RISC-V: Support FP floor to i/l/ll diff size autovec

This patch would like to support the FP below API auto vectorization
with different type size

+----------+-----------+----------+
| API      | RV64      | RV32     |
+----------+-----------+----------+
| ifloor   | DF => SI  | DF => SI |
| ifloorf  | -         | -        |
| lfloor   | -         | DF => SI |
| lfloorf  | SF => DI  | -        |
| llfloor  | -         | -        |
| llfloorf | SF => DI  | SF => DI |
+----------+-----------+----------+

Given below code:
void
test_lfloorf (long *out, float *in, unsigned count)
{
  for (unsigned i = 0; i < count; i++)
    out[i] = __builtin_lceilf (in[i]);
}

Before this patch:
.L3:
  flw      fa0,0(s0)
  addi     s0,s0,4
  addi     s1,s1,8
  call     floorf
  fcvt.l.s a5,fa0,rtz
  sd       a5,-8(s1)
  bne      s2,s0,.L3

After this patch:
  fsrmi        2  // RDN mode
.L3:
  vsetvli      a5,a2,e32,mf2,ta,ma
  vle32.v      v2,0(a1)
  slli         a3,a5,2
  slli         a4,a5,3
  vfwcvt.x.f.v v1,v2
  sub          a2,a2,a5
  vse64.v      v1,0(a0)
  add          a1,a1,a3
  add          a0,a0,a4
  bne          a2,zero,.L3

Unfortunately, the HF mode is not include due to it requires
additional middle-end support from internal-fun.def.

gcc/ChangeLog:

	* config/riscv/autovec.md: Remove the size check of lfloor.
	* config/riscv/riscv-v.cc (expand_vec_lfloor): Leverage
	emit_vec_rounding_to_integer for floor.

gcc/testsuite/ChangeLog:

	* gcc.target/riscv/rvv/autovec/unop/math-ifloor-1.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-ifloor-run-1.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-run-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-run-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-llfloorf-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-llfloorf-run-0.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/math-ifloor-1.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/math-lfloor-rv32-0.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/math-lfloorf-rv64-0.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/math-llfloorf-0.c: New test.

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

diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
index b59bb880a45f6c73dd75cc43d9eab123ddf2b9e5..973dc4ac2352dc4dd87de6035de6460f602c4703 100644 (file)
--- a/gcc/config/riscv/autovec.md
+++ b/gcc/config/riscv/autovec.md
@@ -2486,8 +2486,7 @@
 (define_expand "lfloor<mode><v_f2si_convert>2"
   [(match_operand:<V_F2SI_CONVERT>   0 "register_operand")
    (match_operand:V_VLS_F_CONVERT_SI 1 "register_operand")]
-  "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math
-    && known_eq (GET_MODE_SIZE (<MODE>mode), GET_MODE_SIZE (<V_F2SI_CONVERT>mode))"
+  "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math"
   {
     riscv_vector::expand_vec_lfloor (operands[0], operands[1], <MODE>mode, <V_F2SI_CONVERT>mode);
     DONE;
@@ -2497,8 +2496,7 @@
 (define_expand "lfloor<mode><v_f2di_convert>2"
   [(match_operand:<V_F2DI_CONVERT>   0 "register_operand")
    (match_operand:V_VLS_F_CONVERT_DI 1 "register_operand")]
-  "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math
-    && known_eq (GET_MODE_SIZE (<MODE>mode), GET_MODE_SIZE (<V_F2DI_CONVERT>mode))"
+  "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math"
   {
     riscv_vector::expand_vec_lfloor (operands[0], operands[1], <MODE>mode, <V_F2DI_CONVERT>mode);
     DONE;

diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index 65a04c47ddf4a5ab79a803251980eac23f7cddce..c32cd8abe6cfe414c979ac2d4cf96b83803d58c9 100644 (file)
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -4184,12 +4184,10 @@ expand_vec_lceil (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
 
 void
 expand_vec_lfloor (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
-                  machine_mode vec_long_mode)
+                  machine_mode vec_int_mode)
 {
-  gcc_assert (known_eq (GET_MODE_SIZE (vec_fp_mode),
-                       GET_MODE_SIZE (vec_long_mode)));
-
-  emit_vec_cvt_x_f (op_0, op_1, UNARY_OP_FRM_RDN, vec_fp_mode);
+  emit_vec_rounding_to_integer (op_0, op_1, vec_fp_mode, vec_int_mode,
+                               UNARY_OP_FRM_RDN);
 }
 
