[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]
[PATCH v2 09/15] target/arm: Wrap vector uqadd/sqadd/uqsub/sqsub GVecGen
From: |
Richard Henderson |
Subject: |
[PATCH v2 09/15] target/arm: Wrap vector uqadd/sqadd/uqsub/sqsub GVecGen4 in GVecGen3Fn |
Date: |
Sat, 2 May 2020 15:44:57 -0700 |
Provide a functional interface for the vector expansion.
This fits better with the existing set of helpers that
we provide for other operations.
Signed-off-by: Richard Henderson <address@hidden>
---
target/arm/translate.h | 13 +-
target/arm/translate-a64.c | 22 ++--
target/arm/translate.c | 248 +++++++++++++++++++++----------------
3 files changed, 157 insertions(+), 126 deletions(-)
diff --git a/target/arm/translate.h b/target/arm/translate.h
index b3e47e7a7f..ada84d411d 100644
--- a/target/arm/translate.h
+++ b/target/arm/translate.h
@@ -298,16 +298,21 @@ void gen_gvec_sshl(unsigned vece, uint32_t rd_ofs,
uint32_t rn_ofs,
void gen_gvec_ushl(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz);
-extern const GVecGen4 uqadd_op[4];
-extern const GVecGen4 sqadd_op[4];
-extern const GVecGen4 uqsub_op[4];
-extern const GVecGen4 sqsub_op[4];
void gen_cmtst_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b);
void gen_ushl_i32(TCGv_i32 d, TCGv_i32 a, TCGv_i32 b);
void gen_sshl_i32(TCGv_i32 d, TCGv_i32 a, TCGv_i32 b);
void gen_ushl_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b);
void gen_sshl_i64(TCGv_i64 d, TCGv_i64 a, TCGv_i64 b);
+void gen_gvec_uqadd_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz);
+void gen_gvec_sqadd_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz);
+void gen_gvec_uqsub_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz);
+void gen_gvec_sqsub_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz);
+
void gen_gvec_ssra(unsigned vece, uint32_t rd_ofs, uint32_t rm_ofs,
int64_t shift, uint32_t opr_sz, uint32_t max_sz);
void gen_gvec_usra(unsigned vece, uint32_t rd_ofs, uint32_t rm_ofs,
diff --git a/target/arm/translate-a64.c b/target/arm/translate-a64.c
index 2be6ab541e..eeaa92b9f1 100644
--- a/target/arm/translate-a64.c
+++ b/target/arm/translate-a64.c
@@ -11185,20 +11185,18 @@ static void disas_simd_3same_int(DisasContext *s,
uint32_t insn)
switch (opcode) {
case 0x01: /* SQADD, UQADD */
- tcg_gen_gvec_4(vec_full_reg_offset(s, rd),
- offsetof(CPUARMState, vfp.qc),
- vec_full_reg_offset(s, rn),
- vec_full_reg_offset(s, rm),
- is_q ? 16 : 8, vec_full_reg_size(s),
- (u ? uqadd_op : sqadd_op) + size);
+ if (u) {
+ gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_uqadd_qc, size);
+ } else {
+ gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_sqadd_qc, size);
+ }
return;
case 0x05: /* SQSUB, UQSUB */
- tcg_gen_gvec_4(vec_full_reg_offset(s, rd),
- offsetof(CPUARMState, vfp.qc),
- vec_full_reg_offset(s, rn),
- vec_full_reg_offset(s, rm),
- is_q ? 16 : 8, vec_full_reg_size(s),
- (u ? uqsub_op : sqsub_op) + size);
+ if (u) {
+ gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_uqsub_qc, size);
+ } else {
+ gen_gvec_fn3(s, is_q, rd, rn, rm, gen_gvec_sqsub_qc, size);
+ }
return;
case 0x08: /* SSHL, USHL */
if (u) {
diff --git a/target/arm/translate.c b/target/arm/translate.c
index e5aa78c88a..8e6c6f7b00 100644
--- a/target/arm/translate.c
+++ b/target/arm/translate.c
@@ -5197,32 +5197,37 @@ static void gen_uqadd_vec(unsigned vece, TCGv_vec t,
TCGv_vec sat,
tcg_temp_free_vec(x);
}
