From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: qemu-arm@nongnu.org, qemu-ppc@nongnu.org, qemu-riscv@nongnu.org,
qemu-s390x@nongnu.org, jcmvbkbc@gmail.com,
kbastian@mail.uni-paderborn.de, ysato@users.sourceforge.jp,
gaosong@loongson.cn, jiaxun.yang@flygoat.com,
tsimpson@quicinc.com, ale@rev.ng, mrolnik@gmail.com,
edgar.iglesias@gmail.com
Subject: [PATCH 09/70] target/arm: Avoid tcg_const_ptr in disas_simd_zip_trn
Date: Sun, 26 Feb 2023 19:41:32 -1000 [thread overview]
Message-ID: <20230227054233.390271-10-richard.henderson@linaro.org> (raw)
In-Reply-To: <20230227054233.390271-1-richard.henderson@linaro.org>
It is easy enough to use mov instead of or-with-zero
and relying on the optimizer to fold away the or.
Use an array for the output, rather than separate
tcg_res{l,h} variables.
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
target/arm/translate-a64.c | 41 +++++++++++++++++++-------------------
1 file changed, 21 insertions(+), 20 deletions(-)
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index 9e682f36c9..1726ec622d 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -7442,10 +7442,10 @@ static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
bool part = extract32(insn, 14, 1);
bool is_q = extract32(insn, 30, 1);
int esize = 8 << size;
- int i, ofs;
+ int i;
int datasize = is_q ? 128 : 64;
int elements = datasize / esize;
- TCGv_i64 tcg_res, tcg_resl, tcg_resh;
+ TCGv_i64 tcg_res[2], tcg_ele;
if (opcode == 0 || (size == 3 && !is_q)) {
unallocated_encoding(s);
@@ -7456,37 +7456,39 @@ static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
return;
}
- tcg_resl = tcg_const_i64(0);
- tcg_resh = is_q ? tcg_const_i64(0) : NULL;
- tcg_res = tcg_temp_new_i64();
+ tcg_res[0] = tcg_temp_new_i64();
+ tcg_res[1] = is_q ? tcg_temp_new_i64() : NULL;
+ tcg_ele = tcg_temp_new_i64();
for (i = 0; i < elements; i++) {
+ int o, w;
+
switch (opcode) {
case 1: /* UZP1/2 */
{
int midpoint = elements / 2;
if (i < midpoint) {
- read_vec_element(s, tcg_res, rn, 2 * i + part, size);
+ read_vec_element(s, tcg_ele, rn, 2 * i + part, size);
} else {
- read_vec_element(s, tcg_res, rm,
+ read_vec_element(s, tcg_ele, rm,
2 * (i - midpoint) + part, size);
}
break;
}
case 2: /* TRN1/2 */
if (i & 1) {
- read_vec_element(s, tcg_res, rm, (i & ~1) + part, size);
+ read_vec_element(s, tcg_ele, rm, (i & ~1) + part, size);
} else {
- read_vec_element(s, tcg_res, rn, (i & ~1) + part, size);
+ read_vec_element(s, tcg_ele, rn, (i & ~1) + part, size);
}
break;
case 3: /* ZIP1/2 */
{
int base = part * elements / 2;
if (i & 1) {
- read_vec_element(s, tcg_res, rm, base + (i >> 1), size);
+ read_vec_element(s, tcg_ele, rm, base + (i >> 1), size);
} else {
- read_vec_element(s, tcg_res, rn, base + (i >> 1), size);
+ read_vec_element(s, tcg_ele, rn, base + (i >> 1), size);
}
break;
}
@@ -7494,19 +7496,18 @@ static void disas_simd_zip_trn(DisasContext *s, uint32_t insn)
g_assert_not_reached();
}
- ofs = i * esize;
- if (ofs < 64) {
- tcg_gen_shli_i64(tcg_res, tcg_res, ofs);
- tcg_gen_or_i64(tcg_resl, tcg_resl, tcg_res);
+ w = (i * esize) / 64;
+ o = (i * esize) % 64;
+ if (o == 0) {
+ tcg_gen_mov_i64(tcg_res[w], tcg_ele);
} else {
- tcg_gen_shli_i64(tcg_res, tcg_res, ofs - 64);
- tcg_gen_or_i64(tcg_resh, tcg_resh, tcg_res);
+ tcg_gen_shli_i64(tcg_ele, tcg_ele, o);
+ tcg_gen_or_i64(tcg_res[w], tcg_res[w], tcg_ele);
}
}
- write_vec_element(s, tcg_resl, rd, 0, MO_64);
- if (is_q) {
- write_vec_element(s, tcg_resh, rd, 1, MO_64);
