)
    out[i] = in1[i] > in2[i] ? in1[i] : in2[i];
    // Or out[i] = fmax (in1[i], in2[i]);
}

Before this patch:
test:
  csrr    a4,vlenb
  slli    a4,a4,1
  li      a5,128
  bleu    a5,a4,.L2
  mv      a5,a4
.L2:
  vsetvli zero,a5,e32,m8,ta,ma
  vle32.v v16,0(a1)
  vle32.v v8,0(a2)
  vsetvli a3,zero,e32,m8,ta,ma
  vmfgt.vv        v0,v16,v8
  vmerge.vvm      v8,v8,v16,v0
  vsetvli zero,a5,e32,m8,ta,ma
  vse32.v v8,0(a0)
  ret

After this patch:
test:
  li      a5,128
  vsetvli zero,a5,e32,m1,ta,ma
  vle32.v v1,0(a1)
  vle32.v v2,0(a2)
  vfmax.vv        v1,v1,v2
  vse32.v v1,0(a0)
  ret

This MAX/MIN autovec acts on function call like fmaxf/fmax in math.h
too. And it depends on the option -ffast-math.

Signed-off-by: Pan Li <pan2.li@intel.com>

gcc/ChangeLog:

	* config/riscv/autovec-vls.md (<optab><mode>3): New pattern for
	fmax/fmin
	* config/riscv/vector.md: Add VLS modes to vfmax/vfmin.

gcc/testsuite/ChangeLog:

	* gcc.target/riscv/rvv/autovec/vls/def.h: New macros.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c: New test.
---
 gcc/config/riscv/autovec-vls.md               | 23 ++++++++++
 gcc/config/riscv/vector.md                    | 12 +++---
 .../gcc.target/riscv/rvv/autovec/vls/def.h    | 16 +++++++
 .../rvv/autovec/vls/floating-point-max-1.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-2.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-3.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-4.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-5.c    | 31 +++++++++++++
 .../rvv/autovec/vls/floating-point-min-1.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-2.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-3.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-4.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-5.c    | 31 +++++++++++++
 13 files changed, 451 insertions(+), 6 deletions(-)
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c

diff --git a/gcc/config/riscv/autovec-vls.md b/gcc/config/riscv/autovec-vls.md
index 4ca640c11e2..7ef29637e33 100644
--- a/gcc/config/riscv/autovec-vls.md
+++ b/gcc/config/riscv/autovec-vls.md
@@ -232,6 +232,29 @@ (define_insn_and_split "<optab><mode>3"
 [(set_attr "type" "vector")]
 )
 
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - vfmin.vv/vfmax.vv
+;; - vfmin.vf/vfmax.vf
+;; - fmax/fmaxf in math.h
+;; -------------------------------------------------------------------------
+(define_insn_and_split "<optab><mode>3"
+  [(set (match_operand:VLSF 0 "register_operand")
+    (any_float_binop_nofrm:VLSF
+     (match_operand:VLSF 1 "<binop_rhs1_predicate>")
+     (match_operand:VLSF 2 "<binop_rhs2_predicate>")))]
+  "TARGET_VECTOR && can_create_pseudo_p ()"
+  "#"
+  "&& 1"
+  [(const_int 0)]
+{
+  riscv_vector::emit_vlmax_insn (code_for_pred (<CODE>, <MODE>mode),
+				 riscv_vector::BINARY_OP, operands);
+  DONE;
+}
+[(set_attr "type" "vector")]
+)
+
 ;; -------------------------------------------------------------------------------
 ;; ---- [INT] Unary operations
 ;; -------------------------------------------------------------------------------
diff --git a/gcc/config/riscv/vector.md b/gcc/config/riscv/vector.md
index 6fe750ca8a4..9d7b4bbe1d4 100644
--- a/gcc/config/riscv/vector.md
+++ b/gcc/config/riscv/vector.md
@@ -6050,8 +6050,8 @@ (define_insn "@pred_<optab><mode>"
 	(symbol_ref "riscv_vector::get_frm_mode (operands[9])"))])
 
 (define_insn "@pred_<optab><mode>"
-  [(set (match_operand:VF 0 "register_operand"           "=vd, vd, vr, vr")
-	(if_then_else:VF
+  [(set (match_operand:V_VLSF 0 "register_operand"           "=vd, vd, vr, vr")
+	(if_then_else:V_VLSF
 	  (unspec:<VM>
 	    [(match_operand:<VM> 1 "vector_mask_operand" " vm, vm,Wc1,Wc1")
 	     (match_operand 5 "vector_length_operand"    " rK, rK, rK, rK")
@@ -6060,10 +6060,10 @@ (define_insn "@pred_<optab><mode>"
 	     (match_operand 8 "const_int_operand"        "  i,  i,  i,  i")
 	     (reg:SI VL_REGNUM)
 	     (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE)
-	  (any_float_binop_nofrm:VF
-	    (match_operand:VF 3 "register_operand"       " vr, vr, vr, vr")
-	    (match_operand:VF 4 "register_operand"       " vr, vr, vr, vr"))
-	  (match_operand:VF 2 "vector_merge_operand"     " vu,  0, vu,  0")))]
+	  (any_float_binop_nofrm:V_VLSF
+	    (match_operand:V_VLSF 3 "register_operand"       " vr, vr, vr, vr")
+	    (match_operand:V_VLSF 4 "register_operand"       " vr, vr, vr, vr"))
+	  (match_operand:V_VLSF 2 "vector_merge_operand"     " vu,  0, vu,  0")))]
   "TARGET_VECTOR"
   "vf<insn>.vv\t%0,%3,%4%p1"
   [(set_attr "type" "<float_insn_type>")
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
index 476f966c427..2e07e908736 100644
--- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
@@ -143,6 +143,14 @@ typedef double v512df __attribute__ ((vector_size (4096)));
       a[i] = b[i] OP c[i] ? b[i] : c[i];                                       \
   }
 
+#define DEF_MINMAX_VX(PREFIX, NUM, TYPE, OP)                                   \
+  void __attribute__ ((noinline, noclone))                                     \
+  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE c)            \
+  {                                                                            \
+    for (int i = 0; i < NUM; ++i)                                              \
+      a[i] = b[i] OP c ? b[i] : c;                                             \
+  }
+
 #define DEF_OP_VI_7(PREFIX, NUM, TYPE, OP)                                     \
   void __attribute__ ((noinline, noclone))                                     \
   PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c)  \
@@ -159,6 +167,14 @@ typedef double v512df __attribute__ ((vector_size (4096)));
       a[i] = OP b[i];                                                          \
   }
 