 /* Vectorize popcount by the Wilkes-Wheeler-Gill algorithm that libgcc uses as

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-1.c
new file mode 100644 (file)
index 0000000..4eb63ae
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-1.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math -fno-schedule-insns -fno-schedule-insns2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+#include "test-math.h"
+
+/*
+** test_double_int___builtin_ifloor:
+**   frrm\s+[atx][0-9]+
+**   ...
+**   fsrmi\s+2
+**   ...
+**   vfncvt\.x\.f\.w\s+v[0-9]+,\s*v[0-9]+
+**   ...
+**   fsrm\s+[atx][0-9]+
+**   ret
+*/
+TEST_UNARY_CALL_CVT (double, int, __builtin_ifloor)

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-run-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-run-1.c
new file mode 100644 (file)
index 0000000..ee68f71
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-ifloor-run-1.c
@@ -0,0 +1,83 @@
+/* { dg-do run { target { riscv_v } } } */
+/* { dg-additional-options "-std=c99 -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math" } */
+
+#include "test-math.h"
+
+#define ARRAY_SIZE 128
+
+double in[ARRAY_SIZE];
+int out[ARRAY_SIZE];
+int ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL_CVT (double, int, __builtin_ifloor)
+TEST_ASSERT (int)
+
+TEST_INIT_CVT (double, 0.0, int, __builtin_ifloor (-0.0), 1)
+TEST_INIT_CVT (double, -0.0, int, __builtin_ifloor (-0.0), 2)
+TEST_INIT_CVT (double, 1.2, int, __builtin_ifloor (1.2), 3)
+TEST_INIT_CVT (double, -1.2, int, __builtin_ifloor (-1.2), 4)
+TEST_INIT_CVT (double, 1.5, int, __builtin_ifloor (1.5), 5)
+TEST_INIT_CVT (double, -1.5, int, __builtin_ifloor (-1.5), 6)
+TEST_INIT_CVT (double, 1.8, int, __builtin_ifloor (1.8), 7)
+TEST_INIT_CVT (double, -1.8, int, __builtin_ifloor (-1.8), 8)
+TEST_INIT_CVT (double, 0.2, int, __builtin_ifloor (0.2), 9)
+TEST_INIT_CVT (double, -0.2, int, __builtin_ifloor (-0.2), 10)
+TEST_INIT_CVT (double, 0.5, int, __builtin_ifloor (0.5), 11)
+TEST_INIT_CVT (double, -0.5, int, __builtin_ifloor (-0.5), 12)
+TEST_INIT_CVT (double, 0.8, int, __builtin_ifloor (0.8), 13)
+TEST_INIT_CVT (double, -0.8, int, __builtin_ifloor (-0.8), 14)
+TEST_INIT_CVT (double, 4.0, int, __builtin_ifloor (4.0), 15)
+TEST_INIT_CVT (double, -4.0, int, __builtin_ifloor (-4.0), 16)
+TEST_INIT_CVT (double, 4503599627370495.5, int, __builtin_ifloor (4503599627370495.5), 17)
+TEST_INIT_CVT (double, 4503599627370497.0, int, __builtin_ifloor (4503599627370497.0), 18)
+TEST_INIT_CVT (double, -4503599627370495.5, int, __builtin_ifloor (-4503599627370495.5), 19)
+TEST_INIT_CVT (double, -4503599627370496.0, int, __builtin_ifloor (-4503599627370496.0), 20)
+TEST_INIT_CVT (double, 2147483647.0, int, __builtin_ifloor (2147483647.0), 21)
+TEST_INIT_CVT (double, -2147483648.0, int, __builtin_ifloor (-2147483648.0), 22)
+TEST_INIT_CVT (double, 2147483647.0000002384185791, int, __builtin_ifloor (2147483647.0000002384185791), 23)
+TEST_INIT_CVT (double, -2147483648.0000004768371582, int, 0x80000000, 24)
+TEST_INIT_CVT (double, 9223372036854774784.0, int, __builtin_ifloor (9223372036854774784.0), 25)
+TEST_INIT_CVT (double, 9223372036854775808.0, int, 0x7fffffff, 26)
+TEST_INIT_CVT (double, -9223372036854775808.0, int, __builtin_ifloor (-9223372036854775808.0), 27)
+TEST_INIT_CVT (double, -9223372036854777856.0, int, 0x80000000, 28)
+TEST_INIT_CVT (double, __builtin_inf (), int, __builtin_ifloor (__builtin_inf ()), 29)
+TEST_INIT_CVT (double, -__builtin_inf (), int, __builtin_ifloor (-__builtin_inf ()), 30)
+TEST_INIT_CVT (double, __builtin_nan (""), int, 0x7fffffff, 31)
+
+int
+main ()
+{
+  RUN_TEST_CVT (double, int, 1, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 2, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 3, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 4, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 5, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 6, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 7, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 8, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 9, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 10, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 11, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 12, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 13, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 14, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 15, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 16, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 17, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 18, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 19, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 20, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 21, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 22, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 23, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 24, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 25, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 26, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 27, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 28, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 29, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 30, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, int, 31, __builtin_ifloor, in, out, ref, ARRAY_SIZE);
+
+  return 0;
+}