+ for (i = 0; i <= is_q; ++i) {
+ write_vec_element(s, tcg_res[i], rd, i, MO_64);
}
clear_vec_high(s, is_q, rd);
}
--
2.34.1
next prev parent reply other threads:[~2023-02-27 5:51 UTC|newest]
Thread overview: 140+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-02-27 5:41 [PATCH 00/70] tcg: Remove tcg_const_* Richard Henderson
2023-02-27 5:41 ` [PATCH 01/70] target/arm: Use rmode >= 0 for need_rmode Richard Henderson
2023-03-06 13:54 ` Philippe Mathieu-Daudé
2023-03-06 13:56 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 02/70] target/arm: Handle FPROUNDING_ODD in arm_rmode_to_sf Richard Henderson
2023-02-27 5:41 ` [PATCH 03/70] target/arm: Improve arm_rmode_to_sf Richard Henderson
2023-03-06 14:00 ` Philippe Mathieu-Daudé
2023-03-06 19:20 ` Richard Henderson
2023-02-27 5:41 ` [PATCH 04/70] target/arm: Consistently use ARMFPRounding during translation Richard Henderson
2023-03-06 13:58 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 05/70] target/arm: Create gen_set_rmode, gen_restore_rmode Richard Henderson
2023-02-27 5:41 ` [PATCH 06/70] target/arm: Improve trans_BFCI Richard Henderson
2023-02-27 5:41 ` [PATCH 07/70] target/arm: Avoid tcg_const_ptr in gen_sve_{ldr,str} Richard Henderson
2023-03-06 15:11 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 08/70] target/arm: Avoid tcg_const_* in translate-mve.c Richard Henderson
2023-02-27 5:41 ` Richard Henderson [this message]
2023-02-27 5:41 ` [PATCH 10/70] target/arm: Avoid tcg_const_ptr in handle_vec_simd_sqshrn Richard Henderson
2023-03-06 15:15 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 11/70] target/arm: Avoid tcg_const_ptr in handle_rev Richard Henderson
2023-03-06 15:22 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 12/70] target/avr: Avoid use of tcg_const_i32 in SBIC, SBIS Richard Henderson
2023-03-06 13:51 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 13/70] target/avr: Avoid use of tcg_const_i32 throughout Richard Henderson
2023-03-06 23:49 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 14/70] target/cris: " Richard Henderson
2023-03-07 0:30 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 15/70] target/hexagon: Use tcg_constant_* for gen_constant_from_imm Richard Henderson
2023-02-27 21:55 ` Taylor Simpson
2023-02-27 5:41 ` [PATCH 16/70] target/hexagon/idef-parser: Use gen_tmp for LPCFG Richard Henderson
2023-02-27 21:55 ` Taylor Simpson
2023-02-27 5:41 ` [PATCH 17/70] target/hexagon/idef-parser: Use gen_tmp for gen_pred_assign Richard Henderson
2023-02-27 21:55 ` Taylor Simpson
2023-02-27 5:41 ` [PATCH 18/70] target/hexagon/idef-parser: Use gen_tmp for gen_rvalue_pred Richard Henderson
2023-02-27 21:55 ` Taylor Simpson
2023-02-27 5:41 ` [PATCH 19/70] target/hexagon/idef-parser: Use gen_constant for gen_extend_tcg_width_op Richard Henderson
2023-02-27 21:55 ` Taylor Simpson
2023-02-27 22:00 ` Richard Henderson
2023-02-27 22:38 ` Taylor Simpson
2023-02-27 5:41 ` [PATCH 20/70] target/hppa: Avoid tcg_const_i64 in trans_fid_f Richard Henderson
2023-03-06 13:50 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 21/70] target/hppa: Avoid use of tcg_const_i32 throughout Richard Henderson
2023-03-06 23:51 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 22/70] target/i386: Simplify POPF Richard Henderson
2023-02-27 9:04 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 23/70] target/i386: Avoid use of tcg_const_* throughout Richard Henderson