+#define DEF_CALL_VV(PREFIX, NUM, TYPE, CALL)                                   \
+  void __attribute__ ((noinline, noclone))                                     \
+  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c)  \
+  {                                                                            \
+    for (int i = 0; i < NUM; ++i)                                              \
+      a[i] = CALL (b[i], c[i]);                                                \
+  }
+
 #define DEF_CONST(TYPE, VAL, NUM)                                              \
   void const_##TYPE##_##NUM (TYPE *restrict a)                                 \
   {                                                                            \
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
new file mode 100644
index 00000000000..8d3cd2aa538
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (max, 1, _Float16, >)
+DEF_MINMAX_VV (max, 2, _Float16, >)
+DEF_MINMAX_VV (max, 4, _Float16, >)
+DEF_MINMAX_VV (max, 8, _Float16, >)
+DEF_MINMAX_VV (max, 16, _Float16, >)
+DEF_MINMAX_VV (max, 32, _Float16, >)
+DEF_MINMAX_VV (max, 64, _Float16, >)
+DEF_MINMAX_VV (max, 128, _Float16, >)
+DEF_MINMAX_VV (max, 256, _Float16, >)
+DEF_MINMAX_VV (max, 512, _Float16, >)
+DEF_MINMAX_VV (max, 1024, _Float16, >)
+DEF_MINMAX_VV (max, 2048, _Float16, >)
+
+DEF_MINMAX_VV (max, 1, float, >)
+DEF_MINMAX_VV (max, 2, float, >)
+DEF_MINMAX_VV (max, 4, float, >)
+DEF_MINMAX_VV (max, 8, float, >)
+DEF_MINMAX_VV (max, 16, float, >)
+DEF_MINMAX_VV (max, 32, float, >)
+DEF_MINMAX_VV (max, 64, float, >)
+DEF_MINMAX_VV (max, 128, float, >)
+DEF_MINMAX_VV (max, 256, float, >)
+DEF_MINMAX_VV (max, 512, float, >)
+DEF_MINMAX_VV (max, 1024, float, >)
+
+DEF_MINMAX_VV (max, 1, double, >)
+DEF_MINMAX_VV (max, 2, double, >)
+DEF_MINMAX_VV (max, 4, double, >)
+DEF_MINMAX_VV (max, 8, double, >)
+DEF_MINMAX_VV (max, 16, double, >)
+DEF_MINMAX_VV (max, 32, double, >)
+DEF_MINMAX_VV (max, 64, double, >)
+DEF_MINMAX_VV (max, 128, double, >)
+DEF_MINMAX_VV (max, 256, double, >)
+DEF_MINMAX_VV (max, 512, double, >)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
new file mode 100644
index 00000000000..a13de042041
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (max, 1, _Float16, >)
+DEF_MINMAX_VX (max, 2, _Float16, >)
+DEF_MINMAX_VX (max, 4, _Float16, >)
+DEF_MINMAX_VX (max, 8, _Float16, >)
+DEF_MINMAX_VX (max, 16, _Float16, >)
+DEF_MINMAX_VX (max, 32, _Float16, >)
+DEF_MINMAX_VX (max, 64, _Float16, >)
+DEF_MINMAX_VX (max, 128, _Float16, >)
+DEF_MINMAX_VX (max, 256, _Float16, >)
+DEF_MINMAX_VX (max, 512, _Float16, >)
+DEF_MINMAX_VX (max, 1024, _Float16, >)
+DEF_MINMAX_VX (max, 2048, _Float16, >)
+
+DEF_MINMAX_VX (max, 1, float, >)
+DEF_MINMAX_VX (max, 2, float, >)
+DEF_MINMAX_VX (max, 4, float, >)
+DEF_MINMAX_VX (max, 8, float, >)
+DEF_MINMAX_VX (max, 16, float, >)
+DEF_MINMAX_VX (max, 32, float, >)
+DEF_MINMAX_VX (max, 64, float, >)
+DEF_MINMAX_VX (max, 128, float, >)
+DEF_MINMAX_VX (max, 256, float, >)
+DEF_MINMAX_VX (max, 512, float, >)
+DEF_MINMAX_VX (max, 1024, float, >)
+
+DEF_MINMAX_VX (max, 1, double, >)
+DEF_MINMAX_VX (max, 2, double, >)
+DEF_MINMAX_VX (max, 4, double, >)
+DEF_MINMAX_VX (max, 8, double, >)
+DEF_MINMAX_VX (max, 16, double, >)
+DEF_MINMAX_VX (max, 32, double, >)
+DEF_MINMAX_VX (max, 64, double, >)
+DEF_MINMAX_VX (max, 128, double, >)
+DEF_MINMAX_VX (max, 256, double, >)
+DEF_MINMAX_VX (max, 512, double, >)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
new file mode 100644
index 00000000000..108a883bba5
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (max, 1, _Float16, >=)
+DEF_MINMAX_VV (max, 2, _Float16, >=)
+DEF_MINMAX_VV (max, 4, _Float16, >=)
+DEF_MINMAX_VV (max, 8, _Float16, >=)
+DEF_MINMAX_VV (max, 16, _Float16, >=)
+DEF_MINMAX_VV (max, 32, _Float16, >=)
+DEF_MINMAX_VV (max, 64, _Float16, >=)
+DEF_MINMAX_VV (max, 128, _Float16, >=)
+DEF_MINMAX_VV (max, 256, _Float16, >=)
+DEF_MINMAX_VV (max, 512, _Float16, >=)
+DEF_MINMAX_VV (max, 1024, _Float16, >=)
+DEF_MINMAX_VV (max, 2048, _Float16, >=)
+
+DEF_MINMAX_VV (max, 1, float, >=)
+DEF_MINMAX_VV (max, 2, float, >=)
+DEF_MINMAX_VV (max, 4, float, >=)
+DEF_MINMAX_VV (max, 8, float, >=)
+DEF_MINMAX_VV (max, 16, float, >=)
+DEF_MINMAX_VV (max, 32, float, >=)
+DEF_MINMAX_VV (max, 64, float, >=)
+DEF_MINMAX_VV (max, 128, float, >=)
+DEF_MINMAX_VV (max, 256, float, >=)
+DEF_MINMAX_VV (max, 512, float, >=)
+DEF_MINMAX_VV (max, 1024, float, >=)
+
+DEF_MINMAX_VV (max, 1, double, >=)
+DEF_MINMAX_VV (max, 2, double, >=)
+DEF_MINMAX_VV (max, 4, double, >=)
+DEF_MINMAX_VV (max, 8, double, >=)
+DEF_MINMAX_VV (max, 16, double, >=)
+DEF_MINMAX_VV (max, 32, double, >=)
+DEF_MINMAX_VV (max, 64, double, >=)
+DEF_MINMAX_VV (max, 128, double, >=)
+DEF_MINMAX_VV (max, 256, double, >=)
+DEF_MINMAX_VV (max, 512, double, >=)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
new file mode 100644
index 00000000000..d74801887b6
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (max, 1, _Float16, >=)
+DEF_MINMAX_VX (max, 2, _Float16, >=)
+DEF_MINMAX_VX (max, 4, _Float16, >=)
+DEF_MINMAX_VX (max, 8, _Float16, >=)
+DEF_MINMAX_VX (max, 16, _Float16, >=)
+DEF_MINMAX_VX (max, 32, _Float16, >=)
+DEF_MINMAX_VX (max, 64, _Float16, >=)
+DEF_MINMAX_VX (max, 128, _Float16, >=)
+DEF_MINMAX_VX (max, 256, _Float16, >=)
+DEF_MINMAX_VX (max, 512, _Float16, >=)
+DEF_MINMAX_VX (max, 1024, _Float16, >=)
+DEF_MINMAX_VX (max, 2048, _Float16, >=)
+
+DEF_MINMAX_VX (max, 1, float, >=)
+DEF_MINMAX_VX (max, 2, float, >=)
+DEF_MINMAX_VX (max, 4, float, >=)
+DEF_MINMAX_VX (max, 8, float, >=)
+DEF_MINMAX_VX (max, 16, float, >=)
+DEF_MINMAX_VX (max, 32, float, >=)
+DEF_MINMAX_VX (max, 64, float, >=)
+DEF_MINMAX_VX (max, 128, float, >=)
+DEF_MINMAX_VX (max, 256, float, >=)
+DEF_MINMAX_VX (max, 512, float, >=)
+DEF_MINMAX_VX (max, 1024, float, >=)
+
+DEF_MINMAX_VX (max, 1, double, >=)
+DEF_MINMAX_VX (max, 2, double, >=)
+DEF_MINMAX_VX (max, 4, double, >=)
+DEF_MINMAX_VX (max, 8, double, >=)
+DEF_MINMAX_VX (max, 16, double, >=)
+DEF_MINMAX_VX (max, 32, double, >=)
+DEF_MINMAX_VX (max, 64, double, >=)
+DEF_MINMAX_VX (max, 128, double, >=)
+DEF_MINMAX_VX (max, 256, double, >=)
+DEF_MINMAX_VX (max, 512, double, >=)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
new file mode 100644
index 00000000000..775ddb1d25e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
@@ -0,0 +1,31 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+#include "math.h"
+
+DEF_CALL_VV (max, 1, float, fmaxf)
+DEF_CALL_VV (max, 2, float, fmaxf)
+DEF_CALL_VV (max, 4, float, fmaxf)
+DEF_CALL_VV (max, 8, float, fmaxf)
+DEF_CALL_VV (max, 16, float, fmaxf)
+DEF_CALL_VV (max, 32, float, fmaxf)
+DEF_CALL_VV (max, 64, float, fmaxf)
+DEF_CALL_VV (max, 128, float, fmaxf)
+DEF_CALL_VV (max, 256, float, fmaxf)
+DEF_CALL_VV (max, 512, float, fmaxf)
+DEF_CALL_VV (max, 1024, float, fmaxf)
+
+DEF_CALL_VV (max, 1, double, fmax)