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-0.c
new file mode 100644 (file)
index 0000000..29c45f3
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-0.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv32gcv -mabi=ilp32d -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math -fno-schedule-insns -fno-schedule-insns2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+#include "test-math.h"
+
+/*
+** test_double_int___builtin_lfloor:
+**   frrm\s+[atx][0-9]+
+**   ...
+**   fsrmi\s+2
+**   ...
+**   vfncvt\.x\.f\.w\s+v[0-9]+,\s*v[0-9]+
+**   ...
+**   fsrm\s+[atx][0-9]+
+**   ret
+*/
+TEST_UNARY_CALL_CVT (double, int, __builtin_lfloor)

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-run-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-run-0.c
new file mode 100644 (file)
index 0000000..dcecee0
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-rv32-run-0.c
@@ -0,0 +1,83 @@
+/* { dg-do run { target { riscv_v && rv32 } } } */
+/* { dg-additional-options "-std=c99 -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math" } */
+
+#include "test-math.h"
+
+#define ARRAY_SIZE 128
+
+double in[ARRAY_SIZE];
+long out[ARRAY_SIZE];
+long ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL_CVT (double, long, __builtin_lfloor)
+TEST_ASSERT (long)
+
+TEST_INIT_CVT (double, 0.0, long, __builtin_lfloor (-0.0), 1)
+TEST_INIT_CVT (double, -0.0, long, __builtin_lfloor (-0.0), 2)
+TEST_INIT_CVT (double, 1.2, long, __builtin_lfloor (1.2), 3)
+TEST_INIT_CVT (double, -1.2, long, __builtin_lfloor (-1.2), 4)
+TEST_INIT_CVT (double, 1.5, long, __builtin_lfloor (1.5), 5)
+TEST_INIT_CVT (double, -1.5, long, __builtin_lfloor (-1.5), 6)
+TEST_INIT_CVT (double, 1.8, long, __builtin_lfloor (1.8), 7)
+TEST_INIT_CVT (double, -1.8, long, __builtin_lfloor (-1.8), 8)
+TEST_INIT_CVT (double, 0.2, long, __builtin_lfloor (0.2), 9)
+TEST_INIT_CVT (double, -0.2, long, __builtin_lfloor (-0.2), 10)
+TEST_INIT_CVT (double, 0.5, long, __builtin_lfloor (0.5), 11)
+TEST_INIT_CVT (double, -0.5, long, __builtin_lfloor (-0.5), 12)
+TEST_INIT_CVT (double, 0.8, long, __builtin_lfloor (0.8), 13)
+TEST_INIT_CVT (double, -0.8, long, __builtin_lfloor (-0.8), 14)
+TEST_INIT_CVT (double, 4.0, long, __builtin_lfloor (4.0), 15)
+TEST_INIT_CVT (double, -4.0, long, __builtin_lfloor (-4.0), 16)
+TEST_INIT_CVT (double, 4503599627370495.5, long, __builtin_lfloor (4503599627370495.5), 17)
+TEST_INIT_CVT (double, 4503599627370497.0, long, __builtin_lfloor (4503599627370497.0), 18)
+TEST_INIT_CVT (double, -4503599627370495.5, long, __builtin_lfloor (-4503599627370495.5), 19)
+TEST_INIT_CVT (double, -4503599627370496.0, long, __builtin_lfloor (-4503599627370496.0), 20)
+TEST_INIT_CVT (double, 2147483647.0, long, __builtin_lfloor (2147483647.0), 21)
+TEST_INIT_CVT (double, -2147483648.0, long, __builtin_lfloor (-2147483648.0), 22)
+TEST_INIT_CVT (double, 2147483647.0000002384185791, long, __builtin_lfloor (2147483647.0000002384185791), 23)
+TEST_INIT_CVT (double, -2147483648.0000004768371582, long, 0x80000000, 24)
