RISC-V: Support FP rint auto-vectorization

This patch would like to support auto-vectorization for the
rint API in math.h. It depends on the -ffast-math option.

When we would like to call rint/rintf like v2 = rint (v1),
we will convert it into below insns (reference the implementation of llvm).

* vfcvt.x.f v3, v1
* vfcvt.f.x v2, v3

However, the floating point value may not need the cvt as above if
its mantissa is zero. Take single precision floating point as example:

Assume we have RTZ rounding mode

  +------------+---------------+-----------------+
  | raw float  | binary layout | after int       |
  +------------+---------------+-----------------+
  | -8388607.5 | 0xcaffffff    | -8388607.0      |
  | 8388607.5  | 0x4affffff    | 8388607.0       |
  | 8388608.0  | 0x4b000000    | 8388608.0       |
  | 8388609.0  | 0x4b000001    | 8388609.0       |
  +------------+---------------+-----------------+

All single floating point >= 8388608.0 will have all zero mantisaa.
We leverage vmflt and mask to filter them out in vector and only do
the cvt on mask.

Befor this patch:
math-rint-1.c:21:1: missed: couldn't vectorize loop
  ...
.L3:
  flw     fa0,0(s0)
  addi    s0,s0,4
  addi    s1,s1,4
  call    rint
  fsw     fa0,-4(s1)
  bne     s0,s2,.L3

After this patch:
  vfabs.v     v2,v1
  vmflt.vf    v0,v2,fa5
  vfcvt.x.f.v v4,v1,v0.t
  vfcvt.f.x.v v2,v4,v0.t
  vfsgnj.vv   v2,v2,v1

Please note VLS mode is also involved in this patch and covered by the
test cases.

gcc/ChangeLog:

	* config/riscv/autovec.md (rint<mode>2): New pattern.
	* config/riscv/riscv-protos.h (expand_vec_rint): New function decl.
	* config/riscv/riscv-v.cc (expand_vec_rint): New function impl.

gcc/testsuite/ChangeLog:

	* gcc.target/riscv/rvv/autovec/unop/math-rint-0.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-rint-1.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-rint-2.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-rint-3.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-rint-run-1.c: New test.
	* gcc.target/riscv/rvv/autovec/unop/math-rint-run-2.c: New test.
	* gcc.target/riscv/rvv/autovec/vls/math-rint-1.c: New test.

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

10 files changed, 261 insertions, 0 deletions

diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
index b47f086..1d2fca6 100644
--- a/gcc/config/riscv/autovec.md
+++ b/gcc/config/riscv/autovec.md
@@ -2241,3 +2241,13 @@
     DONE;
   }
 )
+
+(define_expand "rint<mode>2"
+  [(match_operand:V_VLSF 0 "register_operand")
+   (match_operand:V_VLSF 1 "register_operand")]
+  "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math"
+  {
+    riscv_vector::expand_vec_rint (operands[0], operands[1], <MODE>mode, <VCONVERT>mode);
+    DONE;
+  }
+)
diff --git a/gcc/config/riscv/riscv-protos.h b/gcc/config/riscv/riscv-protos.h
index f87bdef..629adee 100644
--- a/gcc/config/riscv/riscv-protos.h
+++ b/gcc/config/riscv/riscv-protos.h
@@ -462,6 +462,7 @@ void expand_reduction (unsigned, unsigned, rtx *, rtx);
 void expand_vec_ceil (rtx, rtx, machine_mode, machine_mode);
 void expand_vec_floor (rtx, rtx, machine_mode, machine_mode);
 void expand_vec_nearbyint (rtx, rtx, machine_mode, machine_mode);
+void expand_vec_rint (rtx, rtx, machine_mode, machine_mode);
 #endif
 bool sew64_scalar_helper (rtx *, rtx *, rtx, machine_mode,
 			  bool, void (*)(rtx *, rtx));
diff --git a/gcc/config/riscv/riscv-v.cc b/gcc/config/riscv/riscv-v.cc
index 5d3d458..445ed00 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -3698,4 +3698,26 @@ expand_vec_nearbyint (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
   emit_vec_copysign (op_0, op_0, op_1, vec_fp_mode);
 }
 