+DEF_CALL_VV (max, 2, double, fmax)
+DEF_CALL_VV (max, 4, double, fmax)
+DEF_CALL_VV (max, 8, double, fmax)
+DEF_CALL_VV (max, 16, double, fmax)
+DEF_CALL_VV (max, 32, double, fmax)
+DEF_CALL_VV (max, 64, double, fmax)
+DEF_CALL_VV (max, 128, double, fmax)
+DEF_CALL_VV (max, 256, double, fmax)
+DEF_CALL_VV (max, 512, double, fmax)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
new file mode 100644
index 00000000000..e082c47b044
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (min, 1, _Float16, <)
+DEF_MINMAX_VV (min, 2, _Float16, <)
+DEF_MINMAX_VV (min, 4, _Float16, <)
+DEF_MINMAX_VV (min, 8, _Float16, <)
+DEF_MINMAX_VV (min, 16, _Float16, <)
+DEF_MINMAX_VV (min, 32, _Float16, <)
+DEF_MINMAX_VV (min, 64, _Float16, <)
+DEF_MINMAX_VV (min, 128, _Float16, <)
+DEF_MINMAX_VV (min, 256, _Float16, <)
+DEF_MINMAX_VV (min, 512, _Float16, <)
+DEF_MINMAX_VV (min, 1024, _Float16, <)
+DEF_MINMAX_VV (min, 2048, _Float16, <)
+
+DEF_MINMAX_VV (min, 1, float, <)
+DEF_MINMAX_VV (min, 2, float, <)
+DEF_MINMAX_VV (min, 4, float, <)
+DEF_MINMAX_VV (min, 8, float, <)
+DEF_MINMAX_VV (min, 16, float, <)
+DEF_MINMAX_VV (min, 32, float, <)
+DEF_MINMAX_VV (min, 64, float, <)
+DEF_MINMAX_VV (min, 128, float, <)
+DEF_MINMAX_VV (min, 256, float, <)
+DEF_MINMAX_VV (min, 512, float, <)
+DEF_MINMAX_VV (min, 1024, float, <)
+
+DEF_MINMAX_VV (min, 1, double, <)
+DEF_MINMAX_VV (min, 2, double, <)
+DEF_MINMAX_VV (min, 4, double, <)
+DEF_MINMAX_VV (min, 8, double, <)
+DEF_MINMAX_VV (min, 16, double, <)
+DEF_MINMAX_VV (min, 32, double, <)
+DEF_MINMAX_VV (min, 64, double, <)
+DEF_MINMAX_VV (min, 128, double, <)
+DEF_MINMAX_VV (min, 256, double, <)
+DEF_MINMAX_VV (min, 512, double, <)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
new file mode 100644
index 00000000000..1b900522750
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (min, 1, _Float16, <)
+DEF_MINMAX_VX (min, 2, _Float16, <)
+DEF_MINMAX_VX (min, 4, _Float16, <)
+DEF_MINMAX_VX (min, 8, _Float16, <)
+DEF_MINMAX_VX (min, 16, _Float16, <)
+DEF_MINMAX_VX (min, 32, _Float16, <)
+DEF_MINMAX_VX (min, 64, _Float16, <)
+DEF_MINMAX_VX (min, 128, _Float16, <)
+DEF_MINMAX_VX (min, 256, _Float16, <)
+DEF_MINMAX_VX (min, 512, _Float16, <)
+DEF_MINMAX_VX (min, 1024, _Float16, <)
+DEF_MINMAX_VX (min, 2048, _Float16, <)
+
+DEF_MINMAX_VX (min, 1, float, <)
+DEF_MINMAX_VX (min, 2, float, <)
+DEF_MINMAX_VX (min, 4, float, <)
+DEF_MINMAX_VX (min, 8, float, <)
+DEF_MINMAX_VX (min, 16, float, <)
+DEF_MINMAX_VX (min, 32, float, <)
+DEF_MINMAX_VX (min, 64, float, <)
+DEF_MINMAX_VX (min, 128, float, <)
+DEF_MINMAX_VX (min, 256, float, <)
+DEF_MINMAX_VX (min, 512, float, <)
+DEF_MINMAX_VX (min, 1024, float, <)
+
+DEF_MINMAX_VX (min, 1, double, <)
+DEF_MINMAX_VX (min, 2, double, <)
+DEF_MINMAX_VX (min, 4, double, <)
+DEF_MINMAX_VX (min, 8, double, <)
+DEF_MINMAX_VX (min, 16, double, <)
+DEF_MINMAX_VX (min, 32, double, <)
+DEF_MINMAX_VX (min, 64, double, <)
+DEF_MINMAX_VX (min, 128, double, <)
+DEF_MINMAX_VX (min, 256, double, <)
+DEF_MINMAX_VX (min, 512, double, <)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
new file mode 100644
index 00000000000..ad05800572f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (min, 1, _Float16, <=)
+DEF_MINMAX_VV (min, 2, _Float16, <=)
+DEF_MINMAX_VV (min, 4, _Float16, <=)
+DEF_MINMAX_VV (min, 8, _Float16, <=)
+DEF_MINMAX_VV (min, 16, _Float16, <=)
+DEF_MINMAX_VV (min, 32, _Float16, <=)
+DEF_MINMAX_VV (min, 64, _Float16, <=)
+DEF_MINMAX_VV (min, 128, _Float16, <=)
+DEF_MINMAX_VV (min, 256, _Float16, <=)
+DEF_MINMAX_VV (min, 512, _Float16, <=)
+DEF_MINMAX_VV (min, 1024, _Float16, <=)
+DEF_MINMAX_VV (min, 2048, _Float16, <=)
+
+DEF_MINMAX_VV (min, 1, float, <=)
+DEF_MINMAX_VV (min, 2, float, <=)
+DEF_MINMAX_VV (min, 4, float, <=)
+DEF_MINMAX_VV (min, 8, float, <=)
+DEF_MINMAX_VV (min, 16, float, <=)
+DEF_MINMAX_VV (min, 32, float, <=)
+DEF_MINMAX_VV (min, 64, float, <=)
+DEF_MINMAX_VV (min, 128, float, <=)
+DEF_MINMAX_VV (min, 256, float, <=)
+DEF_MINMAX_VV (min, 512, float, <=)
+DEF_MINMAX_VV (min, 1024, float, <=)
+
+DEF_MINMAX_VV (min, 1, double, <=)
+DEF_MINMAX_VV (min, 2, double, <=)
+DEF_MINMAX_VV (min, 4, double, <=)
+DEF_MINMAX_VV (min, 8, double, <=)
+DEF_MINMAX_VV (min, 16, double, <=)
+DEF_MINMAX_VV (min, 32, double, <=)
+DEF_MINMAX_VV (min, 64, double, <=)
+DEF_MINMAX_VV (min, 128, double, <=)
+DEF_MINMAX_VV (min, 256, double, <=)
+DEF_MINMAX_VV (min, 512, double, <=)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
new file mode 100644
index 00000000000..5d4109aa3c2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (min, 1, _Float16, <=)
+DEF_MINMAX_VX (min, 2, _Float16, <=)
+DEF_MINMAX_VX (min, 4, _Float16, <=)
+DEF_MINMAX_VX (min, 8, _Float16, <=)
+DEF_MINMAX_VX (min, 16, _Float16, <=)
+DEF_MINMAX_VX (min, 32, _Float16, <=)
+DEF_MINMAX_VX (min, 64, _Float16, <=)
+DEF_MINMAX_VX (min, 128, _Float16, <=)
+DEF_MINMAX_VX (min, 256, _Float16, <=)
+DEF_MINMAX_VX (min, 512, _Float16, <=)
+DEF_MINMAX_VX (min, 1024, _Float16, <=)
+DEF_MINMAX_VX (min, 2048, _Float16, <=)
+
+DEF_MINMAX_VX (min, 1, float, <=)
+DEF_MINMAX_VX (min, 2, float, <=)
+DEF_MINMAX_VX (min, 4, float, <=)
+DEF_MINMAX_VX (min, 8, float, <=)
+DEF_MINMAX_VX (min, 16, float, <=)
+DEF_MINMAX_VX (min, 32, float, <=)
+DEF_MINMAX_VX (min, 64, float, <=)
+DEF_MINMAX_VX (min, 128, float, <=)
+DEF_MINMAX_VX (min, 256, float, <=)
+DEF_MINMAX_VX (min, 512, float, <=)
+DEF_MINMAX_VX (min, 1024, float, <=)
+
+DEF_MINMAX_VX (min, 1, double, <=)
+DEF_MINMAX_VX (min, 2, double, <=)
+DEF_MINMAX_VX (min, 4, double, <=)
+DEF_MINMAX_VX (min, 8, double, <=)
+DEF_MINMAX_VX (min, 16, double, <=)
+DEF_MINMAX_VX (min, 32, double, <=)
+DEF_MINMAX_VX (min, 64, double, <=)
+DEF_MINMAX_VX (min, 128, double, <=)
+DEF_MINMAX_VX (min, 256, double, <=)
+DEF_MINMAX_VX (min, 512, double, <=)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
new file mode 100644
index 00000000000..1e9ff7d5054
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
@@ -0,0 +1,31 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+#include "math.h"
+
+DEF_CALL_VV (min, 1, float, fminf)
+DEF_CALL_VV (min, 2, float, fminf)
+DEF_CALL_VV (min, 4, float, fminf)
+DEF_CALL_VV (min, 8, float, fminf)
+DEF_CALL_VV (min, 16, float, fminf)
+DEF_CALL_VV (min, 32, float, fminf)
+DEF_CALL_VV (min, 64, float, fminf)
+DEF_CALL_VV (min, 128, float, fminf)
+DEF_CALL_VV (min, 256, float, fminf)
+DEF_CALL_VV (min, 512, float, fminf)
+DEF_CALL_VV (min, 1024, float, fminf)
+
+DEF_CALL_VV (min, 1, double, fmin)
+DEF_CALL_VV (min, 2, double, fmin)
+DEF_CALL_VV (min, 4, double, fmin)
+DEF_CALL_VV (min, 8, double, fmin)
+DEF_CALL_VV (min, 16, double, fmin)
+DEF_CALL_VV (min, 32, double, fmin)
+DEF_CALL_VV (min, 64, double, fmin)
+DEF_CALL_VV (min, 128, double, fmin)
+DEF_CALL_VV (min, 256, double, fmin)
+DEF_CALL_VV (min, 512, double, fmin)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
-- 
2.34.1