-static const TCGOpcode vecop_list_uqadd[] = {
- INDEX_op_usadd_vec, INDEX_op_cmp_vec, INDEX_op_add_vec, 0
-};
-
-const GVecGen4 uqadd_op[4] = {
- { .fniv = gen_uqadd_vec,
- .fno = gen_helper_gvec_uqadd_b,
- .write_aofs = true,
- .opt_opc = vecop_list_uqadd,
- .vece = MO_8 },
- { .fniv = gen_uqadd_vec,
- .fno = gen_helper_gvec_uqadd_h,
- .write_aofs = true,
- .opt_opc = vecop_list_uqadd,
- .vece = MO_16 },
- { .fniv = gen_uqadd_vec,
- .fno = gen_helper_gvec_uqadd_s,
- .write_aofs = true,
- .opt_opc = vecop_list_uqadd,
- .vece = MO_32 },
- { .fniv = gen_uqadd_vec,
- .fno = gen_helper_gvec_uqadd_d,
- .write_aofs = true,
- .opt_opc = vecop_list_uqadd,
- .vece = MO_64 },
-};
+void gen_gvec_uqadd_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz)
+{
+ static const TCGOpcode vecop_list[] = {
+ INDEX_op_usadd_vec, INDEX_op_cmp_vec, INDEX_op_add_vec, 0
+ };
+ static const GVecGen4 ops[4] = {
+ { .fniv = gen_uqadd_vec,
+ .fno = gen_helper_gvec_uqadd_b,
+ .write_aofs = true,
+ .opt_opc = vecop_list,
+ .vece = MO_8 },
+ { .fniv = gen_uqadd_vec,
+ .fno = gen_helper_gvec_uqadd_h,
+ .write_aofs = true,
+ .opt_opc = vecop_list,
+ .vece = MO_16 },
+ { .fniv = gen_uqadd_vec,
+ .fno = gen_helper_gvec_uqadd_s,
+ .write_aofs = true,
+ .opt_opc = vecop_list,
+ .vece = MO_32 },
+ { .fniv = gen_uqadd_vec,
+ .fno = gen_helper_gvec_uqadd_d,
+ .write_aofs = true,
+ .opt_opc = vecop_list,
+ .vece = MO_64 },
+ };
+ tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
+ rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]);
+}
static void gen_sqadd_vec(unsigned vece, TCGv_vec t, TCGv_vec sat,
TCGv_vec a, TCGv_vec b)
@@ -5235,32 +5240,37 @@ static void gen_sqadd_vec(unsigned vece, TCGv_vec t,
TCGv_vec sat,
tcg_temp_free_vec(x);
}
-static const TCGOpcode vecop_list_sqadd[] = {
- INDEX_op_ssadd_vec, INDEX_op_cmp_vec, INDEX_op_add_vec, 0
-};
-
-const GVecGen4 sqadd_op[4] = {
- { .fniv = gen_sqadd_vec,
- .fno = gen_helper_gvec_sqadd_b,
- .opt_opc = vecop_list_sqadd,
- .write_aofs = true,
- .vece = MO_8 },
- { .fniv = gen_sqadd_vec,
- .fno = gen_helper_gvec_sqadd_h,
- .opt_opc = vecop_list_sqadd,
- .write_aofs = true,
- .vece = MO_16 },
- { .fniv = gen_sqadd_vec,
- .fno = gen_helper_gvec_sqadd_s,
- .opt_opc = vecop_list_sqadd,
- .write_aofs = true,
- .vece = MO_32 },
- { .fniv = gen_sqadd_vec,
- .fno = gen_helper_gvec_sqadd_d,
- .opt_opc = vecop_list_sqadd,
- .write_aofs = true,
- .vece = MO_64 },
-};
+void gen_gvec_sqadd_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz)
+{
+ static const TCGOpcode vecop_list[] = {
+ INDEX_op_ssadd_vec, INDEX_op_cmp_vec, INDEX_op_add_vec, 0
+ };
+ static const GVecGen4 ops[4] = {
+ { .fniv = gen_sqadd_vec,
+ .fno = gen_helper_gvec_sqadd_b,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_8 },
+ { .fniv = gen_sqadd_vec,
+ .fno = gen_helper_gvec_sqadd_h,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_16 },
+ { .fniv = gen_sqadd_vec,
+ .fno = gen_helper_gvec_sqadd_s,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_32 },
+ { .fniv = gen_sqadd_vec,
+ .fno = gen_helper_gvec_sqadd_d,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_64 },
+ };
+ tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
+ rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]);
+}
static void gen_uqsub_vec(unsigned vece, TCGv_vec t, TCGv_vec sat,
TCGv_vec a, TCGv_vec b)
@@ -5273,32 +5283,37 @@ static void gen_uqsub_vec(unsigned vece, TCGv_vec t,
