From: Juzhe-Zhong <juzhe.zh...@rivai.ai>
This patch is to enhance vwmul.vv combine optimizations.
Consider this following code:
void
vwadd_int16_t_int8_t (int16_t *__restrict dst, int16_t *__restrict dst2,
int16_t *__restrict dst3, int16_t *__restrict dst4,
int8_t *__restrict a, int8_t *__restrict b,
int8_t *__restrict a2, int8_t *__restrict b2, int n)
{
for (int i = 0; i < n; i++)
{
dst[i] = (int16_t) a[i] * (int16_t) b[i];
dst2[i] = (int16_t) a2[i] * (int16_t) b[i];
dst3[i] = (int16_t) a2[i] * (int16_t) a[i];
dst4[i] = (int16_t) a[i] * (int16_t) b2[i];
}
}
In such complicate case, the operand is not single used, used by multiple
statements.
GCC combine optimization will iterate the combination of the operands.
First round -> combine one of the operand and change vsext + vmul into vwmul.wv
Second round -> combine the other operand and change vwmul.wv into vwmul.vv
Notice when I add a pseudo vwmul.wv pattern, it makes vwmulsu.vv testcase fail
since GCC prefer such pattern order:
(mul: (zero_extend)
(sign_exted))
So change vwmulsu.vv instruction operands order.
gcc/ChangeLog:
* config/riscv/vector.md: Shift zero_extend and sign_extend order.
* config/riscv/autovec-opt.md: New file.
gcc/testsuite/ChangeLog:
* gcc.target/riscv/rvv/autovec/widen/widen-7.c: New test.
* gcc.target/riscv/rvv/autovec/widen/widen-complicate-3.c: New test.
* gcc.target/riscv/rvv/autovec/widen/widen_run-7.c: New test.
---
gcc/config/riscv/autovec-opt.md | 56 +++++++++++++++++++
gcc/config/riscv/vector.md | 9 +--
.../riscv/rvv/autovec/widen/widen-7.c | 27 +++++++++
.../rvv/autovec/widen/widen-complicate-3.c | 32 +++++++++++
.../riscv/rvv/autovec/widen/widen_run-7.c | 34 +++++++++++
5 files changed, 154 insertions(+), 4 deletions(-)
create mode 100644 gcc/config/riscv/autovec-opt.md
create mode 100644 gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-7.c
create mode 100644
gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-complicate-3.c
create mode 100644
gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen_run-7.c
diff --git a/gcc/config/riscv/autovec-opt.md b/gcc/config/riscv/autovec-opt.md
new file mode 100644
index 00000000000..5b7dc9bef8c
--- /dev/null
+++ b/gcc/config/riscv/autovec-opt.md
@@ -0,0 +1,56 @@
+;; Machine description for optimization of RVV auto-vectorization.
+;; Copyright (C) 2023 Free Software Foundation, Inc.
+;; Contributed by Juzhe Zhong (juzhe.zh...@rivai.ai), RiVAI Technologies Ltd.
+
+;; This file is part of GCC.
+
+;; GCC is free software; you can redistribute it and/or modify
+;; it under the terms of the GNU General Public License as published by
+;; the Free Software Foundation; either version 3, or (at your option)
+;; any later version.
+
+;; GCC is distributed in the hope that it will be useful,
+;; but WITHOUT ANY WARRANTY; without even the implied warranty of
+;; MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
+;; GNU General Public License for more details.
+
+;; You should have received a copy of the GNU General Public License
+;; along with GCC; see the file COPYING3. If not see
+;; <http://www.gnu.org/licenses/>.
+
+;; We don't have vwmul.wv instruction like vwadd.wv in RVV.
+;; This pattern is an intermediate RTL IR as a pseudo vwmul.wv to enhance
+;; optimization of instructions combine.