^ permalink raw reply	[flat|nested] 3+ messages in thread

* Re: [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode
  2023-09-02  8:53 [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode pan2.li
@ 2023-09-02 15:40 ` Kito Cheng
  2023-09-03  0:18   ` Li, Pan2
  0 siblings, 1 reply; 3+ messages in thread
From: Kito Cheng @ 2023-09-02 15:40 UTC (permalink / raw)
  To: pan2.li; +Cc: gcc-patches, juzhe.zhong, yanzhang.wang

[-- Attachment #1: Type: text/plain, Size: 28448 bytes --]

Ok

Pan Li via Gcc-patches <gcc-patches@gcc.gnu.org>於 2023年9月2日 週六,16:54寫道:

> From: Pan Li <pan2.li@intel.com>
>
> This patch would like to allow the VLS mode autovec for the
> floating-point binary operation MAX/MIN.
>
> Given below code example:
>
> test (float *out, float *in1, float *in2)
> {
>   for (int i = 0; i < 128; i++)
>     out[i] = in1[i] > in2[i] ? in1[i] : in2[i];
>     // Or out[i] = fmax (in1[i], in2[i]);
> }
>
> Before this patch:
> test:
>   csrr    a4,vlenb
>   slli    a4,a4,1
>   li      a5,128
>   bleu    a5,a4,.L2
>   mv      a5,a4
> .L2:
>   vsetvli zero,a5,e32,m8,ta,ma
>   vle32.v v16,0(a1)
>   vle32.v v8,0(a2)
>   vsetvli a3,zero,e32,m8,ta,ma
>   vmfgt.vv        v0,v16,v8
>   vmerge.vvm      v8,v8,v16,v0
>   vsetvli zero,a5,e32,m8,ta,ma
>   vse32.v v8,0(a0)
>   ret
>
> After this patch:
> test:
>   li      a5,128
>   vsetvli zero,a5,e32,m1,ta,ma
>   vle32.v v1,0(a1)
>   vle32.v v2,0(a2)
>   vfmax.vv        v1,v1,v2
>   vse32.v v1,0(a0)
>   ret
>
> This MAX/MIN autovec acts on function call like fmaxf/fmax in math.h
> too. And it depends on the option -ffast-math.
>
> Signed-off-by: Pan Li <pan2.li@intel.com>
>
> gcc/ChangeLog:
>
>         * config/riscv/autovec-vls.md (<optab><mode>3): New pattern for
>         fmax/fmin
>         * config/riscv/vector.md: Add VLS modes to vfmax/vfmin.
>
> gcc/testsuite/ChangeLog:
>
>         * gcc.target/riscv/rvv/autovec/vls/def.h: New macros.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c: New
> test.
>         * gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c: New
> test.
> ---
>  gcc/config/riscv/autovec-vls.md               | 23 ++++++++++
>  gcc/config/riscv/vector.md                    | 12 +++---
>  .../gcc.target/riscv/rvv/autovec/vls/def.h    | 16 +++++++
>  .../rvv/autovec/vls/floating-point-max-1.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-max-2.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-max-3.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-max-4.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-max-5.c    | 31 +++++++++++++
>  .../rvv/autovec/vls/floating-point-min-1.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-min-2.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-min-3.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-min-4.c    | 43 +++++++++++++++++++
>  .../rvv/autovec/vls/floating-point-min-5.c    | 31 +++++++++++++
>  13 files changed, 451 insertions(+), 6 deletions(-)
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
>  create mode 100644
> gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
>
> diff --git a/gcc/config/riscv/autovec-vls.md
> b/gcc/config/riscv/autovec-vls.md
> index 4ca640c11e2..7ef29637e33 100644
> --- a/gcc/config/riscv/autovec-vls.md
> +++ b/gcc/config/riscv/autovec-vls.md
> @@ -232,6 +232,29 @@ (define_insn_and_split "<optab><mode>3"
>  [(set_attr "type" "vector")]
>  )
>
> +;;
> -------------------------------------------------------------------------
> +;; Includes:
> +;; - vfmin.vv/vfmax.vv
> +;; - vfmin.vf/vfmax.vf
> +;; - fmax/fmaxf in math.h
> +;;
> -------------------------------------------------------------------------
> +(define_insn_and_split "<optab><mode>3"
> +  [(set (match_operand:VLSF 0 "register_operand")
> +    (any_float_binop_nofrm:VLSF
> +     (match_operand:VLSF 1 "<binop_rhs1_predicate>")
> +     (match_operand:VLSF 2 "<binop_rhs2_predicate>")))]
> +  "TARGET_VECTOR && can_create_pseudo_p ()"
> +  "#"
> +  "&& 1"
> +  [(const_int 0)]
> +{
> +  riscv_vector::emit_vlmax_insn (code_for_pred (<CODE>, <MODE>mode),
> +                                riscv_vector::BINARY_OP, operands);
> +  DONE;
> +}
> +[(set_attr "type" "vector")]
> +)
> +
>  ;;
> -------------------------------------------------------------------------------
>  ;; ---- [INT] Unary operations
>  ;;
> -------------------------------------------------------------------------------
> diff --git a/gcc/config/riscv/vector.md b/gcc/config/riscv/vector.md
> index 6fe750ca8a4..9d7b4bbe1d4 100644
> --- a/gcc/config/riscv/vector.md
> +++ b/gcc/config/riscv/vector.md
> @@ -6050,8 +6050,8 @@ (define_insn "@pred_<optab><mode>"
>         (symbol_ref "riscv_vector::get_frm_mode (operands[9])"))])
>
>  (define_insn "@pred_<optab><mode>"
> -  [(set (match_operand:VF 0 "register_operand"           "=vd, vd, vr,
> vr")
> -       (if_then_else:VF
> +  [(set (match_operand:V_VLSF 0 "register_operand"           "=vd, vd,
> vr, vr")
> +       (if_then_else:V_VLSF
>           (unspec:<VM>
>             [(match_operand:<VM> 1 "vector_mask_operand" " vm, vm,Wc1,Wc1")
>              (match_operand 5 "vector_length_operand"    " rK, rK, rK, rK")
> @@ -6060,10 +6060,10 @@ (define_insn "@pred_<optab><mode>"
>              (match_operand 8 "const_int_operand"        "  i,  i,  i,  i")
>              (reg:SI VL_REGNUM)
>              (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE)
> -         (any_float_binop_nofrm:VF
> -           (match_operand:VF 3 "register_operand"       " vr, vr, vr, vr")
> -           (match_operand:VF 4 "register_operand"       " vr, vr, vr,
> vr"))
> -         (match_operand:VF 2 "vector_merge_operand"     " vu,  0, vu,
> 0")))]
> +         (any_float_binop_nofrm:V_VLSF
> +           (match_operand:V_VLSF 3 "register_operand"       " vr, vr, vr,
> vr")
> +           (match_operand:V_VLSF 4 "register_operand"       " vr, vr, vr,
> vr"))
> +         (match_operand:V_VLSF 2 "vector_merge_operand"     " vu,  0,
> vu,  0")))]
>    "TARGET_VECTOR"
>    "vf<insn>.vv\t%0,%3,%4%p1"
>    [(set_attr "type" "<float_insn_type>")
> diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
> index 476f966c427..2e07e908736 100644
> --- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
> @@ -143,6 +143,14 @@ typedef double v512df __attribute__ ((vector_size
> (4096)));
>        a[i] = b[i] OP c[i] ? b[i] : c[i];
>      \
>    }
>
> +#define DEF_MINMAX_VX(PREFIX, NUM, TYPE, OP)
>      \
> +  void __attribute__ ((noinline, noclone))
>      \
> +  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE c)
>       \
> +  {
>       \
> +    for (int i = 0; i < NUM; ++i)
>       \
> +      a[i] = b[i] OP c ? b[i] : c;
>      \
> +  }
> +
>  #define DEF_OP_VI_7(PREFIX, NUM, TYPE, OP)
>      \
>    void __attribute__ ((noinline, noclone))
>      \
>    PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE
> *restrict c)  \
> @@ -159,6 +167,14 @@ typedef double v512df __attribute__ ((vector_size
> (4096)));
>        a[i] = OP b[i];
>       \
>    }
>
> +#define DEF_CALL_VV(PREFIX, NUM, TYPE, CALL)
>      \
> +  void __attribute__ ((noinline, noclone))
>      \
> +  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE
> *restrict c)  \
> +  {
>       \
> +    for (int i = 0; i < NUM; ++i)
>       \
> +      a[i] = CALL (b[i], c[i]);
>       \
> +  }
> +
>  #define DEF_CONST(TYPE, VAL, NUM)
>       \
>    void const_##TYPE##_##NUM (TYPE *restrict a)
>      \
>    {
>       \
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
> new file mode 100644
> index 00000000000..8d3cd2aa538
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VV (max, 1, _Float16, >)
> +DEF_MINMAX_VV (max, 2, _Float16, >)
> +DEF_MINMAX_VV (max, 4, _Float16, >)
> +DEF_MINMAX_VV (max, 8, _Float16, >)
> +DEF_MINMAX_VV (max, 16, _Float16, >)
> +DEF_MINMAX_VV (max, 32, _Float16, >)
> +DEF_MINMAX_VV (max, 64, _Float16, >)
> +DEF_MINMAX_VV (max, 128, _Float16, >)
> +DEF_MINMAX_VV (max, 256, _Float16, >)
> +DEF_MINMAX_VV (max, 512, _Float16, >)
> +DEF_MINMAX_VV (max, 1024, _Float16, >)
> +DEF_MINMAX_VV (max, 2048, _Float16, >)
> +
> +DEF_MINMAX_VV (max, 1, float, >)
> +DEF_MINMAX_VV (max, 2, float, >)
> +DEF_MINMAX_VV (max, 4, float, >)
> +DEF_MINMAX_VV (max, 8, float, >)
> +DEF_MINMAX_VV (max, 16, float, >)
> +DEF_MINMAX_VV (max, 32, float, >)
> +DEF_MINMAX_VV (max, 64, float, >)
> +DEF_MINMAX_VV (max, 128, float, >)
> +DEF_MINMAX_VV (max, 256, float, >)
> +DEF_MINMAX_VV (max, 512, float, >)
> +DEF_MINMAX_VV (max, 1024, float, >)
> +
> +DEF_MINMAX_VV (max, 1, double, >)
> +DEF_MINMAX_VV (max, 2, double, >)
> +DEF_MINMAX_VV (max, 4, double, >)
> +DEF_MINMAX_VV (max, 8, double, >)
> +DEF_MINMAX_VV (max, 16, double, >)
> +DEF_MINMAX_VV (max, 32, double, >)
> +DEF_MINMAX_VV (max, 64, double, >)
> +DEF_MINMAX_VV (max, 128, double, >)
> +DEF_MINMAX_VV (max, 256, double, >)
> +DEF_MINMAX_VV (max, 512, double, >)
> +
> +/* { dg-final { scan-assembler-times
> {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
> new file mode 100644
> index 00000000000..a13de042041
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VX (max, 1, _Float16, >)
> +DEF_MINMAX_VX (max, 2, _Float16, >)
> +DEF_MINMAX_VX (max, 4, _Float16, >)
> +DEF_MINMAX_VX (max, 8, _Float16, >)
> +DEF_MINMAX_VX (max, 16, _Float16, >)
> +DEF_MINMAX_VX (max, 32, _Float16, >)
> +DEF_MINMAX_VX (max, 64, _Float16, >)
> +DEF_MINMAX_VX (max, 128, _Float16, >)
> +DEF_MINMAX_VX (max, 256, _Float16, >)
> +DEF_MINMAX_VX (max, 512, _Float16, >)
> +DEF_MINMAX_VX (max, 1024, _Float16, >)
> +DEF_MINMAX_VX (max, 2048, _Float16, >)
> +
> +DEF_MINMAX_VX (max, 1, float, >)
> +DEF_MINMAX_VX (max, 2, float, >)
> +DEF_MINMAX_VX (max, 4, float, >)
> +DEF_MINMAX_VX (max, 8, float, >)
> +DEF_MINMAX_VX (max, 16, float, >)
> +DEF_MINMAX_VX (max, 32, float, >)
> +DEF_MINMAX_VX (max, 64, float, >)
> +DEF_MINMAX_VX (max, 128, float, >)
> +DEF_MINMAX_VX (max, 256, float, >)
> +DEF_MINMAX_VX (max, 512, float, >)
> +DEF_MINMAX_VX (max, 1024, float, >)
> +
> +DEF_MINMAX_VX (max, 1, double, >)
> +DEF_MINMAX_VX (max, 2, double, >)
> +DEF_MINMAX_VX (max, 4, double, >)
> +DEF_MINMAX_VX (max, 8, double, >)
> +DEF_MINMAX_VX (max, 16, double, >)
> +DEF_MINMAX_VX (max, 32, double, >)
> +DEF_MINMAX_VX (max, 64, double, >)
> +DEF_MINMAX_VX (max, 128, double, >)
> +DEF_MINMAX_VX (max, 256, double, >)
> +DEF_MINMAX_VX (max, 512, double, >)
> +
> +/* { dg-final { scan-assembler-times
> {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
> new file mode 100644
> index 00000000000..108a883bba5
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VV (max, 1, _Float16, >=)
> +DEF_MINMAX_VV (max, 2, _Float16, >=)
> +DEF_MINMAX_VV (max, 4, _Float16, >=)
> +DEF_MINMAX_VV (max, 8, _Float16, >=)
> +DEF_MINMAX_VV (max, 16, _Float16, >=)
> +DEF_MINMAX_VV (max, 32, _Float16, >=)
> +DEF_MINMAX_VV (max, 64, _Float16, >=)
> +DEF_MINMAX_VV (max, 128, _Float16, >=)
> +DEF_MINMAX_VV (max, 256, _Float16, >=)
> +DEF_MINMAX_VV (max, 512, _Float16, >=)
> +DEF_MINMAX_VV (max, 1024, _Float16, >=)
> +DEF_MINMAX_VV (max, 2048, _Float16, >=)
> +
> +DEF_MINMAX_VV (max, 1, float, >=)