TCGv_vec sat,
tcg_temp_free_vec(x);
}
-static const TCGOpcode vecop_list_uqsub[] = {
- INDEX_op_ussub_vec, INDEX_op_cmp_vec, INDEX_op_sub_vec, 0
-};
-
-const GVecGen4 uqsub_op[4] = {
- { .fniv = gen_uqsub_vec,
- .fno = gen_helper_gvec_uqsub_b,
- .opt_opc = vecop_list_uqsub,
- .write_aofs = true,
- .vece = MO_8 },
- { .fniv = gen_uqsub_vec,
- .fno = gen_helper_gvec_uqsub_h,
- .opt_opc = vecop_list_uqsub,
- .write_aofs = true,
- .vece = MO_16 },
- { .fniv = gen_uqsub_vec,
- .fno = gen_helper_gvec_uqsub_s,
- .opt_opc = vecop_list_uqsub,
- .write_aofs = true,
- .vece = MO_32 },
- { .fniv = gen_uqsub_vec,
- .fno = gen_helper_gvec_uqsub_d,
- .opt_opc = vecop_list_uqsub,
- .write_aofs = true,
- .vece = MO_64 },
-};
+void gen_gvec_uqsub_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz)
+{
+ static const TCGOpcode vecop_list[] = {
+ INDEX_op_ussub_vec, INDEX_op_cmp_vec, INDEX_op_sub_vec, 0
+ };
+ static const GVecGen4 ops[4] = {
+ { .fniv = gen_uqsub_vec,
+ .fno = gen_helper_gvec_uqsub_b,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_8 },
+ { .fniv = gen_uqsub_vec,
+ .fno = gen_helper_gvec_uqsub_h,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_16 },
+ { .fniv = gen_uqsub_vec,
+ .fno = gen_helper_gvec_uqsub_s,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_32 },
+ { .fniv = gen_uqsub_vec,
+ .fno = gen_helper_gvec_uqsub_d,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_64 },
+ };
+ tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
+ rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]);
+}
static void gen_sqsub_vec(unsigned vece, TCGv_vec t, TCGv_vec sat,
TCGv_vec a, TCGv_vec b)
@@ -5311,32 +5326,37 @@ static void gen_sqsub_vec(unsigned vece, TCGv_vec t,
TCGv_vec sat,
tcg_temp_free_vec(x);
}
-static const TCGOpcode vecop_list_sqsub[] = {
- INDEX_op_sssub_vec, INDEX_op_cmp_vec, INDEX_op_sub_vec, 0
-};
-
-const GVecGen4 sqsub_op[4] = {
- { .fniv = gen_sqsub_vec,
- .fno = gen_helper_gvec_sqsub_b,
- .opt_opc = vecop_list_sqsub,
- .write_aofs = true,
- .vece = MO_8 },
- { .fniv = gen_sqsub_vec,
- .fno = gen_helper_gvec_sqsub_h,
- .opt_opc = vecop_list_sqsub,
- .write_aofs = true,
- .vece = MO_16 },
- { .fniv = gen_sqsub_vec,
- .fno = gen_helper_gvec_sqsub_s,
- .opt_opc = vecop_list_sqsub,
- .write_aofs = true,
- .vece = MO_32 },
- { .fniv = gen_sqsub_vec,
- .fno = gen_helper_gvec_sqsub_d,
- .opt_opc = vecop_list_sqsub,
- .write_aofs = true,
- .vece = MO_64 },
-};
+void gen_gvec_sqsub_qc(unsigned vece, uint32_t rd_ofs, uint32_t rn_ofs,
+ uint32_t rm_ofs, uint32_t opr_sz, uint32_t max_sz)
+{
+ static const TCGOpcode vecop_list[] = {
+ INDEX_op_sssub_vec, INDEX_op_cmp_vec, INDEX_op_sub_vec, 0
+ };
+ static const GVecGen4 ops[4] = {
+ { .fniv = gen_sqsub_vec,
+ .fno = gen_helper_gvec_sqsub_b,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_8 },
+ { .fniv = gen_sqsub_vec,
+ .fno = gen_helper_gvec_sqsub_h,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_16 },
+ { .fniv = gen_sqsub_vec,
+ .fno = gen_helper_gvec_sqsub_s,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_32 },
+ { .fniv = gen_sqsub_vec,
+ .fno = gen_helper_gvec_sqsub_d,
+ .opt_opc = vecop_list,
+ .write_aofs = true,
+ .vece = MO_64 },
+ };
+ tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
+ rn_ofs, rm_ofs, opr_sz, max_sz, &ops[vece]);
+}
/* Translate a NEON data processing instruction. Return nonzero if the
instruction is invalid.
