Re: [PATCH qemu v2 02/10] target/riscv: rvv: Add mask agnostic for vector load / store instructions

[Weiwei Li]

在 2022/3/17 下午3:47, ~eopxd 写道:
> From: Yueh-Ting (eop) Chen <eop.chen@sifive.com>
>
> Signed-off-by: eop Chen <eop.chen@sifive.com>
> Reviewed-by: Frank Chang <frank.chang@sifive.com>
> ---
>   target/riscv/insn_trans/trans_rvv.c.inc |  9 +++++++
>   target/riscv/vector_helper.c            | 35 +++++++++++++++++--------
>   2 files changed, 33 insertions(+), 11 deletions(-)
>
> diff --git a/target/riscv/insn_trans/trans_rvv.c.inc 
> b/target/riscv/insn_trans/trans_rvv.c.inc
> index 63ddd54669..9a2d54313a 100644
> --- a/target/riscv/insn_trans/trans_rvv.c.inc
> +++ b/target/riscv/insn_trans/trans_rvv.c.inc
> @@ -712,6 +712,7 @@ static bool ld_us_op(DisasContext *s, arg_r2nfvm *a, 
> uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
>   }
>   
> @@ -750,6 +751,7 @@ static bool st_us_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
>   }
>   
> @@ -778,6 +780,7 @@ static bool ld_us_mask_op(DisasContext *s, arg_vlm_v *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, NF, 1);
>       /* Mask destination register are always tail-agnostic */
>       data = FIELD_DP32(data, VDATA, VTA, s->cfg_vta_all_1s);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_us_trans(a->rd, a->rs1, data, fn, s, false);
>   }
>   
> @@ -797,6 +800,7 @@ static bool st_us_mask_op(DisasContext *s, arg_vsm_v *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, NF, 1);
>       /* Mask destination register are always tail-agnostic */
>       data = FIELD_DP32(data, VDATA, VTA, s->cfg_vta_all_1s);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_us_trans(a->rd, a->rs1, data, fn, s, true);
>   }
>   
> @@ -869,6 +873,7 @@ static bool ld_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_stride_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
>   }
>   
> @@ -899,6 +904,7 @@ static bool st_stride_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       fn = fns[eew];
>       if (fn == NULL) {
>           return false;
> @@ -1000,6 +1006,7 @@ static bool ld_index_op(DisasContext *s, arg_rnfvm *a, 
> uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, false);
>   }
>   
> @@ -1053,6 +1060,7 @@ static bool st_index_op(DisasContext *s, arg_rnfvm *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldst_index_trans(a->rd, a->rs1, a->rs2, data, fn, s, true);
>   }
>   
> @@ -1119,6 +1127,7 @@ static bool ldff_op(DisasContext *s, arg_r2nfvm *a, uint8_t eew)
>       data = FIELD_DP32(data, VDATA, LMUL, emul);
>       data = FIELD_DP32(data, VDATA, NF, a->nf);
>       data = FIELD_DP32(data, VDATA, VTA, s->vta);
> +    data = FIELD_DP32(data, VDATA, VMA, s->vma);
>       return ldff_trans(a->rd, a->rs1, data, fn, s);
>   }
>   
> diff --git a/target/riscv/vector_helper.c b/target/riscv/vector_helper.c
> index 934b283db2..89eea33eb3 100644
> --- a/target/riscv/vector_helper.c
> +++ b/target/riscv/vector_helper.c
> @@ -282,14 +282,18 @@ vext_ldst_stride(void *vd, void *v0, target_ulong base,
>       uint32_t esz = 1 << log2_esz;
>       uint32_t total_elems = vext_get_total_elems(env, desc, esz);
>       uint32_t vta = vext_vta(desc);
> +    uint32_t vma = vext_vma(desc);
>   
>       for (i = env->vstart; i < env->vl; i++, env->vstart++) {
> -        if (!vm && !vext_elem_mask(v0, i)) {
> -            continue;
> -        }
> -
>           k = 0;
>           while (k < nf) {
> +            if (!vm && !vext_elem_mask(v0, i)) {
> +                /* set masked-off elements to 1s */
> +                vext_set_elems_1s(vd, vma,(i + k * max_elems) * esz,

seems lack a space here. the same to following cases.

Regards,

Weiwei Li