From: Richard Henderson <richard.henderson@linaro.org>
To: qemu-devel@nongnu.org
Cc: Jiajie Chen <c@jia.je>
Subject: [PULL 10/39] tcg/loongarch64: Lower add/sub_vec to vadd/vsub
Date: Fri, 15 Sep 2023 20:29:42 -0700 [thread overview]
Message-ID: <20230916033011.479144-11-richard.henderson@linaro.org> (raw)
In-Reply-To: <20230916033011.479144-1-richard.henderson@linaro.org>
From: Jiajie Chen <c@jia.je>
Lower the following ops:
- add_vec
- sub_vec
Signed-off-by: Jiajie Chen <c@jia.je>
Reviewed-by: Richard Henderson <richard.henderson@linaro.org>
Message-Id: <20230908022302.180442-6-c@jia.je>
Signed-off-by: Richard Henderson <richard.henderson@linaro.org>
---
tcg/loongarch64/tcg-target-con-set.h | 1 +
tcg/loongarch64/tcg-target-con-str.h | 1 +
tcg/loongarch64/tcg-target.c.inc | 61 ++++++++++++++++++++++++++++
3 files changed, 63 insertions(+)
diff --git a/tcg/loongarch64/tcg-target-con-set.h b/tcg/loongarch64/tcg-target-con-set.h
index 8c8ea5d919..2d5dce75c3 100644
--- a/tcg/loongarch64/tcg-target-con-set.h
+++ b/tcg/loongarch64/tcg-target-con-set.h
@@ -32,4 +32,5 @@ C_O1_I2(r, rZ, ri)
C_O1_I2(r, rZ, rJ)
C_O1_I2(r, rZ, rZ)
C_O1_I2(w, w, wM)
+C_O1_I2(w, w, wA)
C_O1_I4(r, rZ, rJ, rZ, rZ)
diff --git a/tcg/loongarch64/tcg-target-con-str.h b/tcg/loongarch64/tcg-target-con-str.h
index a8a1c44014..2ba9c135ac 100644
--- a/tcg/loongarch64/tcg-target-con-str.h
+++ b/tcg/loongarch64/tcg-target-con-str.h
@@ -27,3 +27,4 @@ CONST('Z', TCG_CT_CONST_ZERO)
CONST('C', TCG_CT_CONST_C12)
CONST('W', TCG_CT_CONST_WSZ)
CONST('M', TCG_CT_CONST_VCMP)
+CONST('A', TCG_CT_CONST_VADD)
diff --git a/tcg/loongarch64/tcg-target.c.inc b/tcg/loongarch64/tcg-target.c.inc
index 129dd92910..1a369b237c 100644
--- a/tcg/loongarch64/tcg-target.c.inc
+++ b/tcg/loongarch64/tcg-target.c.inc
@@ -177,6 +177,7 @@ static TCGReg tcg_target_call_oarg_reg(TCGCallReturnKind kind, int slot)
#define TCG_CT_CONST_C12 0x1000
#define TCG_CT_CONST_WSZ 0x2000
#define TCG_CT_CONST_VCMP 0x4000
+#define TCG_CT_CONST_VADD 0x8000
#define ALL_GENERAL_REGS MAKE_64BIT_MASK(0, 32)
#define ALL_VECTOR_REGS MAKE_64BIT_MASK(32, 32)
@@ -214,6 +215,9 @@ static bool tcg_target_const_match(int64_t val, TCGType type, int ct, int vece)
if ((ct & TCG_CT_CONST_VCMP) && -0x10 <= vec_val && vec_val <= 0x1f) {
return true;
}
+ if ((ct & TCG_CT_CONST_VADD) && -0x1f <= vec_val && vec_val <= 0x1f) {
+ return true;
+ }
return false;
}
@@ -1621,6 +1625,51 @@ static void tcg_out_dupi_vec(TCGContext *s, TCGType type, unsigned vece,
}
}
+static void tcg_out_addsub_vec(TCGContext *s, unsigned vece, const TCGArg a0,
+ const TCGArg a1, const TCGArg a2,
+ bool a2_is_const, bool is_add)
+{
+ static const LoongArchInsn add_vec_insn[4] = {
+ OPC_VADD_B, OPC_VADD_H, OPC_VADD_W, OPC_VADD_D
+ };
+ static const LoongArchInsn add_vec_imm_insn[4] = {
+ OPC_VADDI_BU, OPC_VADDI_HU, OPC_VADDI_WU, OPC_VADDI_DU
+ };
+ static const LoongArchInsn sub_vec_insn[4] = {
+ OPC_VSUB_B, OPC_VSUB_H, OPC_VSUB_W, OPC_VSUB_D
+ };
+ static const LoongArchInsn sub_vec_imm_insn[4] = {