2023-03-07 0:37 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 24/70] target/m68k: Reject immediate as destination in gen_ea_mode Richard Henderson
2023-02-27 5:41 ` [PATCH 25/70] target/m68k: Use tcg_constant_i32 " Richard Henderson
2023-03-06 14:14 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 26/70] target/m68k: Avoid tcg_const_i32 when modified Richard Henderson
2023-03-06 23:53 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 27/70] target/m68k: Avoid tcg_const_i32 in bfop_reg Richard Henderson
2023-03-07 0:03 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 28/70] target/m68k: Avoid tcg_const_* throughout Richard Henderson
2023-03-06 23:59 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 29/70] target/microblaze: " Richard Henderson
2023-02-27 8:56 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 30/70] target/mips: Split out gen_lxl Richard Henderson
2023-03-06 13:31 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 31/70] target/mips: Split out gen_lxr Richard Henderson
2023-03-06 13:40 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 32/70] target/mips: Avoid tcg_const_tl in gen_r6_ld Richard Henderson
2023-03-06 13:41 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 33/70] target/mips: Avoid tcg_const_* throughout Richard Henderson
2023-03-06 13:46 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 34/70] target/ppc: Split out gen_vx_vmul10 Richard Henderson
2023-03-06 15:08 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 35/70] target/ppc: Avoid tcg_const_i64 in do_vector_shift_quad Richard Henderson
2023-03-06 14:16 ` Philippe Mathieu-Daudé
2023-02-27 5:41 ` [PATCH 36/70] target/ppc: Avoid tcg_const_i64 in do_vcntmb Richard Henderson
2023-02-27 5:42 ` [PATCH 37/70] target/ppc: Avoid tcg_const_* in vmx-impl.c.inc Richard Henderson
2023-02-27 5:42 ` [PATCH 38/70] target/ppc: Avoid tcg_const_* in xxeval Richard Henderson
2023-02-27 5:42 ` [PATCH 39/70] target/ppc: Avoid tcg_const_* in vsx-impl.c.inc Richard Henderson
2023-02-27 5:42 ` [PATCH 40/70] target/ppc: Avoid tcg_const_* in fp-impl.c.inc Richard Henderson
2023-02-27 5:42 ` [PATCH 41/70] target/ppc: Avoid tcg_const_* in power8-pmu-regs.c.inc Richard Henderson
2023-02-27 5:42 ` [PATCH 42/70] target/ppc: Rewrite trans_ADDG6S Richard Henderson
2023-02-27 5:42 ` [PATCH 43/70] target/ppc: Fix gen_tlbsx_booke206 Richard Henderson
2023-02-27 5:42 ` [PATCH 44/70] target/ppc: Avoid tcg_const_* in translate.c Richard Henderson
2023-02-27 5:42 ` [PATCH 45/70] target/riscv: Avoid tcg_const_* Richard Henderson
2023-02-27 9:05 ` Philippe Mathieu-Daudé
2023-03-06 13:53 ` liweiwei
2023-02-27 5:42 ` [PATCH 46/70] target/rx: Use tcg_gen_abs_i32 Richard Henderson
2023-03-06 13:48 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 47/70] target/rx: Use cpu_psw_z as temp in flags computation Richard Henderson
2023-03-07 0:32 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 48/70] target/rx: Avoid tcg_const_i32 when new temp needed Richard Henderson
2023-03-06 14:18 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 49/70] target/rx: Avoid tcg_const_i32 Richard Henderson
2023-03-07 0:27 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 50/70] target/s390x: Split out gen_ri2 Richard Henderson
2023-02-27 9:09 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 51/70] target/s390x: Avoid tcg_const_i64 Richard Henderson
2023-03-07 0:21 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 52/70] target/sh4: Avoid tcg_const_i32 for TAS.B Richard Henderson