> +DEF_MINMAX_VV (max, 2, float, >=)
> +DEF_MINMAX_VV (max, 4, float, >=)
> +DEF_MINMAX_VV (max, 8, float, >=)
> +DEF_MINMAX_VV (max, 16, float, >=)
> +DEF_MINMAX_VV (max, 32, float, >=)
> +DEF_MINMAX_VV (max, 64, float, >=)
> +DEF_MINMAX_VV (max, 128, float, >=)
> +DEF_MINMAX_VV (max, 256, float, >=)
> +DEF_MINMAX_VV (max, 512, float, >=)
> +DEF_MINMAX_VV (max, 1024, float, >=)
> +
> +DEF_MINMAX_VV (max, 1, double, >=)
> +DEF_MINMAX_VV (max, 2, double, >=)
> +DEF_MINMAX_VV (max, 4, double, >=)
> +DEF_MINMAX_VV (max, 8, double, >=)
> +DEF_MINMAX_VV (max, 16, double, >=)
> +DEF_MINMAX_VV (max, 32, double, >=)
> +DEF_MINMAX_VV (max, 64, double, >=)
> +DEF_MINMAX_VV (max, 128, double, >=)
> +DEF_MINMAX_VV (max, 256, double, >=)
> +DEF_MINMAX_VV (max, 512, double, >=)
> +
> +/* { dg-final { scan-assembler-times
> {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
> new file mode 100644
> index 00000000000..d74801887b6
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VX (max, 1, _Float16, >=)
> +DEF_MINMAX_VX (max, 2, _Float16, >=)
> +DEF_MINMAX_VX (max, 4, _Float16, >=)
> +DEF_MINMAX_VX (max, 8, _Float16, >=)
> +DEF_MINMAX_VX (max, 16, _Float16, >=)
> +DEF_MINMAX_VX (max, 32, _Float16, >=)
> +DEF_MINMAX_VX (max, 64, _Float16, >=)
> +DEF_MINMAX_VX (max, 128, _Float16, >=)
> +DEF_MINMAX_VX (max, 256, _Float16, >=)
> +DEF_MINMAX_VX (max, 512, _Float16, >=)
> +DEF_MINMAX_VX (max, 1024, _Float16, >=)
> +DEF_MINMAX_VX (max, 2048, _Float16, >=)
> +
> +DEF_MINMAX_VX (max, 1, float, >=)
> +DEF_MINMAX_VX (max, 2, float, >=)
> +DEF_MINMAX_VX (max, 4, float, >=)
> +DEF_MINMAX_VX (max, 8, float, >=)
> +DEF_MINMAX_VX (max, 16, float, >=)
> +DEF_MINMAX_VX (max, 32, float, >=)
> +DEF_MINMAX_VX (max, 64, float, >=)
> +DEF_MINMAX_VX (max, 128, float, >=)
> +DEF_MINMAX_VX (max, 256, float, >=)
> +DEF_MINMAX_VX (max, 512, float, >=)
> +DEF_MINMAX_VX (max, 1024, float, >=)
> +
> +DEF_MINMAX_VX (max, 1, double, >=)
> +DEF_MINMAX_VX (max, 2, double, >=)
> +DEF_MINMAX_VX (max, 4, double, >=)
> +DEF_MINMAX_VX (max, 8, double, >=)
> +DEF_MINMAX_VX (max, 16, double, >=)
> +DEF_MINMAX_VX (max, 32, double, >=)
> +DEF_MINMAX_VX (max, 64, double, >=)
> +DEF_MINMAX_VX (max, 128, double, >=)
> +DEF_MINMAX_VX (max, 256, double, >=)
> +DEF_MINMAX_VX (max, 512, double, >=)
> +
> +/* { dg-final { scan-assembler-times
> {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
> new file mode 100644
> index 00000000000..775ddb1d25e
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
> @@ -0,0 +1,31 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +#include "math.h"
> +
> +DEF_CALL_VV (max, 1, float, fmaxf)
> +DEF_CALL_VV (max, 2, float, fmaxf)
> +DEF_CALL_VV (max, 4, float, fmaxf)
> +DEF_CALL_VV (max, 8, float, fmaxf)
> +DEF_CALL_VV (max, 16, float, fmaxf)
> +DEF_CALL_VV (max, 32, float, fmaxf)
> +DEF_CALL_VV (max, 64, float, fmaxf)
> +DEF_CALL_VV (max, 128, float, fmaxf)
> +DEF_CALL_VV (max, 256, float, fmaxf)
> +DEF_CALL_VV (max, 512, float, fmaxf)
> +DEF_CALL_VV (max, 1024, float, fmaxf)
> +
> +DEF_CALL_VV (max, 1, double, fmax)
> +DEF_CALL_VV (max, 2, double, fmax)
> +DEF_CALL_VV (max, 4, double, fmax)
> +DEF_CALL_VV (max, 8, double, fmax)
> +DEF_CALL_VV (max, 16, double, fmax)
> +DEF_CALL_VV (max, 32, double, fmax)
> +DEF_CALL_VV (max, 64, double, fmax)
> +DEF_CALL_VV (max, 128, double, fmax)
> +DEF_CALL_VV (max, 256, double, fmax)
> +DEF_CALL_VV (max, 512, double, fmax)
> +
> +/* { dg-final { scan-assembler-times
> {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
> new file mode 100644
> index 00000000000..e082c47b044
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VV (min, 1, _Float16, <)
> +DEF_MINMAX_VV (min, 2, _Float16, <)
> +DEF_MINMAX_VV (min, 4, _Float16, <)
> +DEF_MINMAX_VV (min, 8, _Float16, <)
> +DEF_MINMAX_VV (min, 16, _Float16, <)
> +DEF_MINMAX_VV (min, 32, _Float16, <)
> +DEF_MINMAX_VV (min, 64, _Float16, <)
> +DEF_MINMAX_VV (min, 128, _Float16, <)
> +DEF_MINMAX_VV (min, 256, _Float16, <)
> +DEF_MINMAX_VV (min, 512, _Float16, <)
> +DEF_MINMAX_VV (min, 1024, _Float16, <)
> +DEF_MINMAX_VV (min, 2048, _Float16, <)
> +
> +DEF_MINMAX_VV (min, 1, float, <)
> +DEF_MINMAX_VV (min, 2, float, <)
> +DEF_MINMAX_VV (min, 4, float, <)
> +DEF_MINMAX_VV (min, 8, float, <)
> +DEF_MINMAX_VV (min, 16, float, <)
> +DEF_MINMAX_VV (min, 32, float, <)
> +DEF_MINMAX_VV (min, 64, float, <)
> +DEF_MINMAX_VV (min, 128, float, <)
> +DEF_MINMAX_VV (min, 256, float, <)
> +DEF_MINMAX_VV (min, 512, float, <)
> +DEF_MINMAX_VV (min, 1024, float, <)
> +
> +DEF_MINMAX_VV (min, 1, double, <)
> +DEF_MINMAX_VV (min, 2, double, <)
> +DEF_MINMAX_VV (min, 4, double, <)
> +DEF_MINMAX_VV (min, 8, double, <)
> +DEF_MINMAX_VV (min, 16, double, <)
> +DEF_MINMAX_VV (min, 32, double, <)
> +DEF_MINMAX_VV (min, 64, double, <)
> +DEF_MINMAX_VV (min, 128, double, <)
> +DEF_MINMAX_VV (min, 256, double, <)
> +DEF_MINMAX_VV (min, 512, double, <)
> +
> +/* { dg-final { scan-assembler-times
> {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
> new file mode 100644
> index 00000000000..1b900522750
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VX (min, 1, _Float16, <)
> +DEF_MINMAX_VX (min, 2, _Float16, <)
> +DEF_MINMAX_VX (min, 4, _Float16, <)
> +DEF_MINMAX_VX (min, 8, _Float16, <)
> +DEF_MINMAX_VX (min, 16, _Float16, <)
> +DEF_MINMAX_VX (min, 32, _Float16, <)
> +DEF_MINMAX_VX (min, 64, _Float16, <)
> +DEF_MINMAX_VX (min, 128, _Float16, <)
> +DEF_MINMAX_VX (min, 256, _Float16, <)
> +DEF_MINMAX_VX (min, 512, _Float16, <)
> +DEF_MINMAX_VX (min, 1024, _Float16, <)
> +DEF_MINMAX_VX (min, 2048, _Float16, <)
> +
> +DEF_MINMAX_VX (min, 1, float, <)
> +DEF_MINMAX_VX (min, 2, float, <)
> +DEF_MINMAX_VX (min, 4, float, <)
> +DEF_MINMAX_VX (min, 8, float, <)
> +DEF_MINMAX_VX (min, 16, float, <)
> +DEF_MINMAX_VX (min, 32, float, <)
> +DEF_MINMAX_VX (min, 64, float, <)
> +DEF_MINMAX_VX (min, 128, float, <)
> +DEF_MINMAX_VX (min, 256, float, <)
> +DEF_MINMAX_VX (min, 512, float, <)
> +DEF_MINMAX_VX (min, 1024, float, <)
> +
> +DEF_MINMAX_VX (min, 1, double, <)
> +DEF_MINMAX_VX (min, 2, double, <)
> +DEF_MINMAX_VX (min, 4, double, <)
> +DEF_MINMAX_VX (min, 8, double, <)
> +DEF_MINMAX_VX (min, 16, double, <)
> +DEF_MINMAX_VX (min, 32, double, <)
> +DEF_MINMAX_VX (min, 64, double, <)
> +DEF_MINMAX_VX (min, 128, double, <)
> +DEF_MINMAX_VX (min, 256, double, <)
> +DEF_MINMAX_VX (min, 512, double, <)
> +
> +/* { dg-final { scan-assembler-times
> {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
> new file mode 100644
> index 00000000000..ad05800572f
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VV (min, 1, _Float16, <=)
> +DEF_MINMAX_VV (min, 2, _Float16, <=)
> +DEF_MINMAX_VV (min, 4, _Float16, <=)
> +DEF_MINMAX_VV (min, 8, _Float16, <=)
> +DEF_MINMAX_VV (min, 16, _Float16, <=)
> +DEF_MINMAX_VV (min, 32, _Float16, <=)
> +DEF_MINMAX_VV (min, 64, _Float16, <=)
> +DEF_MINMAX_VV (min, 128, _Float16, <=)
> +DEF_MINMAX_VV (min, 256, _Float16, <=)
> +DEF_MINMAX_VV (min, 512, _Float16, <=)
> +DEF_MINMAX_VV (min, 1024, _Float16, <=)
> +DEF_MINMAX_VV (min, 2048, _Float16, <=)
> +
> +DEF_MINMAX_VV (min, 1, float, <=)
> +DEF_MINMAX_VV (min, 2, float, <=)
> +DEF_MINMAX_VV (min, 4, float, <=)
> +DEF_MINMAX_VV (min, 8, float, <=)
> +DEF_MINMAX_VV (min, 16, float, <=)
> +DEF_MINMAX_VV (min, 32, float, <=)
> +DEF_MINMAX_VV (min, 64, float, <=)
> +DEF_MINMAX_VV (min, 128, float, <=)
> +DEF_MINMAX_VV (min, 256, float, <=)
> +DEF_MINMAX_VV (min, 512, float, <=)
> +DEF_MINMAX_VV (min, 1024, float, <=)
> +
> +DEF_MINMAX_VV (min, 1, double, <=)
> +DEF_MINMAX_VV (min, 2, double, <=)
> +DEF_MINMAX_VV (min, 4, double, <=)
> +DEF_MINMAX_VV (min, 8, double, <=)
> +DEF_MINMAX_VV (min, 16, double, <=)
> +DEF_MINMAX_VV (min, 32, double, <=)
> +DEF_MINMAX_VV (min, 64, double, <=)
> +DEF_MINMAX_VV (min, 128, double, <=)
> +DEF_MINMAX_VV (min, 256, double, <=)
> +DEF_MINMAX_VV (min, 512, double, <=)
> +
> +/* { dg-final { scan-assembler-times
> {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
> new file mode 100644
> index 00000000000..5d4109aa3c2
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
> @@ -0,0 +1,43 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +
> +DEF_MINMAX_VX (min, 1, _Float16, <=)
> +DEF_MINMAX_VX (min, 2, _Float16, <=)
> +DEF_MINMAX_VX (min, 4, _Float16, <=)
> +DEF_MINMAX_VX (min, 8, _Float16, <=)
> +DEF_MINMAX_VX (min, 16, _Float16, <=)
> +DEF_MINMAX_VX (min, 32, _Float16, <=)
> +DEF_MINMAX_VX (min, 64, _Float16, <=)
> +DEF_MINMAX_VX (min, 128, _Float16, <=)
> +DEF_MINMAX_VX (min, 256, _Float16, <=)
> +DEF_MINMAX_VX (min, 512, _Float16, <=)
> +DEF_MINMAX_VX (min, 1024, _Float16, <=)
> +DEF_MINMAX_VX (min, 2048, _Float16, <=)
> +
> +DEF_MINMAX_VX (min, 1, float, <=)
> +DEF_MINMAX_VX (min, 2, float, <=)
> +DEF_MINMAX_VX (min, 4, float, <=)
> +DEF_MINMAX_VX (min, 8, float, <=)
> +DEF_MINMAX_VX (min, 16, float, <=)
> +DEF_MINMAX_VX (min, 32, float, <=)
> +DEF_MINMAX_VX (min, 64, float, <=)
> +DEF_MINMAX_VX (min, 128, float, <=)
> +DEF_MINMAX_VX (min, 256, float, <=)
> +DEF_MINMAX_VX (min, 512, float, <=)
> +DEF_MINMAX_VX (min, 1024, float, <=)
> +
> +DEF_MINMAX_VX (min, 1, double, <=)
> +DEF_MINMAX_VX (min, 2, double, <=)
> +DEF_MINMAX_VX (min, 4, double, <=)
> +DEF_MINMAX_VX (min, 8, double, <=)
> +DEF_MINMAX_VX (min, 16, double, <=)
> +DEF_MINMAX_VX (min, 32, double, <=)
> +DEF_MINMAX_VX (min, 64, double, <=)
> +DEF_MINMAX_VX (min, 128, double, <=)
> +DEF_MINMAX_VX (min, 256, double, <=)
> +DEF_MINMAX_VX (min, 512, double, <=)
> +
> +/* { dg-final { scan-assembler-times
> {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> diff --git
> a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
> b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
> new file mode 100644
> index 00000000000..1e9ff7d5054
> --- /dev/null
> +++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
> @@ -0,0 +1,31 @@
> +/* { dg-do compile } */
> +/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3
> -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8
> -ffast-math" } */
> +
> +#include "def.h"
> +#include "math.h"
> +
> +DEF_CALL_VV (min, 1, float, fminf)
> +DEF_CALL_VV (min, 2, float, fminf)
> +DEF_CALL_VV (min, 4, float, fminf)
> +DEF_CALL_VV (min, 8, float, fminf)
> +DEF_CALL_VV (min, 16, float, fminf)
> +DEF_CALL_VV (min, 32, float, fminf)
> +DEF_CALL_VV (min, 64, float, fminf)
> +DEF_CALL_VV (min, 128, float, fminf)
> +DEF_CALL_VV (min, 256, float, fminf)
> +DEF_CALL_VV (min, 512, float, fminf)
> +DEF_CALL_VV (min, 1024, float, fminf)
> +
> +DEF_CALL_VV (min, 1, double, fmin)
> +DEF_CALL_VV (min, 2, double, fmin)
> +DEF_CALL_VV (min, 4, double, fmin)
> +DEF_CALL_VV (min, 8, double, fmin)
> +DEF_CALL_VV (min, 16, double, fmin)
> +DEF_CALL_VV (min, 32, double, fmin)
> +DEF_CALL_VV (min, 64, double, fmin)
> +DEF_CALL_VV (min, 128, double, fmin)
> +DEF_CALL_VV (min, 256, double, fmin)
> +DEF_CALL_VV (min, 512, double, fmin)
> +
> +/* { dg-final { scan-assembler-times
> {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
> +/* { dg-final { scan-assembler-not {csrr} } } */
> --
> 2.34.1
>
>