+ OPC_VSUBI_BU, OPC_VSUBI_HU, OPC_VSUBI_WU, OPC_VSUBI_DU
+ };
+
+ if (a2_is_const) {
+ int64_t value = sextract64(a2, 0, 8 << vece);
+ if (!is_add) {
+ value = -value;
+ }
+
+ /* Try vaddi/vsubi */
+ if (0 <= value && value <= 0x1f) {
+ tcg_out32(s, encode_vdvjuk5_insn(add_vec_imm_insn[vece], a0, \
+ a1, value));
+ return;
+ } else if (-0x1f <= value && value < 0) {
+ tcg_out32(s, encode_vdvjuk5_insn(sub_vec_imm_insn[vece], a0, \
+ a1, -value));
+ return;
+ }
+
+ /* constraint TCG_CT_CONST_VADD ensures unreachable */
+ g_assert_not_reached();
+ }
+
+ if (is_add) {
+ tcg_out32(s, encode_vdvjvk_insn(add_vec_insn[vece], a0, a1, a2));
+ } else {
+ tcg_out32(s, encode_vdvjvk_insn(sub_vec_insn[vece], a0, a1, a2));
+ }
+}
+
static void tcg_out_vec_op(TCGContext *s, TCGOpcode opc,
unsigned vecl, unsigned vece,
const TCGArg args[TCG_MAX_OP_ARGS],
@@ -1712,6 +1761,12 @@ static void tcg_out_vec_op(TCGContext *s, TCGOpcode opc,
}
tcg_out32(s, encode_vdvjvk_insn(insn, a0, a1, a2));
break;
+ case INDEX_op_add_vec:
+ tcg_out_addsub_vec(s, vece, a0, a1, a2, const_args[2], true);
+ break;
+ case INDEX_op_sub_vec:
+ tcg_out_addsub_vec(s, vece, a0, a1, a2, const_args[2], false);
+ break;
case INDEX_op_dupm_vec:
tcg_out_dupm_vec(s, type, vece, a0, a1, a2);
break;
@@ -1728,6 +1783,8 @@ int tcg_can_emit_vec_op(TCGOpcode opc, TCGType type, unsigned vece)
case INDEX_op_dup_vec:
case INDEX_op_dupm_vec:
case INDEX_op_cmp_vec:
+ case INDEX_op_add_vec:
+ case INDEX_op_sub_vec:
return 1;
default:
return 0;
@@ -1892,6 +1949,10 @@ static TCGConstraintSetIndex tcg_target_op_def(TCGOpcode op)
case INDEX_op_cmp_vec:
return C_O1_I2(w, w, wM);
+ case INDEX_op_add_vec:
+ case INDEX_op_sub_vec:
+ return C_O1_I2(w, w, wA);
+
default:
g_assert_not_reached();
}
--
2.34.1
next prev parent reply other threads:[~2023-09-16 3:32 UTC|newest]
Thread overview: 45+ messages / expand[flat|nested] mbox.gz Atom feed top
2023-09-16 3:29 [PULL 00/39] tcg patch queue Richard Henderson
2023-09-16 3:29 ` [PULL 01/39] accel/tcg: mttcg remove false-negative halted assertion Richard Henderson
2023-09-18 6:44 ` Nicholas Piggin
2023-09-18 7:59 ` Alex Bennée
2023-09-18 10:53 ` Nicholas Piggin
2023-09-18 12:19 ` Alex Bennée
2023-09-16 3:29 ` [PULL 02/39] accel/tcg: Fix the comment for CPUTLBEntryFull Richard Henderson
2023-09-16 3:29 ` [PULL 03/39] util: Delete checks for old host definitions Richard Henderson
2023-09-16 3:29 ` [PULL 04/39] softmmu: " Richard Henderson
2023-09-16 3:29 ` [PULL 05/39] thunk: " Richard Henderson
2023-09-16 3:29 ` [PULL 06/39] tcg/loongarch64: Import LSX instructions Richard Henderson
2023-09-16 3:29 ` [PULL 07/39] tcg/loongarch64: Lower basic tcg vec ops to LSX Richard Henderson
2023-09-16 3:29 ` [PULL 08/39] tcg: pass vece to tcg_target_const_match() Richard Henderson
2023-09-16 3:29 ` [PULL 09/39] tcg/loongarch64: Lower cmp_vec to vseq/vsle/vslt Richard Henderson
2023-09-16 3:29 ` Richard Henderson [this message]
2023-09-16 3:29 ` [PULL 11/39] tcg/loongarch64: Lower vector bitwise operations Richard Henderson