+(define_insn_and_split "@pred_single_widen_mul<any_extend:su><mode>"
+ [(set (match_operand:VWEXTI 0 "register_operand" "=&vr,&vr")
+ (if_then_else:VWEXTI
+ (unspec:<VM>
+ [(match_operand:<VM> 1 "vector_mask_operand"
"vmWc1,vmWc1")
+ (match_operand 5 "vector_length_operand" " rK,
rK")
+ (match_operand 6 "const_int_operand" " i,
i")
+ (match_operand 7 "const_int_operand" " i,
i")
+ (match_operand 8 "const_int_operand" " i,
i")
+ (reg:SI VL_REGNUM)
+ (reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE)
+ (mult:VWEXTI
+ (any_extend:VWEXTI
+ (match_operand:<V_DOUBLE_TRUNC> 4 "register_operand" " vr,
vr"))
+ (match_operand:VWEXTI 3 "register_operand" " vr,
vr"))
+ (match_operand:VWEXTI 2 "vector_merge_operand" " vu,
0")))]
+ "TARGET_VECTOR"
+ "#"
+ "&& can_create_pseudo_p ()"
+ [(const_int 0)]
+ {
+ insn_code icode = code_for_pred_vf2 (<CODE>, <MODE>mode);
+ rtx tmp = gen_reg_rtx (<MODE>mode);
+ rtx ops[] = {tmp, operands[4]};
+ riscv_vector::emit_vlmax_insn (icode, riscv_vector::RVV_UNOP, ops);
+
+ emit_insn (gen_pred (MULT, <MODE>mode, operands[0], operands[1],
operands[2],
+ operands[3], tmp, operands[5], operands[6],
+ operands[7], operands[8]));
+ DONE;
+ }
+ [(set_attr "type" "viwmul")
+ (set_attr "mode" "<MODE>")])
diff --git a/gcc/config/riscv/vector.md b/gcc/config/riscv/vector.md
index c74dce89db6..bb333a9b410 100644
--- a/gcc/config/riscv/vector.md
+++ b/gcc/config/riscv/vector.md
@@ -24,7 +24,7 @@
;;
;; - Intrinsics (https://github.com/riscv/rvv-intrinsic-doc)
;; - Auto-vectorization (autovec.md)
-;; - Combine optimization (TBD)
+;; - Optimization (autovec-opt.md)
(include "vector-iterators.md")
@@ -3207,10 +3207,10 @@
(reg:SI VL_REGNUM)
(reg:SI VTYPE_REGNUM)] UNSPEC_VPREDICATE)
(mult:VWEXTI
- (sign_extend:VWEXTI
- (match_operand:<V_DOUBLE_TRUNC> 3 "register_operand" " vr,
vr"))
(zero_extend:VWEXTI
- (match_operand:<V_DOUBLE_TRUNC> 4 "register_operand" " vr,
vr")))
+ (match_operand:<V_DOUBLE_TRUNC> 4 "register_operand" " vr,
vr"))
+ (sign_extend:VWEXTI
+ (match_operand:<V_DOUBLE_TRUNC> 3 "register_operand" " vr,
vr")))
(match_operand:VWEXTI 2 "vector_merge_operand" " vu,
0")))]
"TARGET_VECTOR"
"vwmulsu.vv\t%0,%3,%4%p1"
@@ -8422,3 +8422,4 @@
)
(include "autovec.md")
+(include "autovec-opt.md")
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-7.c
b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-7.c
new file mode 100644
index 00000000000..cc43d9ba3fe
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-7.c
@@ -0,0 +1,27 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv -mabi=ilp32d
--param=riscv-autovec-preference=scalable" } */
+
+#include <stdint-gcc.h>
+
+#define TEST_TYPE(TYPE1, TYPE2)
\
+ __attribute__ ((noipa)) void vwmul_##TYPE1_##TYPE2 (TYPE1 *__restrict dst,
\
+ TYPE2 *__restrict a, \
+ TYPE1 *__restrict b, \
+ int n) \
+ {
\
+ for (int i = 0; i < n; i++)
\
+ dst[i] = ((TYPE1) a[i]) * b[i];