+TEST_INIT_CVT (double, 9223372036854774784.0, long, __builtin_lfloor (9223372036854774784.0), 25)
+TEST_INIT_CVT (double, 9223372036854775808.0, long, 0x7fffffff, 26)
+TEST_INIT_CVT (double, -9223372036854775808.0, long, __builtin_lfloor (-9223372036854775808.0), 27)
+TEST_INIT_CVT (double, -9223372036854777856.0, long, 0x80000000, 28)
+TEST_INIT_CVT (double, __builtin_inf (), long, __builtin_lfloor (__builtin_inf ()), 29)
+TEST_INIT_CVT (double, -__builtin_inf (), long, __builtin_lfloor (-__builtin_inf ()), 30)
+TEST_INIT_CVT (double, __builtin_nan (""), long, 0x7fffffff, 31)
+
+long
+main ()
+{
+  RUN_TEST_CVT (double, long, 1, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 2, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 3, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 4, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 5, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 6, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 7, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 8, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 9, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 10, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 11, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 12, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 13, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 14, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 15, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 16, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 17, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 18, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 19, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 20, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 21, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 22, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 23, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 24, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 25, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 26, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 27, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 28, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 29, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 30, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (double, long, 31, __builtin_lfloor, in, out, ref, ARRAY_SIZE);
+
+  return 0;
+}

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-0.c
new file mode 100644 (file)
index 0000000..5c4a619
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-0.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math -fno-schedule-insns -fno-schedule-insns2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+#include "test-math.h"
+
+/*
+** test_float_long___builtin_lfloorf:
+**   frrm\s+[atx][0-9]+
+**   ...
+**   fsrmi\s+2
+**   ...
+**   vfwcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+
+**   ...
+**   fsrm\s+[atx][0-9]+
+**   ret
+*/
+TEST_UNARY_CALL_CVT (float, long, __builtin_lfloorf)