@@ -5522,15 +5542,23 @@ static int disas_neon_data_insn(DisasContext *s,
uint32_t insn)
return 0;
case NEON_3R_VQADD:
- tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
- rn_ofs, rm_ofs, vec_size, vec_size,
- (u ? uqadd_op : sqadd_op) + size);
+ if (u) {
+ gen_gvec_uqadd_qc(size, rd_ofs, rn_ofs, rm_ofs,
+ vec_size, vec_size);
+ } else {
+ gen_gvec_sqadd_qc(size, rd_ofs, rn_ofs, rm_ofs,
+ vec_size, vec_size);
+ }
return 0;
case NEON_3R_VQSUB:
- tcg_gen_gvec_4(rd_ofs, offsetof(CPUARMState, vfp.qc),
- rn_ofs, rm_ofs, vec_size, vec_size,
- (u ? uqsub_op : sqsub_op) + size);
+ if (u) {
+ gen_gvec_uqsub_qc(size, rd_ofs, rn_ofs, rm_ofs,
+ vec_size, vec_size);
+ } else {
+ gen_gvec_sqsub_qc(size, rd_ofs, rn_ofs, rm_ofs,
+ vec_size, vec_size);
+ }
return 0;
case NEON_3R_VMUL: /* VMUL */
--
2.20.1
- [PATCH v2 00/15] target/arm: partial vector cleanup, Richard Henderson, 2020/05/02
- [PATCH v2 01/15] target/arm: Create gen_gvec_[us]sra, Richard Henderson, 2020/05/02
- [PATCH v2 03/15] target/arm: Create gen_gvec_{sri,sli}, Richard Henderson, 2020/05/02
- [PATCH v2 05/15] target/arm: Tidy handle_vec_simd_shri, Richard Henderson, 2020/05/02
- [PATCH v2 02/15] target/arm: Create gen_gvec_{u,s}{rshr,rsra}, Richard Henderson, 2020/05/02
- [PATCH v2 04/15] target/arm: Remove unnecessary range check for VSHL, Richard Henderson, 2020/05/02
- [PATCH v2 06/15] target/arm: Wrap vector compare zero GVecGen2 in GVecGen2Fn, Richard Henderson, 2020/05/02
- [PATCH v2 08/15] target/arm: Wrap vector cmtst/ushl/sshl GVecGen3 in GVecGen3Fn, Richard Henderson, 2020/05/02
- [PATCH v2 11/15] target/arm: Wrap vector qrdmla/qrdmls in GVecGen3Fn, Richard Henderson, 2020/05/02
- [PATCH v2 10/15] target/arm: Remove fp_status from helper_{recpe, rsqrte}_u32, Richard Henderson, 2020/05/02
- [PATCH v2 09/15] target/arm: Wrap vector uqadd/sqadd/uqsub/sqsub GVecGen4 in GVecGen3Fn,
Richard Henderson <=
- [PATCH v2 12/15] target/arm: Pass pointer to qc to qrdmla/qrdmls, Richard Henderson, 2020/05/02
- [PATCH v2 07/15] target/arm: Wrap vector mla/mls GVecGen3 in GVecGen3Fn, Richard Henderson, 2020/05/02
- [PATCH v2 13/15] target/arm: Clear tail in gvec_fmul_idx_*, gvec_fmla_idx_*, Richard Henderson, 2020/05/02
- [PATCH v2 14/15] target/arm: Vectorize SABD/UABD, Richard Henderson, 2020/05/02
- [PATCH v2 15/15] target/arm: Vectorize SABA/UABA, Richard Henderson, 2020/05/02
- Re: [PATCH v2 00/15] target/arm: partial vector cleanup, Peter Maydell, 2020/05/05