public inbox for gcc-cvs@sourceware.org
help / color / mirror / Atom feed
* [gcc r14-4605] RISC-V: Support FP lfloor/lfloorf auto vectorization
@ 2023-10-13 2:25 Pan Li
0 siblings, 0 replies; only message in thread
From: Pan Li @ 2023-10-13 2:25 UTC (permalink / raw)
To: gcc-cvs
https://gcc.gnu.org/g:8f52040e5f08ae0bbc63ee49c07212f07997a59f
commit r14-4605-g8f52040e5f08ae0bbc63ee49c07212f07997a59f
Author: Pan Li <pan2.li@intel.com>
Date: Fri Oct 13 09:30:55 2023 +0800
RISC-V: Support FP lfloor/lfloorf auto vectorization
This patch would like to support the FP lfloor/lfloorf auto vectorization.
* long lfloor (double) for rv64
* long lfloorf (float) for rv32
Due to the limitation that only the same size of data type are allowed
in the vectorier, the standard name lfloormn2 only act on DF => DI for
rv64, and SF => SI for rv32.
Given we have code like:
void
test_lfloor (long *out, double *in, unsigned count)
{
for (unsigned i = 0; i < count; i++)
out[i] = __builtin_lfloor (in[i]);
}
Before this patch:
.L3:
...
fld fa5,0(a1)
fcvt.l.d a5,fa5,rdn
sd a5,-8(a0)
...
bne a1,a4,.L3
After this patch:
frrm a6
...
fsrmi 2 // RDN
.L3:
...
vsetvli a3,zero,e64,m1,ta,ma
vfcvt.x.f.v v1,v1
vsetvli zero,a2,e64,m1,ta,ma
vse32.v v1,0(a0)
...
bne a2,zero,.L3
...
fsrm a6
The rest part like SF => DI/HF => DI/DF => SI/HF => SI will be covered
by TARGET_VECTORIZE_BUILTIN_VECTORIZED_FUNCTION.
gcc/ChangeLog:
* config/riscv/autovec.md (lfloor<mode><v_i_l_ll_convert>2): New
pattern for lfloor/lfloorf.
* config/riscv/riscv-protos.h (enum insn_type): New enum value.
(expand_vec_lfloor): New func decl for expanding lfloor.
* config/riscv/riscv-v.cc (expand_vec_lfloor): New func impl
for expanding lfloor.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/unop/math-lfloor-0.c: New test.
* gcc.target/riscv/rvv/autovec/unop/math-lfloor-1.c: New test.
* gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-0.c: New test.
* gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-1.c: New test.
* gcc.target/riscv/rvv/autovec/vls/math-lfloor-0.c: New test.
* gcc.target/riscv/rvv/autovec/vls/math-lfloor-1.c: New test.
Signed-off-by: Pan Li <pan2.li@intel.com>
Diff:
---
gcc/config/riscv/autovec.md | 11 ++++
gcc/config/riscv/riscv-protos.h | 2 +
gcc/config/riscv/riscv-v.cc | 10 ++++
.../riscv/rvv/autovec/unop/math-lfloor-0.c | 19 ++++++
.../riscv/rvv/autovec/unop/math-lfloor-1.c | 19 ++++++
.../riscv/rvv/autovec/unop/math-lfloor-run-0.c | 69 ++++++++++++++++++++++
.../riscv/rvv/autovec/unop/math-lfloor-run-1.c | 69 ++++++++++++++++++++++
.../riscv/rvv/autovec/vls/math-lfloor-0.c | 30 ++++++++++
.../riscv/rvv/autovec/vls/math-lfloor-1.c | 30 ++++++++++
9 files changed, 259 insertions(+)
diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
index 267691a0095..c5b1e52cbf9 100644
--- a/gcc/config/riscv/autovec.md
+++ b/gcc/config/riscv/autovec.md
@@ -2242,6 +2242,7 @@
;; - lrint/lrintf
;; - irintf
;; - lceil/lceilf
+;; - lfloor/lfloorf
;; -------------------------------------------------------------------------
(define_expand "ceil<mode>2"
[(match_operand:V_VLSF 0 "register_operand")
@@ -2342,3 +2343,13 @@
DONE;
}
)
+
+(define_expand "lfloor<mode><v_i_l_ll_convert>2"
+ [(match_operand:<V_I_L_LL_CONVERT> 0 "register_operand")
+ (match_operand:V_VLS_FCONVERT_I_L_LL 1 "register_operand")]
+ "TARGET_VECTOR && !flag_trapping_math && !flag_rounding_math"
+ {
+ riscv_vector::expand_vec_lfloor (operands[0], operands[1], <MODE>mode, <V_I_L_LL_CONVERT>mode);
+ DONE;
+ }
+)
diff --git a/gcc/config/riscv/riscv-protos.h b/gcc/config/riscv/riscv-protos.h
index ab65ab19524..49bdcdf2f93 100644
--- a/gcc/config/riscv/riscv-protos.h
+++ b/gcc/config/riscv/riscv-protos.h
@@ -304,6 +304,7 @@ enum insn_type : unsigned int
UNARY_OP_FRM_DYN = UNARY_OP | FRM_DYN_P,
UNARY_OP_FRM_RMM = UNARY_OP | FRM_RMM_P,
UNARY_OP_FRM_RUP = UNARY_OP | FRM_RUP_P,
+ UNARY_OP_FRM_RDN = UNARY_OP | FRM_RDN_P,
UNARY_OP_TAMU_FRM_DYN = UNARY_OP_TAMU | FRM_DYN_P,
UNARY_OP_TAMU_FRM_RUP = UNARY_OP_TAMU | FRM_RUP_P,