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-run-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-run-0.c
new file mode 100644 (file)
index 0000000..59af94b
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloorf-rv64-run-0.c
@@ -0,0 +1,84 @@
+/* { dg-do run { target { riscv_v && rv64 } } } */
+/* { dg-additional-options "-std=c99 -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math" } */
+
+#include <stdint-gcc.h>
+#include "test-math.h"
+
+#define ARRAY_SIZE 128
+
+float in[ARRAY_SIZE];
+int64_t out[ARRAY_SIZE];
+int64_t ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL_CVT (float, int64_t, __builtin_lfloorf)
+TEST_ASSERT (int64_t)
+
+TEST_INIT_CVT (float, 0.0, int64_t, __builtin_lfloorf (-0.0), 1)
+TEST_INIT_CVT (float, -0.0, int64_t, __builtin_lfloorf (-0.0), 2)
+TEST_INIT_CVT (float, 1.2, int64_t, __builtin_lfloorf (1.2), 3)
+TEST_INIT_CVT (float, -1.2, int64_t, __builtin_lfloorf (-1.2), 4)
+TEST_INIT_CVT (float, 1.5, int64_t, __builtin_lfloorf (1.5), 5)
+TEST_INIT_CVT (float, -1.5, int64_t, __builtin_lfloorf (-1.5), 6)
+TEST_INIT_CVT (float, 1.8, int64_t, __builtin_lfloorf (1.8), 7)
+TEST_INIT_CVT (float, -1.8, int64_t, __builtin_lfloorf (-1.8), 8)
+TEST_INIT_CVT (float, 0.2, int64_t, __builtin_lfloorf (0.2), 9)
+TEST_INIT_CVT (float, -0.2, int64_t, __builtin_lfloorf (-0.2), 10)
+TEST_INIT_CVT (float, 0.5, int64_t, __builtin_lfloorf (0.5), 11)
+TEST_INIT_CVT (float, -0.5, int64_t, __builtin_lfloorf (-0.5), 12)
+TEST_INIT_CVT (float, 0.8, int64_t, __builtin_lfloorf (0.8), 13)
+TEST_INIT_CVT (float, -0.8, int64_t, __builtin_lfloorf (-0.8), 14)
+TEST_INIT_CVT (float, 4.0, int64_t, __builtin_lfloorf (4.0), 15)
+TEST_INIT_CVT (float, -4.0, int64_t, __builtin_lfloorf (-4.0), 16)
+TEST_INIT_CVT (float, 8388607.5, int64_t, __builtin_lfloorf (8388607.5), 17)
+TEST_INIT_CVT (float, 8388609.0, int64_t, __builtin_lfloorf (8388609.0), 18)
+TEST_INIT_CVT (float, -8388607.5, int64_t, __builtin_lfloorf (-8388607.5), 19)
+TEST_INIT_CVT (float, -8388609.0, int64_t, __builtin_lfloorf (-8388609.0), 20)
+TEST_INIT_CVT (float, 2147483520.0, int64_t, __builtin_lfloorf (2147483520.0), 21)
+TEST_INIT_CVT (float, 2147483648.0, int64_t, __builtin_lfloorf (2147483648.0), 22)
+TEST_INIT_CVT (float, -2147483648.0, int64_t, __builtin_lfloorf (-2147483648.0), 23)
+TEST_INIT_CVT (float, -2147483904.0, int64_t, __builtin_lfloorf (-2147483904.0), 24)
+TEST_INIT_CVT (float, 9223371487098961920.0, int64_t, __builtin_lfloorf (9223371487098961920.0), 25)
+TEST_INIT_CVT (float, 9223372036854775808.0, int64_t, 0x7fffffffffffffff, 26)
+TEST_INIT_CVT (float, -9223372036854775808.0, int64_t, __builtin_lfloorf (-9223372036854775808.0), 27)
+TEST_INIT_CVT (float, -9223373136366403584.0, int64_t, 0x8000000000000000, 28)
+TEST_INIT_CVT (float, __builtin_inf (), int64_t, __builtin_lfloorf (__builtin_inf ()), 29)
+TEST_INIT_CVT (float, -__builtin_inf (), int64_t, __builtin_lfloorf (-__builtin_inf ()), 30)
+TEST_INIT_CVT (float, __builtin_nan (""), int64_t, 0x7fffffffffffffff, 31)
+
+int64_t
+main ()
+{
+  RUN_TEST_CVT (float, int64_t, 1, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 2, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 3, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 4, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 5, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 6, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 7, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 8, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 9, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 10, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 11, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 12, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 13, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 14, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 15, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 16, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 17, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 18, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 19, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 20, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 21, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 22, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 23, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 24, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 25, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 26, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 27, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 28, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 29, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 30, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 31, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+
+  return 0;
+}

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-0.c
new file mode 100644 (file)
index 0000000..562fc82
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-0.c
@@ -0,0 +1,19 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv -mabi=lp64d -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math -fno-schedule-insns -fno-schedule-insns2" } */
+/* { dg-final { check-function-bodies "**" "" } } */
+
+#include <stdint-gcc.h>
+#include "test-math.h"
+
+/*
+** test_float_int64_t___builtin_llfloorf:
+**   frrm\s+[atx][0-9]+
+**   ...
+**   fsrmi\s+2
+**   ...
+**   vfwcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+
+**   ...
+**   fsrm\s+[atx][0-9]+
+**   ret
+*/
+TEST_UNARY_CALL_CVT (float, int64_t, __builtin_llfloorf)