^ permalink raw reply	[flat|nested] 3+ messages in thread

* RE: [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode
  2023-09-02 15:40 ` Kito Cheng
@ 2023-09-03  0:18   ` Li, Pan2
  0 siblings, 0 replies; 3+ messages in thread
From: Li, Pan2 @ 2023-09-03  0:18 UTC (permalink / raw)
  To: Kito Cheng; +Cc: gcc-patches, juzhe.zhong, Wang, Yanzhang

[-- Attachment #1: Type: text/plain, Size: 27933 bytes --]

Committed, thanks Kito.

Pan

From: Kito Cheng <kito.cheng@gmail.com>
Sent: Saturday, September 2, 2023 11:41 PM
To: Li, Pan2 <pan2.li@intel.com>
Cc: gcc-patches@gcc.gnu.org; juzhe.zhong@rivai.ai; Wang, Yanzhang <yanzhang.wang@intel.com>
Subject: Re: [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode

Ok

Pan Li via Gcc-patches <gcc-patches@gcc.gnu.org<mailto:gcc-patches@gcc.gnu.org>>於 2023年9月2日 週六,16:54寫道:
From: Pan Li <pan2.li@intel.com<mailto:pan2.li@intel.com>>

This patch would like to allow the VLS mode autovec for the
floating-point binary operation MAX/MIN.

Given below code example:

test (float *out, float *in1, float *in2)
{
  for (int i = 0; i < 128; i++)
    out[i] = in1[i] > in2[i] ? in1[i] : in2[i];
    // Or out[i] = fmax (in1[i], in2[i]);
}

Before this patch:
test:
  csrr    a4,vlenb
  slli    a4,a4,1
  li      a5,128
  bleu    a5,a4,.L2
  mv      a5,a4
.L2:
  vsetvli zero,a5,e32,m8,ta,ma
  vle32.v v16,0(a1)
  vle32.v v8,0(a2)
  vsetvli a3,zero,e32,m8,ta,ma
  vmfgt.vv        v0,v16,v8
  vmerge.vvm      v8,v8,v16,v0
  vsetvli zero,a5,e32,m8,ta,ma
  vse32.v v8,0(a0)
  ret

After this patch:
test:
  li      a5,128
  vsetvli zero,a5,e32,m1,ta,ma
  vle32.v v1,0(a1)
  vle32.v v2,0(a2)
  vfmax.vv        v1,v1,v2
  vse32.v v1,0(a0)
  ret

This MAX/MIN autovec acts on function call like fmaxf/fmax in math.h
too. And it depends on the option -ffast-math.

Signed-off-by: Pan Li <pan2.li@intel.com<mailto:pan2.li@intel.com>>

gcc/ChangeLog:

        * config/riscv/autovec-vls.md (<optab><mode>3): New pattern for
        fmax/fmin
        * config/riscv/vector.md: Add VLS modes to vfmax/vfmin.

gcc/testsuite/ChangeLog:

        * gcc.target/riscv/rvv/autovec/vls/def.h: New macros.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c: New test.
        * gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c: New test.
---
 gcc/config/riscv/autovec-vls.md               | 23 ++++++++++
 gcc/config/riscv/vector.md                    | 12 +++---
 .../gcc.target/riscv/rvv/autovec/vls/def.h    | 16 +++++++
 .../rvv/autovec/vls/floating-point-max-1.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-2.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-3.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-4.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-max-5.c    | 31 +++++++++++++
 .../rvv/autovec/vls/floating-point-min-1.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-2.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-3.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-4.c    | 43 +++++++++++++++++++
 .../rvv/autovec/vls/floating-point-min-5.c    | 31 +++++++++++++
 13 files changed, 451 insertions(+), 6 deletions(-)
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
 create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c

diff --git a/gcc/config/riscv/autovec-vls.md b/gcc/config/riscv/autovec-vls.md
index 4ca640c11e2..7ef29637e33 100644
--- a/gcc/config/riscv/autovec-vls.md
+++ b/gcc/config/riscv/autovec-vls.md
@@ -232,6 +232,29 @@ (define_insn_and_split "<optab><mode>3"
 [(set_attr "type" "vector")]
 )

+;; -------------------------------------------------------------------------
+;; Includes:
+;; - vfmin.vv/vfmax.vv
+;; - vfmin.vf/vfmax.vf
+;; - fmax/fmaxf in math.h
+;; -------------------------------------------------------------------------
+(define_insn_and_split "<optab><mode>3"
+  [(set (match_operand:VLSF 0 "register_operand")
+    (any_float_binop_nofrm:VLSF
+     (match_operand:VLSF 1 "<binop_rhs1_predicate>")
+     (match_operand:VLSF 2 "<binop_rhs2_predicate>")))]
+  "TARGET_VECTOR && can_create_pseudo_p ()"
+  "#"
+  "&& 1"
+  [(const_int 0)]
+{
+  riscv_vector::emit_vlmax_insn (code_for_pred (<CODE>, <MODE>mode),
+                                riscv_vector::BINARY_OP, operands);
+  DONE;
+}
+[(set_attr "type" "vector")]
+)
+
 ;; -------------------------------------------------------------------------------
 ;; ---- [INT] Unary operations
 ;; -------------------------------------------------------------------------------
diff --git a/gcc/config/riscv/vector.md b/gcc/config/riscv/vector.md
index 6fe750ca8a4..9d7b4bbe1d4 100644
--- a/gcc/config/riscv/vector.md
+++ b/gcc/config/riscv/vector.md
@@ -6050,8 +6050,8 @@ (define_insn "@pred_<optab><mode>"
        (symbol_ref "riscv_vector::get_frm_mode (operands[9])"))])

 (define_insn "@pred_<optab><mode>"
-  [(set (match_operand:VF 0 "register_operand"           "=vd, vd, vr, vr")
-       (if_then_else:VF
+  [(set (match_operand:V_VLSF 0 "register_operand"           "=vd, vd, vr, vr")
+       (if_then_else:V_VLSF
          (unspec:<VM>
            [(match_operand:<VM> 1 "vector_mask_operand" " vm, vm,Wc1,Wc1")
             (match_operand 5 "vector_length_operand"    " rK, rK, rK, rK")
@@ -6060,10 +6060,10 @@ (define_insn "@pred_<optab><mode>"
             (match_operand 8 "const_int_operand"        "  i,  i,  i,  i")
             (reg:SI VL_REGNUM)
             (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE)
-         (any_float_binop_nofrm:VF
-           (match_operand:VF 3 "register_operand"       " vr, vr, vr, vr")
-           (match_operand:VF 4 "register_operand"       " vr, vr, vr, vr"))
-         (match_operand:VF 2 "vector_merge_operand"     " vu,  0, vu,  0")))]
+         (any_float_binop_nofrm:V_VLSF
+           (match_operand:V_VLSF 3 "register_operand"       " vr, vr, vr, vr")
+           (match_operand:V_VLSF 4 "register_operand"       " vr, vr, vr, vr"))
+         (match_operand:V_VLSF 2 "vector_merge_operand"     " vu,  0, vu,  0")))]
   "TARGET_VECTOR"
   "vf<insn>.vv\t%0,%3,%4%p1"
   [(set_attr "type" "<float_insn_type>")
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
index 476f966c427..2e07e908736 100644
--- a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/def.h
@@ -143,6 +143,14 @@ typedef double v512df __attribute__ ((vector_size (4096)));
       a[i] = b[i] OP c[i] ? b[i] : c[i];                                       \
   }

+#define DEF_MINMAX_VX(PREFIX, NUM, TYPE, OP)                                   \
+  void __attribute__ ((noinline, noclone))                                     \
+  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE c)            \
+  {                                                                            \
+    for (int i = 0; i < NUM; ++i)                                              \
+      a[i] = b[i] OP c ? b[i] : c;                                             \
+  }
+
 #define DEF_OP_VI_7(PREFIX, NUM, TYPE, OP)                                     \
   void __attribute__ ((noinline, noclone))                                     \
   PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c)  \
@@ -159,6 +167,14 @@ typedef double v512df __attribute__ ((vector_size (4096)));
       a[i] = OP b[i];                                                          \
   }