UNARY_OP_TAMU_FRM_RDN = UNARY_OP_TAMU | FRM_RDN_P,
@@ -479,6 +480,7 @@ void expand_vec_roundeven (rtx, rtx, machine_mode, machine_mode);
void expand_vec_lrint (rtx, rtx, machine_mode, machine_mode);
void expand_vec_lround (rtx, rtx, machine_mode, machine_mode);
void expand_vec_lceil (rtx, rtx, machine_mode, machine_mode);
+void expand_vec_lfloor (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 b03213dd8ed..21d86c3f917 100644
--- a/gcc/config/riscv/riscv-v.cc
+++ b/gcc/config/riscv/riscv-v.cc
@@ -4142,4 +4142,14 @@ expand_vec_lceil (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
emit_vec_cvt_x_f (op_0, op_1, UNARY_OP_FRM_RUP, vec_fp_mode);
}
+void
+expand_vec_lfloor (rtx op_0, rtx op_1, machine_mode vec_fp_mode,
+ machine_mode vec_long_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);
+}
+
} // namespace riscv_vector
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-0.c
new file mode 100644
index 00000000000..ac2d1722300
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-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 "test-math.h"
+
+/*
+** test_double_long___builtin_lfloor:
+** frrm\s+[atx][0-9]+
+** ...
+** fsrmi\s+2
+** ...
+** vsetvli\s+[atx][0-9]+,\s*zero,\s*e64,\s*m1,\s*ta,\s*ma
+** vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+
+** ...
+** fsrm\s+[atx][0-9]+
+** ret
+*/
+TEST_UNARY_CALL_CVT (double, long, __builtin_lfloor)
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-1.c
new file mode 100644
index 00000000000..164e97c17d6
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-1.c
@@ -0,0 +1,19 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv32gcv -mabi=ilp32f -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
+** ...
+** vsetvli\s+[atx][0-9]+,\s*zero,\s*e32,\s*m1,\s*ta,\s*ma
+** vfcvt\.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-lfloor-run-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-0.c
new file mode 100644
index 00000000000..3b710485e88
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-0.c
@@ -0,0 +1,69 @@
+/* { dg-do run { target { riscv_v && rv64 } } } */
+/* { 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, 1.2, long, __builtin_lfloor (1.2), 1)
+TEST_INIT_CVT (double, -1.2, long, __builtin_lfloor (-1.2), 2)
+TEST_INIT_CVT (double, 0.5, long, __builtin_lfloor (0.5), 3)
+TEST_INIT_CVT (double, -0.5, long, __builtin_lfloor (-0.5), 4)
+TEST_INIT_CVT (double, 0.1, long, __builtin_lfloor (0.1), 5)
+TEST_INIT_CVT (double, -0.1, long, __builtin_lfloor (-0.1), 6)
+TEST_INIT_CVT (double, 3.0, long, __builtin_lfloor (3.0), 7)
+TEST_INIT_CVT (double, -3.0, long, __builtin_lfloor (-3.0), 8)
+TEST_INIT_CVT (double, 4503599627370495.5, long, __builtin_lfloor (4503599627370495.5), 9)
+TEST_INIT_CVT (double, 4503599627370497.0, long, __builtin_lfloor (4503599627370497.0), 10)
+TEST_INIT_CVT (double, -4503599627370495.5, long, __builtin_lfloor (-4503599627370495.5), 11)
+TEST_INIT_CVT (double, -4503599627370496.0, long, __builtin_lfloor (-4503599627370496.0), 12)
+TEST_INIT_CVT (double, 0.0, long, __builtin_lfloor (-0.0), 13)
+TEST_INIT_CVT (double, -0.0, long, __builtin_lfloor (-0.0), 14)
+TEST_INIT_CVT (double, 9223372036854774784.0, long, __builtin_lfloor (9223372036854774784.0), 15)
+TEST_INIT_CVT (double, 9223372036854775808.0, long, 0x7fffffffffffffff, 16)
+TEST_INIT_CVT (double, -9223372036854775808.0, long, __builtin_lfloor (-9223372036854775808.0), 17)
+TEST_INIT_CVT (double, -9223372036854777856.0, long, 0x8000000000000000, 18)
+TEST_INIT_CVT (double, __builtin_inf (), long, __builtin_lfloor (__builtin_inf ()), 19)
+TEST_INIT_CVT (double, -__builtin_inf (), long, __builtin_lfloor (-__builtin_inf ()), 20)
+TEST_INIT_CVT (double, __builtin_nan (""), long, 0x7fffffffffffffff, 21)
+
+/*
+ Similar to lround, some reference are hard-code instead of leveraging
+ scalar __builtin_lfloor because the return value of a NaN or an infinity,
+ or the rounded value is too large to be stored in a long is UNSPECIFIED.