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-run-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-run-0.c
new file mode 100644 (file)
index 0000000..da68fd1
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-llfloorf-run-0.c
@@ -0,0 +1,84 @@
+/* { dg-do run { target { riscv_v } } } */
+/* { dg-additional-options "-std=c99 -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math" } */
+
+#include <stdint-gcc.h>
+#include "test-math.h"
+
+#define ARRAY_SIZE 128
+
+float in[ARRAY_SIZE];
+int64_t out[ARRAY_SIZE];
+int64_t ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL_CVT (float, int64_t, __builtin_llfloorf)
+TEST_ASSERT (int64_t)
+
+TEST_INIT_CVT (float, 0.0, int64_t, __builtin_llfloorf (-0.0), 1)
+TEST_INIT_CVT (float, -0.0, int64_t, __builtin_llfloorf (-0.0), 2)
+TEST_INIT_CVT (float, 1.2, int64_t, __builtin_llfloorf (1.2), 3)
+TEST_INIT_CVT (float, -1.2, int64_t, __builtin_llfloorf (-1.2), 4)
+TEST_INIT_CVT (float, 1.5, int64_t, __builtin_llfloorf (1.5), 5)
+TEST_INIT_CVT (float, -1.5, int64_t, __builtin_llfloorf (-1.5), 6)
+TEST_INIT_CVT (float, 1.8, int64_t, __builtin_llfloorf (1.8), 7)
+TEST_INIT_CVT (float, -1.8, int64_t, __builtin_llfloorf (-1.8), 8)
+TEST_INIT_CVT (float, 0.2, int64_t, __builtin_llfloorf (0.2), 9)
+TEST_INIT_CVT (float, -0.2, int64_t, __builtin_llfloorf (-0.2), 10)
+TEST_INIT_CVT (float, 0.5, int64_t, __builtin_llfloorf (0.5), 11)
+TEST_INIT_CVT (float, -0.5, int64_t, __builtin_llfloorf (-0.5), 12)
+TEST_INIT_CVT (float, 0.8, int64_t, __builtin_llfloorf (0.8), 13)
+TEST_INIT_CVT (float, -0.8, int64_t, __builtin_llfloorf (-0.8), 14)
+TEST_INIT_CVT (float, 4.0, int64_t, __builtin_llfloorf (4.0), 15)
+TEST_INIT_CVT (float, -4.0, int64_t, __builtin_llfloorf (-4.0), 16)
+TEST_INIT_CVT (float, 8388607.5, int64_t, __builtin_llfloorf (8388607.5), 17)
+TEST_INIT_CVT (float, 8388609.0, int64_t, __builtin_llfloorf (8388609.0), 18)
+TEST_INIT_CVT (float, -8388607.5, int64_t, __builtin_llfloorf (-8388607.5), 19)
+TEST_INIT_CVT (float, -8388609.0, int64_t, __builtin_llfloorf (-8388609.0), 20)
+TEST_INIT_CVT (float, 2147483520.0, int64_t, __builtin_llfloorf (2147483520.0), 21)
+TEST_INIT_CVT (float, 2147483648.0, int64_t, __builtin_llfloorf (2147483648.0), 22)
+TEST_INIT_CVT (float, -2147483648.0, int64_t, __builtin_llfloorf (-2147483648.0), 23)
+TEST_INIT_CVT (float, -2147483904.0, int64_t, __builtin_llfloorf (-2147483904.0), 24)
+TEST_INIT_CVT (float, 9223371487098961920.0, int64_t, __builtin_llfloorf (9223371487098961920.0), 25)
+TEST_INIT_CVT (float, 9223372036854775808.0, int64_t, 0x7fffffffffffffff, 26)
+TEST_INIT_CVT (float, -9223372036854775808.0, int64_t, __builtin_llfloorf (-9223372036854775808.0), 27)
+TEST_INIT_CVT (float, -9223373136366403584.0, int64_t, 0x8000000000000000, 28)
+TEST_INIT_CVT (float, __builtin_inf (), int64_t, __builtin_llfloorf (__builtin_inf ()), 29)
+TEST_INIT_CVT (float, -__builtin_inf (), int64_t, __builtin_llfloorf (-__builtin_inf ()), 30)
+TEST_INIT_CVT (float, __builtin_nan (""), int64_t, 0x7fffffffffffffff, 31)
+
+int64_t
+main ()
+{
+  RUN_TEST_CVT (float, int64_t, 1, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 2, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 3, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 4, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 5, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 6, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 7, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 8, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 9, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 10, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 11, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 12, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 13, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 14, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 15, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 16, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 17, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 18, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 19, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 20, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 21, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 22, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 23, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 24, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 25, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 26, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 27, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 28, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 29, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 30, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST_CVT (float, int64_t, 31, __builtin_llfloorf, in, out, ref, ARRAY_SIZE);
+
+  return 0;
+}