+void
+expand_vec_rint (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
+		 machine_mode vec_int_mode)
+{
+  /* Step-1: Get the abs float value for mask generation.  */
+  emit_vec_abs (op_0, op_1, vec_fp_mode);
+
+  /* Step-2: Generate the mask on const fp.  */
+  rtx const_fp = get_fp_rounding_coefficient (GET_MODE_INNER (vec_fp_mode));
+  rtx mask = emit_vec_float_cmp_mask (op_0, LT, const_fp, vec_fp_mode);
+
+  /* Step-3: Convert to integer on mask, with dyn rounding (aka rint).  */
+  rtx tmp = gen_reg_rtx (vec_int_mode);
+  emit_vec_cvt_x_f (tmp, op_1, mask, UNARY_OP_TAMU_FRM_DYN, vec_fp_mode);
+
+  /* Step-4: Convert to floating-point on mask for the rint result.  */
+  emit_vec_cvt_f_x (op_0, tmp, mask, UNARY_OP_TAMU_FRM_DYN, vec_fp_mode);
+
+  /* Step-5: Retrieve the sign bit for -0.0.  */
+  emit_vec_copysign (op_0, op_0, op_1, vec_fp_mode);
+}
+
 } // namespace riscv_vector
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-0.c
new file mode 100644
index 0000000..0d44b98
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-0.c
@@ -0,0 +1,18 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh -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__Float16___builtin_rintf16:
+**   ...
+**   vsetvli\s+[atx][0-9]+,\s*zero,\s*e16,\s*m1,\s*ta,\s*mu
+**   vfabs\.v\s+v[0-9]+,\s*v[0-9]+
+**   vmflt\.vf\s+v0,\s*v[0-9]+,\s*[fa]+[0-9]+
+**   vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfcvt\.f\.x\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfsgnj\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+
+**   ...
+*/
+TEST_UNARY_CALL (_Float16, __builtin_rintf16)
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-1.c
new file mode 100644
index 0000000..2ce122a
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-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_float___builtin_rintf:
+**   ...
+**   vsetvli\s+[atx][0-9]+,\s*zero,\s*e32,\s*m1,\s*ta,\s*mu
+**   vfabs\.v\s+v[0-9]+,\s*v[0-9]+
+**   vmflt\.vf\s+v0,\s*v[0-9]+,\s*[fa]+[0-9]+
+**   vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfcvt\.f\.x\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfsgnj\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+
+**   ...
+*/
+TEST_UNARY_CALL (float, __builtin_rintf)
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-2.c
new file mode 100644
index 0000000..e3b911b
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-2.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___builtin_rint:
+**   ...
+**   vsetvli\s+[atx][0-9]+,\s*zero,\s*e64,\s*m1,\s*ta,\s*mu
+**   vfabs\.v\s+v[0-9]+,\s*v[0-9]+
+**   vmflt\.vf\s+v0,\s*v[0-9]+,\s*[fa]+[0-9]+
+**   vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfcvt\.f\.x\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfsgnj\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+
+**   ...
+*/
+TEST_UNARY_CALL (double, __builtin_rint)
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-3.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-3.c
new file mode 100644
index 0000000..541c42c
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-3.c
@@ -0,0 +1,20 @@
+/* { 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___builtin_rintf:
+**   ...
+**   vsetvli\s+[atx][0-9]+,\s*zero,\s*e32,\s*m1,\s*ta,\s*mu
+**   vfabs\.v\s+v[0-9]+,\s*v[0-9]+
+**   vmflt\.vf\s+v0,\s*v[0-9]+,\s*[fa]+[0-9]+
+**   vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfcvt\.f\.x\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t
+**   vfsgnj\.vv\s+v[0-9]+,v[0-9]+,v[0-9]+
+**   ...
+**   vmerge\.vvm\s+v[0-9]+,\s*v[0-9]+,\s*v[0-9]+,\s*v0
+**   ...
+*/
+TEST_COND_UNARY_CALL (float, __builtin_rintf)
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-1.c
new file mode 100644
index 0000000..e96dae4
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-1.c
@@ -0,0 +1,48 @@
+/* { dg-do run { target { riscv_vector } } } */
+/* { dg-additional-options "-std=c99 -O3 -ftree-vectorize -fno-vect-cost-model -ffast-math" } */
+
+#include "test-math.h"
+
+#define ARRAY_SIZE 128
+
+float in[ARRAY_SIZE];
+float out[ARRAY_SIZE];
+float ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL (float, __builtin_rintf)
+TEST_ASSERT (float)
+
+TEST_INIT (float, 1.2, 1.0, 1)
+TEST_INIT (float, -1.2, -1.0, 2)
+TEST_INIT (float, 3.0, 3.0, 3)
+TEST_INIT (float, 8388607.5, 8388607.0, 4)
+TEST_INIT (float, 8388609.0, 8388609.0, 5)
+TEST_INIT (float, 0.0, 0.0, 6)
+TEST_INIT (float, -0.0, -0.0, 7)
+TEST_INIT (float, -8388607.5, -8388607.0, 8)
+TEST_INIT (float, -8388608.0, -8388608.0, 9)
+
+int
+main ()
+{
+  unsigned fflags_before = get_fflags ();
+
+  set_rm (FRM_RTZ);
+