+*/
+
+int
+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);
+
+ return 0;
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-1.c
new file mode 100644
index 00000000000..60f50fc8836
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/unop/math-lfloor-run-1.c
@@ -0,0 +1,69 @@
+/* { 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
+
+float in[ARRAY_SIZE];
+long out[ARRAY_SIZE];
+long ref[ARRAY_SIZE];
+
+TEST_UNARY_CALL_CVT (float, long, __builtin_lfloorf)
+TEST_ASSERT (long)
+
+TEST_INIT_CVT (float, 1.2, long, __builtin_lfloorf (1.2), 1)
+TEST_INIT_CVT (float, -1.2, long, __builtin_lfloorf (-1.2), 2)
+TEST_INIT_CVT (float, 0.5, long, __builtin_lfloorf (0.5), 3)
+TEST_INIT_CVT (float, -0.5, long, __builtin_lfloorf (-0.5), 4)
+TEST_INIT_CVT (float, 0.1, long, __builtin_lfloorf (0.1), 5)
+TEST_INIT_CVT (float, -0.1, long, __builtin_lfloorf (-0.1), 6)
+TEST_INIT_CVT (float, 3.0, long, __builtin_lfloorf (3.0), 7)
+TEST_INIT_CVT (float, -3.0, long, __builtin_lfloorf (-3.0), 8)
+TEST_INIT_CVT (float, 8388607.5, long, __builtin_lfloorf (8388607.5), 9)
+TEST_INIT_CVT (float, 8388609.0, long, __builtin_lfloorf (8388609.0), 10)
+TEST_INIT_CVT (float, -8388607.5, long, __builtin_lfloorf (-8388607.5), 11)
+TEST_INIT_CVT (float, -8388609.0, long, __builtin_lfloorf (-8388609.0), 12)
+TEST_INIT_CVT (float, 0.0, long, __builtin_lfloorf (-0.0), 13)
+TEST_INIT_CVT (float, -0.0, long, __builtin_lfloorf (-0.0), 14)
+TEST_INIT_CVT (float, 2147483520.0, long, __builtin_lfloorf (2147483520.0), 15)
+TEST_INIT_CVT (float, 2147483648.0, long, 0x7fffffff, 16)
+TEST_INIT_CVT (float, -2147483648.0, long, __builtin_lfloorf (-2147483648.0), 17)
+TEST_INIT_CVT (float, -2147483904.0, long, 0x80000000, 18)
+TEST_INIT_CVT (float, __builtin_inf (), long, __builtin_lfloorf (__builtin_inff ()), 19)
+TEST_INIT_CVT (float, -__builtin_inf (), long, __builtin_lfloorf (-__builtin_inff ()), 20)
+TEST_INIT_CVT (float, __builtin_nanf (""), long, 0x7fffffff, 21)
+
+/*
+ Similar to lround, some reference are hard-code instead of leveraging
+ scalar __builtin_lfloor because the return value of a NaN or an infinity,
+ or the rounded value is too large to be stored in a long is UNSPECIFIED.
+*/
+
+int
+main ()
+{
+ RUN_TEST_CVT (float, long, 1, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 2, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 3, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 4, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 5, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 6, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 7, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 8, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 9, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 10, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 11, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 12, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 13, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 14, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 15, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 16, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 17, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 18, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 19, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 20, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+ RUN_TEST_CVT (float, long, 21, __builtin_lfloorf, in, out, ref, ARRAY_SIZE);
+
+ return 0;
+}
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-0.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-0.c
new file mode 100644
index 00000000000..fe61e997de4
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-0.c
@@ -0,0 +1,30 @@
+/* { 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_CVT (lfloor, 1, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 2, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 4, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 8, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 16, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 32, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 64, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 128, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 256, double, long, __builtin_lfloor)
+DEF_OP_V_CVT (lfloor, 512, double, long, __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-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-times {vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 9 } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-1.c b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-1.c
new file mode 100644
index 00000000000..a64e5c4d252
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/vls/math-lfloor-1.c
@@ -0,0 +1,30 @@
+/* { dg-do compile } */
+/* { dg-options "-march=rv32gcv_zvl4096b -mabi=ilp32f -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-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-times {vfcvt\.x\.f\.v\s+v[0-9]+,\s*v[0-9]+} 9 } } */
^ permalink raw reply [flat|nested] only message in thread
only message in thread, other threads:[~2023-10-13 2:25 UTC | newest]
Thread overview: (only message) (download: mbox.gz / follow: Atom feed)
-- links below jump to the message on this page --
2023-10-13 2:25 [gcc r14-4605] RISC-V: Support FP lfloor/lfloorf auto vectorization Pan Li
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).