+#define DEF_CALL_VV(PREFIX, NUM, TYPE, CALL)                                   \
+  void __attribute__ ((noinline, noclone))                                     \
+  PREFIX##_##TYPE##NUM (TYPE *restrict a, TYPE *restrict b, TYPE *restrict c)  \
+  {                                                                            \
+    for (int i = 0; i < NUM; ++i)                                              \
+      a[i] = CALL (b[i], c[i]);                                                \
+  }
+
 #define DEF_CONST(TYPE, VAL, NUM)                                              \
   void const_##TYPE##_##NUM (TYPE *restrict a)                                 \
   {                                                                            \
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
new file mode 100644
index 00000000000..8d3cd2aa538
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-1.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (max, 1, _Float16, >)
+DEF_MINMAX_VV (max, 2, _Float16, >)
+DEF_MINMAX_VV (max, 4, _Float16, >)
+DEF_MINMAX_VV (max, 8, _Float16, >)
+DEF_MINMAX_VV (max, 16, _Float16, >)
+DEF_MINMAX_VV (max, 32, _Float16, >)
+DEF_MINMAX_VV (max, 64, _Float16, >)
+DEF_MINMAX_VV (max, 128, _Float16, >)
+DEF_MINMAX_VV (max, 256, _Float16, >)
+DEF_MINMAX_VV (max, 512, _Float16, >)
+DEF_MINMAX_VV (max, 1024, _Float16, >)
+DEF_MINMAX_VV (max, 2048, _Float16, >)
+
+DEF_MINMAX_VV (max, 1, float, >)
+DEF_MINMAX_VV (max, 2, float, >)
+DEF_MINMAX_VV (max, 4, float, >)
+DEF_MINMAX_VV (max, 8, float, >)
+DEF_MINMAX_VV (max, 16, float, >)
+DEF_MINMAX_VV (max, 32, float, >)
+DEF_MINMAX_VV (max, 64, float, >)
+DEF_MINMAX_VV (max, 128, float, >)
+DEF_MINMAX_VV (max, 256, float, >)
+DEF_MINMAX_VV (max, 512, float, >)
+DEF_MINMAX_VV (max, 1024, float, >)
+
+DEF_MINMAX_VV (max, 1, double, >)
+DEF_MINMAX_VV (max, 2, double, >)
+DEF_MINMAX_VV (max, 4, double, >)
+DEF_MINMAX_VV (max, 8, double, >)
+DEF_MINMAX_VV (max, 16, double, >)
+DEF_MINMAX_VV (max, 32, double, >)
+DEF_MINMAX_VV (max, 64, double, >)
+DEF_MINMAX_VV (max, 128, double, >)
+DEF_MINMAX_VV (max, 256, double, >)
+DEF_MINMAX_VV (max, 512, double, >)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
new file mode 100644
index 00000000000..a13de042041
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-2.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (max, 1, _Float16, >)
+DEF_MINMAX_VX (max, 2, _Float16, >)
+DEF_MINMAX_VX (max, 4, _Float16, >)
+DEF_MINMAX_VX (max, 8, _Float16, >)
+DEF_MINMAX_VX (max, 16, _Float16, >)
+DEF_MINMAX_VX (max, 32, _Float16, >)
+DEF_MINMAX_VX (max, 64, _Float16, >)
+DEF_MINMAX_VX (max, 128, _Float16, >)
+DEF_MINMAX_VX (max, 256, _Float16, >)
+DEF_MINMAX_VX (max, 512, _Float16, >)
+DEF_MINMAX_VX (max, 1024, _Float16, >)
+DEF_MINMAX_VX (max, 2048, _Float16, >)
+
+DEF_MINMAX_VX (max, 1, float, >)
+DEF_MINMAX_VX (max, 2, float, >)
+DEF_MINMAX_VX (max, 4, float, >)
+DEF_MINMAX_VX (max, 8, float, >)
+DEF_MINMAX_VX (max, 16, float, >)
+DEF_MINMAX_VX (max, 32, float, >)
+DEF_MINMAX_VX (max, 64, float, >)
+DEF_MINMAX_VX (max, 128, float, >)
+DEF_MINMAX_VX (max, 256, float, >)
+DEF_MINMAX_VX (max, 512, float, >)
+DEF_MINMAX_VX (max, 1024, float, >)
+
+DEF_MINMAX_VX (max, 1, double, >)
+DEF_MINMAX_VX (max, 2, double, >)
+DEF_MINMAX_VX (max, 4, double, >)
+DEF_MINMAX_VX (max, 8, double, >)
+DEF_MINMAX_VX (max, 16, double, >)
+DEF_MINMAX_VX (max, 32, double, >)
+DEF_MINMAX_VX (max, 64, double, >)
+DEF_MINMAX_VX (max, 128, double, >)
+DEF_MINMAX_VX (max, 256, double, >)
+DEF_MINMAX_VX (max, 512, double, >)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
new file mode 100644
index 00000000000..108a883bba5
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-3.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (max, 1, _Float16, >=)
+DEF_MINMAX_VV (max, 2, _Float16, >=)
+DEF_MINMAX_VV (max, 4, _Float16, >=)
+DEF_MINMAX_VV (max, 8, _Float16, >=)
+DEF_MINMAX_VV (max, 16, _Float16, >=)
+DEF_MINMAX_VV (max, 32, _Float16, >=)
+DEF_MINMAX_VV (max, 64, _Float16, >=)
+DEF_MINMAX_VV (max, 128, _Float16, >=)
+DEF_MINMAX_VV (max, 256, _Float16, >=)
+DEF_MINMAX_VV (max, 512, _Float16, >=)
+DEF_MINMAX_VV (max, 1024, _Float16, >=)
+DEF_MINMAX_VV (max, 2048, _Float16, >=)
+
+DEF_MINMAX_VV (max, 1, float, >=)
+DEF_MINMAX_VV (max, 2, float, >=)
+DEF_MINMAX_VV (max, 4, float, >=)
+DEF_MINMAX_VV (max, 8, float, >=)
+DEF_MINMAX_VV (max, 16, float, >=)
+DEF_MINMAX_VV (max, 32, float, >=)
+DEF_MINMAX_VV (max, 64, float, >=)
+DEF_MINMAX_VV (max, 128, float, >=)
+DEF_MINMAX_VV (max, 256, float, >=)
+DEF_MINMAX_VV (max, 512, float, >=)
+DEF_MINMAX_VV (max, 1024, float, >=)
+
+DEF_MINMAX_VV (max, 1, double, >=)
+DEF_MINMAX_VV (max, 2, double, >=)
+DEF_MINMAX_VV (max, 4, double, >=)
+DEF_MINMAX_VV (max, 8, double, >=)
+DEF_MINMAX_VV (max, 16, double, >=)
+DEF_MINMAX_VV (max, 32, double, >=)
+DEF_MINMAX_VV (max, 64, double, >=)
+DEF_MINMAX_VV (max, 128, double, >=)
+DEF_MINMAX_VV (max, 256, double, >=)
+DEF_MINMAX_VV (max, 512, double, >=)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
new file mode 100644
index 00000000000..d74801887b6
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-4.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (max, 1, _Float16, >=)
+DEF_MINMAX_VX (max, 2, _Float16, >=)
+DEF_MINMAX_VX (max, 4, _Float16, >=)
+DEF_MINMAX_VX (max, 8, _Float16, >=)
+DEF_MINMAX_VX (max, 16, _Float16, >=)
+DEF_MINMAX_VX (max, 32, _Float16, >=)
+DEF_MINMAX_VX (max, 64, _Float16, >=)
+DEF_MINMAX_VX (max, 128, _Float16, >=)
+DEF_MINMAX_VX (max, 256, _Float16, >=)
+DEF_MINMAX_VX (max, 512, _Float16, >=)
+DEF_MINMAX_VX (max, 1024, _Float16, >=)
+DEF_MINMAX_VX (max, 2048, _Float16, >=)
+
+DEF_MINMAX_VX (max, 1, float, >=)
+DEF_MINMAX_VX (max, 2, float, >=)
+DEF_MINMAX_VX (max, 4, float, >=)
+DEF_MINMAX_VX (max, 8, float, >=)
+DEF_MINMAX_VX (max, 16, float, >=)
+DEF_MINMAX_VX (max, 32, float, >=)
+DEF_MINMAX_VX (max, 64, float, >=)
+DEF_MINMAX_VX (max, 128, float, >=)
+DEF_MINMAX_VX (max, 256, float, >=)
+DEF_MINMAX_VX (max, 512, float, >=)
+DEF_MINMAX_VX (max, 1024, float, >=)
+
+DEF_MINMAX_VX (max, 1, double, >=)
+DEF_MINMAX_VX (max, 2, double, >=)
+DEF_MINMAX_VX (max, 4, double, >=)
+DEF_MINMAX_VX (max, 8, double, >=)
+DEF_MINMAX_VX (max, 16, double, >=)
+DEF_MINMAX_VX (max, 32, double, >=)
+DEF_MINMAX_VX (max, 64, double, >=)
+DEF_MINMAX_VX (max, 128, double, >=)
+DEF_MINMAX_VX (max, 256, double, >=)
+DEF_MINMAX_VX (max, 512, double, >=)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
new file mode 100644
index 00000000000..775ddb1d25e
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-max-5.c
@@ -0,0 +1,31 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+#include "math.h"
+
+DEF_CALL_VV (max, 1, float, fmaxf)
+DEF_CALL_VV (max, 2, float, fmaxf)
+DEF_CALL_VV (max, 4, float, fmaxf)
+DEF_CALL_VV (max, 8, float, fmaxf)
+DEF_CALL_VV (max, 16, float, fmaxf)
+DEF_CALL_VV (max, 32, float, fmaxf)
+DEF_CALL_VV (max, 64, float, fmaxf)
+DEF_CALL_VV (max, 128, float, fmaxf)
+DEF_CALL_VV (max, 256, float, fmaxf)
+DEF_CALL_VV (max, 512, float, fmaxf)
+DEF_CALL_VV (max, 1024, float, fmaxf)
+
+DEF_CALL_VV (max, 1, double, fmax)
+DEF_CALL_VV (max, 2, double, fmax)
+DEF_CALL_VV (max, 4, double, fmax)
+DEF_CALL_VV (max, 8, double, fmax)
+DEF_CALL_VV (max, 16, double, fmax)
+DEF_CALL_VV (max, 32, double, fmax)
+DEF_CALL_VV (max, 64, double, fmax)
+DEF_CALL_VV (max, 128, double, fmax)
+DEF_CALL_VV (max, 256, double, fmax)
+DEF_CALL_VV (max, 512, double, fmax)
+