+  RUN_TEST (float, 1, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 2, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 3, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 4, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 5, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 6, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 7, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 8, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (float, 9, __builtin_rintf, in, out, ref, ARRAY_SIZE);
+
+  unsigned fflags_after = get_fflags ();
+
+  if (fflags_before == fflags_after)
+    __builtin_abort ();
+
+  return 0;
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-2.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-2.c
new file mode 100644
index 0000000..f0007d3
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-rint-run-2.c
@@ -0,0 +1,48 @@
+/* { dg-do run { target { riscv_vector } } } */
+/* { 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];
+double out[ARRAY_SIZE];
+double ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL (double, __builtin_rint)
+TEST_ASSERT (double)
+
+TEST_INIT (double, 1.2, 1.0, 1)
+TEST_INIT (double, -1.8, -2.0, 2)
+TEST_INIT (double, 3.0, 3.0, 3)
+TEST_INIT (double, 4503599627370495.5, 4503599627370496.0, 4)
+TEST_INIT (double, 4503599627370497.0, 4503599627370497.0, 5)
+TEST_INIT (double, 0.0, 0.0, 6)
+TEST_INIT (double, -0.0, -0.0, 7)
+TEST_INIT (double, -4503599627370495.5, -4503599627370496.0, 8)
+TEST_INIT (double, -4503599627370496.0, -4503599627370496.0, 9)
+
+int
+main ()
+{
+  unsigned fflags_before = get_fflags ();
+
+  set_rm (FRM_RNE);
+
+  RUN_TEST (double, 1, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 2, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 3, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 4, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 5, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 6, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 7, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 8, __builtin_rint, in, out, ref, ARRAY_SIZE);
+  RUN_TEST (double, 9, __builtin_rint, in, out, ref, ARRAY_SIZE);
+
+  unsigned fflags_after = get_fflags ();
+
+  if (fflags_before == fflags_after)
+    __builtin_abort ();
+
+  return 0;
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-rint-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-rint-1.c
new file mode 100644
index 0000000..cf10d61
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-rint-1.c
@@ -0,0 +1,58 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv64gcv_zvfh_zvl4096b -mabi=lp64d -O3 --param=riscv-autovec-lmul=m8 -ffast-math -fdump-tree-optimized" } */
+
+#include "def.h"
+
+DEF_OP_V (rintf16, 1, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 2, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 4, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 8, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 16, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 32, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 64, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 128, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 256, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 512, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 1024, _Float16, __builtin_rintf16)
+DEF_OP_V (rintf16, 2048, _Float16, __builtin_rintf16)
+
+DEF_OP_V (rintf, 1, float, __builtin_rintf)
+DEF_OP_V (rintf, 2, float, __builtin_rintf)
+DEF_OP_V (rintf, 4, float, __builtin_rintf)
+DEF_OP_V (rintf, 8, float, __builtin_rintf)
+DEF_OP_V (rintf, 16, float, __builtin_rintf)
+DEF_OP_V (rintf, 32, float, __builtin_rintf)
+DEF_OP_V (rintf, 64, float, __builtin_rintf)
+DEF_OP_V (rintf, 128, float, __builtin_rintf)
+DEF_OP_V (rintf, 256, float, __builtin_rintf)
+DEF_OP_V (rintf, 512, float, __builtin_rintf)
+DEF_OP_V (rintf, 1024, float, __builtin_rintf)
+
+DEF_OP_V (rint, 1, double, __builtin_rint)
+DEF_OP_V (rint, 2, double, __builtin_rint)
+DEF_OP_V (rint, 4, double, __builtin_rint)
+DEF_OP_V (rint, 8, double, __builtin_rint)
+DEF_OP_V (rint, 16, double, __builtin_rint)
+DEF_OP_V (rint, 32, double, __builtin_rint)
+DEF_OP_V (rint, 64, double, __builtin_rint)
+DEF_OP_V (rint, 128, double, __builtin_rint)
+DEF_OP_V (rint, 256, double, __builtin_rint)
+DEF_OP_V (rint, 512, double, __builtin_rint)
+
+/* { 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-tree-dump-not "1024,1024" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "2048,2048" "optimized" } } */
+/* { dg-final { scan-tree-dump-not "4096,4096" "optimized" } } */
+/* { dg-final { scan-assembler-not {frflags\s+[atx][0-9]+} } } */
+/* { dg-final { scan-assembler-not {fsflags\s+[atx][0-9]+} } } */
+/* { dg-final { scan-assembler-times {vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t} 30 } } */
+/* { dg-final { scan-assembler-times {vfcvt\.f\.x\.v\s+v[0-9]+,\s*v[0-9]+,\s*v0\.t} 30 } } */