2023-03-07 0:23 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 53/70] target/sh4: Avoid tcg_const_i32 Richard Henderson
2023-03-07 0:21 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 54/70] tcg/sparc: Avoid tcg_const_tl in gen_edge Richard Henderson
2023-03-06 15:36 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 55/70] target/sparc: Avoid tcg_const_{tl,i32} Richard Henderson
2023-03-01 17:02 ` Mark Cave-Ayland
2023-03-06 15:37 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 56/70] target/tricore: Split t_n as constant from temp as variable Richard Henderson
2023-03-07 0:19 ` Philippe Mathieu-Daudé
2023-03-07 2:24 ` Richard Henderson
2023-03-07 10:20 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 57/70] target/tricore: Rename t_off10 and use tcg_constant_i32 Richard Henderson
2023-03-06 15:38 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 58/70] target/tricore: Use min/max for saturate Richard Henderson
2023-02-27 5:53 ` Richard Henderson
2023-02-27 5:42 ` [PATCH 59/70] target/tricore: Use setcondi instead of explicit allocation Richard Henderson
2023-03-06 15:39 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 60/70] target/tricore: Drop some temp initialization Richard Henderson
2023-03-06 15:25 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 61/70] target/tricore: Avoid tcg_const_i32 Richard Henderson
2023-03-07 0:10 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 62/70] target/xtensa: Tidy translate_bb Richard Henderson
2023-02-27 9:19 ` Max Filippov
2023-03-07 0:07 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 63/70] target/xtensa: Tidy translate_clamps Richard Henderson
2023-02-27 9:22 ` Max Filippov
2023-03-07 0:24 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 64/70] target/xtensa: Avoid tcg_const_i32 in translate_l32r Richard Henderson
2023-02-27 9:23 ` Max Filippov
2023-03-06 15:01 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 65/70] target/xtensa: Use tcg_gen_subfi_i32 in translate_sll Richard Henderson
2023-02-27 9:26 ` Max Filippov
2023-03-06 15:01 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 66/70] target/xtensa: Split constant in bit shift Richard Henderson
2023-02-27 9:27 ` Max Filippov
2023-03-06 15:01 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 67/70] target/xtensa: Avoid tcg_const_i32 Richard Henderson
2023-02-27 9:31 ` Max Filippov
2023-03-07 0:06 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 68/70] tcg: Replace tcg_const_i64 in tcg-op.c Richard Henderson
2023-03-06 15:33 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 69/70] tcg: Drop tcg_const_*_vec Richard Henderson
2023-03-06 15:32 ` Philippe Mathieu-Daudé
2023-02-27 5:42 ` [PATCH 70/70] tcg: Drop tcg_const_* Richard Henderson
2023-03-06 15:30 ` Philippe Mathieu-Daudé
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=20230227054233.390271-10-richard.henderson@linaro.org \
--to=richard.henderson@linaro.org \
--cc=ale@rev.ng \
--cc=edgar.iglesias@gmail.com \
--cc=gaosong@loongson.cn \
--cc=jcmvbkbc@gmail.com \
--cc=jiaxun.yang@flygoat.com \
--cc=kbastian@mail.uni-paderborn.de \
--cc=mrolnik@gmail.com \
--cc=qemu-arm@nongnu.org \
--cc=qemu-devel@nongnu.org \
--cc=qemu-ppc@nongnu.org \
--cc=qemu-riscv@nongnu.org \
--cc=qemu-s390x@nongnu.org \
--cc=tsimpson@quicinc.com \
--cc=ysato@users.sourceforge.jp \
/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 NNTP newsgroup(s).