+/* { dg-final { scan-assembler-times {vfmax\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
new file mode 100644
index 00000000000..e082c47b044
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-1.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (min, 1, _Float16, <)
+DEF_MINMAX_VV (min, 2, _Float16, <)
+DEF_MINMAX_VV (min, 4, _Float16, <)
+DEF_MINMAX_VV (min, 8, _Float16, <)
+DEF_MINMAX_VV (min, 16, _Float16, <)
+DEF_MINMAX_VV (min, 32, _Float16, <)
+DEF_MINMAX_VV (min, 64, _Float16, <)
+DEF_MINMAX_VV (min, 128, _Float16, <)
+DEF_MINMAX_VV (min, 256, _Float16, <)
+DEF_MINMAX_VV (min, 512, _Float16, <)
+DEF_MINMAX_VV (min, 1024, _Float16, <)
+DEF_MINMAX_VV (min, 2048, _Float16, <)
+
+DEF_MINMAX_VV (min, 1, float, <)
+DEF_MINMAX_VV (min, 2, float, <)
+DEF_MINMAX_VV (min, 4, float, <)
+DEF_MINMAX_VV (min, 8, float, <)
+DEF_MINMAX_VV (min, 16, float, <)
+DEF_MINMAX_VV (min, 32, float, <)
+DEF_MINMAX_VV (min, 64, float, <)
+DEF_MINMAX_VV (min, 128, float, <)
+DEF_MINMAX_VV (min, 256, float, <)
+DEF_MINMAX_VV (min, 512, float, <)
+DEF_MINMAX_VV (min, 1024, float, <)
+
+DEF_MINMAX_VV (min, 1, double, <)
+DEF_MINMAX_VV (min, 2, double, <)
+DEF_MINMAX_VV (min, 4, double, <)
+DEF_MINMAX_VV (min, 8, double, <)
+DEF_MINMAX_VV (min, 16, double, <)
+DEF_MINMAX_VV (min, 32, double, <)
+DEF_MINMAX_VV (min, 64, double, <)
+DEF_MINMAX_VV (min, 128, double, <)
+DEF_MINMAX_VV (min, 256, double, <)
+DEF_MINMAX_VV (min, 512, double, <)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
new file mode 100644
index 00000000000..1b900522750
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-2.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (min, 1, _Float16, <)
+DEF_MINMAX_VX (min, 2, _Float16, <)
+DEF_MINMAX_VX (min, 4, _Float16, <)
+DEF_MINMAX_VX (min, 8, _Float16, <)
+DEF_MINMAX_VX (min, 16, _Float16, <)
+DEF_MINMAX_VX (min, 32, _Float16, <)
+DEF_MINMAX_VX (min, 64, _Float16, <)
+DEF_MINMAX_VX (min, 128, _Float16, <)
+DEF_MINMAX_VX (min, 256, _Float16, <)
+DEF_MINMAX_VX (min, 512, _Float16, <)
+DEF_MINMAX_VX (min, 1024, _Float16, <)
+DEF_MINMAX_VX (min, 2048, _Float16, <)
+
+DEF_MINMAX_VX (min, 1, float, <)
+DEF_MINMAX_VX (min, 2, float, <)
+DEF_MINMAX_VX (min, 4, float, <)
+DEF_MINMAX_VX (min, 8, float, <)
+DEF_MINMAX_VX (min, 16, float, <)
+DEF_MINMAX_VX (min, 32, float, <)
+DEF_MINMAX_VX (min, 64, float, <)
+DEF_MINMAX_VX (min, 128, float, <)
+DEF_MINMAX_VX (min, 256, float, <)
+DEF_MINMAX_VX (min, 512, float, <)
+DEF_MINMAX_VX (min, 1024, float, <)
+
+DEF_MINMAX_VX (min, 1, double, <)
+DEF_MINMAX_VX (min, 2, double, <)
+DEF_MINMAX_VX (min, 4, double, <)
+DEF_MINMAX_VX (min, 8, double, <)
+DEF_MINMAX_VX (min, 16, double, <)
+DEF_MINMAX_VX (min, 32, double, <)
+DEF_MINMAX_VX (min, 64, double, <)
+DEF_MINMAX_VX (min, 128, double, <)
+DEF_MINMAX_VX (min, 256, double, <)
+DEF_MINMAX_VX (min, 512, double, <)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
new file mode 100644
index 00000000000..ad05800572f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-3.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VV (min, 1, _Float16, <=)
+DEF_MINMAX_VV (min, 2, _Float16, <=)
+DEF_MINMAX_VV (min, 4, _Float16, <=)
+DEF_MINMAX_VV (min, 8, _Float16, <=)
+DEF_MINMAX_VV (min, 16, _Float16, <=)
+DEF_MINMAX_VV (min, 32, _Float16, <=)
+DEF_MINMAX_VV (min, 64, _Float16, <=)
+DEF_MINMAX_VV (min, 128, _Float16, <=)
+DEF_MINMAX_VV (min, 256, _Float16, <=)
+DEF_MINMAX_VV (min, 512, _Float16, <=)
+DEF_MINMAX_VV (min, 1024, _Float16, <=)
+DEF_MINMAX_VV (min, 2048, _Float16, <=)
+
+DEF_MINMAX_VV (min, 1, float, <=)
+DEF_MINMAX_VV (min, 2, float, <=)
+DEF_MINMAX_VV (min, 4, float, <=)
+DEF_MINMAX_VV (min, 8, float, <=)
+DEF_MINMAX_VV (min, 16, float, <=)
+DEF_MINMAX_VV (min, 32, float, <=)
+DEF_MINMAX_VV (min, 64, float, <=)
+DEF_MINMAX_VV (min, 128, float, <=)
+DEF_MINMAX_VV (min, 256, float, <=)
+DEF_MINMAX_VV (min, 512, float, <=)
+DEF_MINMAX_VV (min, 1024, float, <=)
+
+DEF_MINMAX_VV (min, 1, double, <=)
+DEF_MINMAX_VV (min, 2, double, <=)
+DEF_MINMAX_VV (min, 4, double, <=)
+DEF_MINMAX_VV (min, 8, double, <=)
+DEF_MINMAX_VV (min, 16, double, <=)
+DEF_MINMAX_VV (min, 32, double, <=)
+DEF_MINMAX_VV (min, 64, double, <=)
+DEF_MINMAX_VV (min, 128, double, <=)
+DEF_MINMAX_VV (min, 256, double, <=)
+DEF_MINMAX_VV (min, 512, double, <=)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
new file mode 100644
index 00000000000..5d4109aa3c2
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-4.c
@@ -0,0 +1,43 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+
+DEF_MINMAX_VX (min, 1, _Float16, <=)
+DEF_MINMAX_VX (min, 2, _Float16, <=)
+DEF_MINMAX_VX (min, 4, _Float16, <=)
+DEF_MINMAX_VX (min, 8, _Float16, <=)
+DEF_MINMAX_VX (min, 16, _Float16, <=)
+DEF_MINMAX_VX (min, 32, _Float16, <=)
+DEF_MINMAX_VX (min, 64, _Float16, <=)
+DEF_MINMAX_VX (min, 128, _Float16, <=)
+DEF_MINMAX_VX (min, 256, _Float16, <=)
+DEF_MINMAX_VX (min, 512, _Float16, <=)
+DEF_MINMAX_VX (min, 1024, _Float16, <=)
+DEF_MINMAX_VX (min, 2048, _Float16, <=)
+
+DEF_MINMAX_VX (min, 1, float, <=)
+DEF_MINMAX_VX (min, 2, float, <=)
+DEF_MINMAX_VX (min, 4, float, <=)
+DEF_MINMAX_VX (min, 8, float, <=)
+DEF_MINMAX_VX (min, 16, float, <=)
+DEF_MINMAX_VX (min, 32, float, <=)
+DEF_MINMAX_VX (min, 64, float, <=)
+DEF_MINMAX_VX (min, 128, float, <=)
+DEF_MINMAX_VX (min, 256, float, <=)
+DEF_MINMAX_VX (min, 512, float, <=)
+DEF_MINMAX_VX (min, 1024, float, <=)
+
+DEF_MINMAX_VX (min, 1, double, <=)
+DEF_MINMAX_VX (min, 2, double, <=)
+DEF_MINMAX_VX (min, 4, double, <=)
+DEF_MINMAX_VX (min, 8, double, <=)
+DEF_MINMAX_VX (min, 16, double, <=)
+DEF_MINMAX_VX (min, 32, double, <=)
+DEF_MINMAX_VX (min, 64, double, <=)
+DEF_MINMAX_VX (min, 128, double, <=)
+DEF_MINMAX_VX (min, 256, double, <=)
+DEF_MINMAX_VX (min, 512, double, <=)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 30 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
new file mode 100644
index 00000000000..1e9ff7d5054
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/floating-point-min-5.c
@@ -0,0 +1,31 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 -fno-schedule-insns -fno-schedule-insns2 --param=riscv-autovec-lmul=m8 -ffast-math" } */
+
+#include "def.h"
+#include "math.h"
+
+DEF_CALL_VV (min, 1, float, fminf)
+DEF_CALL_VV (min, 2, float, fminf)
+DEF_CALL_VV (min, 4, float, fminf)
+DEF_CALL_VV (min, 8, float, fminf)
+DEF_CALL_VV (min, 16, float, fminf)
+DEF_CALL_VV (min, 32, float, fminf)
+DEF_CALL_VV (min, 64, float, fminf)
+DEF_CALL_VV (min, 128, float, fminf)
+DEF_CALL_VV (min, 256, float, fminf)
+DEF_CALL_VV (min, 512, float, fminf)
+DEF_CALL_VV (min, 1024, float, fminf)
+
+DEF_CALL_VV (min, 1, double, fmin)
+DEF_CALL_VV (min, 2, double, fmin)
+DEF_CALL_VV (min, 4, double, fmin)
+DEF_CALL_VV (min, 8, double, fmin)
+DEF_CALL_VV (min, 16, double, fmin)
+DEF_CALL_VV (min, 32, double, fmin)
+DEF_CALL_VV (min, 64, double, fmin)
+DEF_CALL_VV (min, 128, double, fmin)
+DEF_CALL_VV (min, 256, double, fmin)
+DEF_CALL_VV (min, 512, double, fmin)
+
+/* { dg-final { scan-assembler-times {vfmin\.vv\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+} 19 } } */
+/* { dg-final { scan-assembler-not {csrr} } } */
--
2.34.1

^ permalink raw reply	[flat|nested] 3+ messages in thread

end of thread, other threads:[~2023-09-03  0:18 UTC | newest]

Thread overview: 3+ messages (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-09-02  8:53 [PATCH v1] RISC-V: Support FP MAX/MIN autovec for VLS mode pan2.li
2023-09-02 15:40 ` Kito Cheng
2023-09-03  0:18   ` Li, Pan2

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).