\
+ }
+
+#define TEST_ALL()
\
+ TEST_TYPE (int16_t, int8_t)
\
+ TEST_TYPE (uint16_t, uint8_t)
\
+ TEST_TYPE (int32_t, int16_t)
\
+ TEST_TYPE (uint32_t, uint16_t)
\
+ TEST_TYPE (int64_t, int32_t)
\
+ TEST_TYPE (uint64_t, uint32_t)
+
+TEST_ALL ()
+
+/* { dg-final { scan-assembler-times {\tvsext\.vf2} 3 } } */
+/* { dg-final { scan-assembler-times {\tvzext\.vf2} 3 } } */
diff --git
a/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-complicate-3.c
b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-complicate-3.c
new file mode 100644
index 00000000000..e1fd79430c3
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen-complicate-3.c
@@ -0,0 +1,32 @@
+/* { dg-do compile } */
+/* { dg-additional-options "-march=rv32gcv -mabi=ilp32d
--param=riscv-autovec-preference=scalable" } */
+
+#include <stdint-gcc.h>
+
+#define TEST_TYPE(TYPE1, TYPE2)
\
+ __attribute__ ((noipa)) void vwadd_##TYPE1_##TYPE2 (
\
+ TYPE1 *__restrict dst, TYPE1 *__restrict dst2, TYPE1 *__restrict dst3,
\
+ TYPE1 *__restrict dst4, TYPE2 *__restrict a, TYPE2 *__restrict b,
\
+ TYPE2 *__restrict a2, TYPE2 *__restrict b2, int n)
\
+ {
\
+ for (int i = 0; i < n; i++)
\
+ {
\
+ dst[i] = (TYPE1) a[i] * (TYPE1) b[i]; \
+ dst2[i] = (TYPE1) a2[i] * (TYPE1) b[i]; \
+ dst3[i] = (TYPE1) a2[i] * (TYPE1) a[i]; \
+ dst4[i] = (TYPE1) a[i] * (TYPE1) b2[i]; \
+ }
\
+ }
+
+#define TEST_ALL()
\
+ TEST_TYPE (int16_t, int8_t)
\
+ TEST_TYPE (uint16_t, uint8_t)
\
+ TEST_TYPE (int32_t, int16_t)
\
+ TEST_TYPE (uint32_t, uint16_t)
\
+ TEST_TYPE (int64_t, int32_t)
\
+ TEST_TYPE (uint64_t, uint32_t)
+
+TEST_ALL ()
+
+/* { dg-final { scan-assembler-times {\tvwmul\.vv} 12 } } */
+/* { dg-final { scan-assembler-times {\tvwmulu\.vv} 12 } } */
diff --git a/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen_run-7.c
b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen_run-7.c
new file mode 100644
index 00000000000..4abddd5d718
--- /dev/null
+++ b/gcc/testsuite/gcc.target/riscv/rvv/autovec/widen/widen_run-7.c
@@ -0,0 +1,34 @@
+/* { dg-do run { target { riscv_vector } } } */
+/* { dg-additional-options "--param=riscv-autovec-preference=scalable" } */
+
+#include <assert.h>
+#include "widen-7.c"
+
+#define SZ 512
+
+#define RUN(TYPE1, TYPE2, LIMIT)
\
+ TYPE2 a##TYPE2[SZ];
\
+ TYPE1 b##TYPE1[SZ];
\
+ TYPE1 dst##TYPE1[SZ];
\
+ for (int i = 0; i < SZ; i++)
\
+ {
\
+ a##TYPE2[i] = LIMIT + i % LIMIT;
\
+ b##TYPE1[i] = LIMIT + i & LIMIT;
\
+ }
\
+ vwmul_##TYPE1_##TYPE2 (dst##TYPE1, a##TYPE2, b##TYPE1, SZ);
\
+ for (int i = 0; i < SZ; i++)
\
+ assert (dst##TYPE1[i] == (((TYPE1) a##TYPE2[i]) * b##TYPE1[i]));
+
+#define RUN_ALL()
\
+ RUN (int16_t, int8_t, -128)
\
+ RUN (uint16_t, uint8_t, 255)
\
+ RUN (int32_t, int16_t, -32768)
\
+ RUN (uint32_t, uint16_t, 65535)
\
+ RUN (int64_t, int32_t, -2147483648)
\
+ RUN (uint64_t, uint32_t, 4294967295)
+
+int
+main ()
+{
+ RUN_ALL ()
+}
--
2.36.1