2023-09-16 3:29 ` [PULL 12/39] tcg/loongarch64: Lower neg_vec to vneg Richard Henderson
2023-09-16 3:29 ` [PULL 13/39] tcg/loongarch64: Lower mul_vec to vmul Richard Henderson
2023-09-16 3:29 ` [PULL 14/39] tcg/loongarch64: Lower vector min max ops Richard Henderson
2023-09-16 3:29 ` [PULL 15/39] tcg/loongarch64: Lower vector saturated ops Richard Henderson
2023-09-16 3:29 ` [PULL 16/39] tcg/loongarch64: Lower vector shift vector ops Richard Henderson
2023-09-16 3:29 ` [PULL 17/39] tcg/loongarch64: Lower bitsel_vec to vbitsel Richard Henderson
2023-09-16 3:29 ` [PULL 18/39] tcg/loongarch64: Lower vector shift integer ops Richard Henderson
2023-09-16 3:29 ` [PULL 19/39] tcg/loongarch64: Lower rotv_vec ops to LSX Richard Henderson
2023-09-16 3:29 ` [PULL 20/39] tcg/loongarch64: Lower rotli_vec to vrotri Richard Henderson
2023-09-16 3:29 ` [PULL 21/39] tcg/loongarch64: Implement 128-bit load & store Richard Henderson
2023-09-16 3:29 ` [PULL 22/39] tcg: Add gvec compare with immediate and scalar operand Richard Henderson
2023-09-16 3:29 ` [PULL 23/39] target/arm: Use tcg_gen_gvec_cmpi for compare vs 0 Richard Henderson
2023-09-16 3:29 ` [PULL 24/39] accel/tcg: Simplify tlb_plugin_lookup Richard Henderson
2023-09-16 3:29 ` [PULL 25/39] accel/tcg: Split out io_prepare and io_failed Richard Henderson
2023-09-16 3:29 ` [PULL 26/39] accel/tcg: Use CPUTLBEntryFull.phys_addr in io_failed Richard Henderson
2023-09-16 3:29 ` [PULL 27/39] plugin: Simplify struct qemu_plugin_hwaddr Richard Henderson
2023-09-16 3:30 ` [PULL 28/39] accel/tcg: Merge cpu_transaction_failed into io_failed Richard Henderson
2023-09-16 3:30 ` [PULL 29/39] accel/tcg: Replace direct use of io_readx/io_writex in do_{ld, st}_1 Richard Henderson
2023-09-16 3:30 ` [PULL 30/39] accel/tcg: Merge io_readx into do_ld_mmio_beN Richard Henderson
2023-09-16 3:30 ` [PULL 31/39] accel/tcg: Merge io_writex into do_st_mmio_leN Richard Henderson
2023-09-16 3:30 ` [PULL 32/39] accel/tcg: Introduce do_ld16_mmio_beN Richard Henderson
2023-09-16 3:30 ` [PULL 33/39] accel/tcg: Introduce do_st16_mmio_leN Richard Henderson
2023-09-16 3:30 ` [PULL 34/39] fpu: Add conversions between bfloat16 and [u]int8 Richard Henderson
2023-09-16 3:30 ` [PULL 35/39] fpu: Handle m68k extended precision denormals properly Richard Henderson
2023-09-16 3:30 ` [PULL 36/39] tcg: Add tcg_out_tb_start backend hook Richard Henderson
2023-09-16 3:30 ` [PULL 37/39] util/cpuinfo-aarch64: Add CPUINFO_BTI Richard Henderson
2023-09-16 3:30 ` [PULL 38/39] tcg/aarch64: Emit BTI insns at jump landing pads Richard Henderson
2023-09-16 3:30 ` [PULL 39/39] tcg: Map code_gen_buffer with PROT_BTI Richard Henderson
2023-09-16 4:07 ` [PULL 00/39] tcg patch queue Richard Henderson
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=20230916033011.479144-11-richard.henderson@linaro.org \
--to=richard.henderson@linaro.org \
--cc=c@jia.je \
--cc=qemu-devel@nongnu.org \
/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).