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-ifloor-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-ifloor-1.c
new file mode 100644 (file)
index 0000000..b6a2122
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-ifloor-1.c
@@ -0,0 +1,27 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -ffast-math -fdump-tree-optimized" } */
+
+#include "def.h"
+
+DEF_OP_V_CVT (ifloor, 1, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 2, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 4, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 8, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 16, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 32, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 64, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 128, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 256, double, int, __builtin_ifloor)
+DEF_OP_V_CVT (ifloor, 512, double, int, __builtin_ifloor)
+
+/* { dg-final { scan-assembler-not {csrr} } } */
+/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */
+/* { dg-final { scan-assembler-times {vfncvt\.x\.f\.w\s+v[0-9]+,\s*v[0-9]+} 9 } } */

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-rv32-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-rv32-0.c
new file mode 100644 (file)
index 0000000..f5a8311
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-rv32-0.c
@@ -0,0 +1,27 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv32gcv_zvl4096b -mabi=ilp32d -O3 --param=riscv-autovec-lmul=m8 -ffast-math -fdump-tree-optimized" } */
+
+#include "def.h"
+
+DEF_OP_V_CVT (lfloor, 1, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 2, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 4, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 8, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 16, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 32, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 64, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 128, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 256, double, int, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 512, double, int, __builtin_lfloor)
+
+/* { dg-final { scan-assembler-not {csrr} } } */
+/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */
+/* { dg-final { scan-assembler-times {vfncvt\.x\.f\.w\s+v[0-9]+,\s*v[0-9]+} 9 } } */

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloorf-rv64-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloorf-rv64-0.c
new file mode 100644 (file)
index 0000000..08b1b7f
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloorf-rv64-0.c
@@ -0,0 +1,27 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -ffast-math -fdump-tree-optimized" } */
+
+#include "def.h"
+
+DEF_OP_V_CVT (lfloorf, 1, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 2, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 4, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 8, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 16, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 32, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 64, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 128, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 256, float, long, __builtin_lfloorf)
+DEF_OP_V_CVT (lfloorf, 512, float, long, __builtin_lfloorf)
+
+/* { dg-final { scan-assembler-not {csrr} } } */
+/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */
+/* { dg-final { scan-assembler-times {vfwcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 9 } } */

diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-llfloorf-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-llfloorf-0.c
new file mode 100644 (file)
index 0000000..b72eaf9
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-llfloorf-0.c
@@ -0,0 +1,27 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -ffast-math -fdump-tree-optimized" } */
+
+#include "def.h"
+
+DEF_OP_V_CVT (llfloorf, 1, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 2, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 4, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 8, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 16, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 32, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 64, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 128, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 256, float, int64_t, __builtin_llfloorf)
+DEF_OP_V_CVT (llfloorf, 512, float, int64_t, __builtin_llfloorf)
+
+/* { dg-final { scan-assembler-not {csrr} } } */
+/* { dg-final { scan-tree-dump-not "1,1" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "2,2" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "4,4" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "16,16" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "32,32" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "64,64" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "128,128" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "256,256" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "512,512" "optimized" } } */
+/* { dg-final { scan-assembler-times {vfwcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 9 } } */