From: "juzhe.zhong@rivai.ai" <juzhe.zhong@rivai.ai>
To: "Robin Dapp" <rdapp.gcc@gmail.com>, palmer <palmer@dabbelt.com>
Cc: gcc-patches <gcc-patches@gcc.gnu.org>,
kito.cheng <kito.cheng@gmail.com>,
collison <collison@rivosinc.com>,
jeffreyalaw <jeffreyalaw@gmail.com>,
"Robin Dapp" <rdapp.gcc@gmail.com>
Subject: Re: [PATCH v2] RISC-V: Split off shift patterns for autovectorization.
Date: Thu, 11 May 2023 18:35:43 +0800 [thread overview]
Message-ID: <3D528D4EC594F954+20230511183542481340220@rivai.ai> (raw)
In-Reply-To: <22063fee-8e38-6da4-8658-4e7c80a3199e@gmail.com>
[-- Attachment #1: Type: text/plain, Size: 4518 bytes --]
LGTM. Plz commit it now. Then I can rebase vec_init patch.
juzhe.zhong@rivai.ai
From: Robin Dapp
Date: 2023-05-11 18:33
To: Palmer Dabbelt
CC: gcc-patches; juzhe.zhong; Kito Cheng; collison; jeffreyalaw; rdapp.gcc
Subject: [PATCH v2] RISC-V: Split off shift patterns for autovectorization.
> "csr_operand" does seem wrong, though, as that just accepts constants.
> Maybe "arith_operand" is the way to go? I haven't looked at the
> V immediates though.
I was pondering changing the shift-count operand to QImode everywhere
but that indeed does not help code generation across the board. It can
still work but might require extra patterns here and there.
"csr_operand" accepts 0-31 constants as well as registers which should
be fine here.
No changes from v1 apart from the RISC-V in the subject and a bit of
rebasing and comments.
This patch splits off the shift patterns of the binop patterns.
This is necessary as the scalar shifts require a Pmode operand
as shift count. To this end, a new iterator any_int_binop_no_shift
is introduced. At a later point when the binops are split up
further in commutative and non-commutative patterns (which both
do not include the shift patterns) we might not need this anymore.
gcc/ChangeLog:
* config/riscv/autovec.md (<optab><mode>3): Add scalar shift
pattern.
(v<optab><mode>3): Add vector shift pattern.
* config/riscv/vector-iterators.md: New iterator.
---
gcc/config/riscv/autovec.md | 47 +++++++++++++++++++++++++++-
gcc/config/riscv/vector-iterators.md | 4 +++
2 files changed, 50 insertions(+), 1 deletion(-)
diff --git a/gcc/config/riscv/autovec.md b/gcc/config/riscv/autovec.md
index 58926ed3e67..ac0c939d277 100644
--- a/gcc/config/riscv/autovec.md
+++ b/gcc/config/riscv/autovec.md
@@ -97,7 +97,7 @@ (define_expand "@vec_series<mode>"
(define_expand "<optab><mode>3"
[(set (match_operand:VI 0 "register_operand")
- (any_int_binop:VI
+ (any_int_binop_no_shift:VI
(match_operand:VI 1 "<binop_rhs1_predicate>")
(match_operand:VI 2 "<binop_rhs2_predicate>")))]
"TARGET_VECTOR"
@@ -119,3 +119,48 @@ (define_expand "<optab><mode>3"
NULL, <VM>mode);
DONE;
})
+
+;; -------------------------------------------------------------------------
+;; ---- [INT] Binary shifts by scalar.
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - vsll.vx/vsra.vx/vsrl.vx
+;; - vsll.vi/vsra.vi/vsrl.vi
+;; -------------------------------------------------------------------------
+
+(define_expand "<optab><mode>3"
+ [(set (match_operand:VI 0 "register_operand")
+ (any_shift:VI
+ (match_operand:VI 1 "register_operand")
+ (match_operand:<VEL> 2 "csr_operand")))]
+ "TARGET_VECTOR"
+{
+ if (!CONST_SCALAR_INT_P (operands[2]))
+ operands[2] = gen_lowpart (Pmode, operands[2]);
+ riscv_vector::emit_len_binop (code_for_pred_scalar
+ (<CODE>, <MODE>mode),
+ operands[0], operands[1], operands[2],
+ NULL_RTX, <VM>mode, Pmode);
+ DONE;
+})
+
+;; -------------------------------------------------------------------------
+;; ---- [INT] Binary shifts by scalar.
+;; -------------------------------------------------------------------------
+;; Includes:
+;; - vsll.vv/vsra.vv/vsrl.vv
+;; -------------------------------------------------------------------------
+
+(define_expand "v<optab><mode>3"
+ [(set (match_operand:VI 0 "register_operand")
+ (any_shift:VI
+ (match_operand:VI 1 "register_operand")
+ (match_operand:VI 2 "vector_shift_operand")))]
+ "TARGET_VECTOR"
+{
+ riscv_vector::emit_len_binop (code_for_pred
+ (<CODE>, <MODE>mode),
+ operands[0], operands[1], operands[2],
+ NULL_RTX, <VM>mode);
+ DONE;
+})
diff --git a/gcc/config/riscv/vector-iterators.md b/gcc/config/riscv/vector-iterators.md
index 29c9d77674b..5cf958ba845 100644
--- a/gcc/config/riscv/vector-iterators.md
+++ b/gcc/config/riscv/vector-iterators.md
@@ -1409,6 +1409,10 @@ (define_code_iterator any_commutative_binop [plus and ior xor
(define_code_iterator any_non_commutative_binop [minus div udiv mod umod])
+(define_code_iterator any_int_binop_no_shift
+ [plus minus and ior xor smax umax smin umin mult div udiv mod umod
+])
+
(define_code_iterator any_sat_int_binop [ss_plus ss_minus us_plus us_minus])
(define_code_iterator sat_int_plus_binop [ss_plus us_plus])
(define_code_iterator sat_int_minus_binop [ss_minus us_minus])
--
2.40.0
next prev parent reply other threads:[~2023-05-11 10:36 UTC|newest]
Thread overview: 8+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-05-10 15:24 [PATCH] riscv: " Robin Dapp
2023-05-10 19:19 ` Palmer Dabbelt
2023-05-10 23:00 ` 钟居哲
2023-05-11 10:33 ` [PATCH v2] RISC-V: " Robin Dapp
2023-05-11 10:35 ` juzhe.zhong [this message]
2023-05-11 11:10 ` Kito Cheng
2023-05-11 14:21 ` Jeff Law
2023-05-11 17:51 ` Palmer Dabbelt
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=3D528D4EC594F954+20230511183542481340220@rivai.ai \
--to=juzhe.zhong@rivai.ai \
--cc=collison@rivosinc.com \
--cc=gcc-patches@gcc.gnu.org \
--cc=jeffreyalaw@gmail.com \
--cc=kito.cheng@gmail.com \
--cc=palmer@dabbelt.com \
--cc=rdapp.gcc@gmail.com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
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).