Re: [Qemu-devel] [v2 06/13] Add stxvw4x

[Paolo Bonzini <pbonzini@redhat.com>]

Il 11/10/2013 14:01, Tom Musta ha scritto:
> This patch adds the Store VSX Vector Word*4 Indexed (stxvw4x)
> instruction.
> 
> Signed-off-by: Tom Musta <tommusta@gmail.com>
> ---
>  target-ppc/translate.c |   28 ++++++++++++++++++++++++++++
>  1 files changed, 28 insertions(+), 0 deletions(-)
> 
> diff --git a/target-ppc/translate.c b/target-ppc/translate.c
> index 46b8380..8d53cb5 100644
> --- a/target-ppc/translate.c
> +++ b/target-ppc/translate.c
> @@ -7111,6 +7111,33 @@ static void gen_stxvd2x(DisasContext *ctx)
>      tcg_temp_free(EA);
>  }
> 
> +static void gen_stxvw4x(DisasContext *ctx)
> +{
> +    TCGv EA, tmp;
> +    if (unlikely(!ctx->vsx_enabled)) {
> +        gen_exception(ctx, POWERPC_EXCP_VSXU);
> +        return;
> +    }
> +    gen_set_access_type(ctx, ACCESS_INT);
> +    EA = tcg_temp_new();
> +    gen_addr_reg_index(ctx, EA);
> +    tmp = tcg_temp_new();
> +
> +    tcg_gen_shri_i64(tmp, cpu_vsrh(xS(ctx->opcode)), 32);
> +    gen_qemu_st32(ctx, tmp, EA);
> +    tcg_gen_addi_tl(EA, EA, 4);
> +    gen_qemu_st32(ctx, cpu_vsrh(xS(ctx->opcode)), EA);
> +
> +    tcg_gen_shri_i64(tmp, cpu_vsrl(xS(ctx->opcode)), 32);
> +    tcg_gen_addi_tl(EA, EA, 4);
> +    gen_qemu_st32(ctx, tmp, EA);
> +    tcg_gen_addi_tl(EA, EA, 4);
> +    gen_qemu_st32(ctx, cpu_vsrl(xS(ctx->opcode)), EA);
> +
> +    tcg_temp_free(EA);
> +    tcg_temp_free(tmp);
> +}
> +
>  static void gen_xxpermdi(DisasContext *ctx)
>  {
>      TCGv_i64 xh, xl;
> @@ -9597,6 +9624,7 @@ GEN_HANDLER_E(lxvw4x, 0x1F, 0x0C, 0x18, 0,
> PPC_NONE, PPC2_VSX),
> 
>  GEN_HANDLER_E(stxsdx, 0x1F, 0xC, 0x16, 0, PPC_NONE, PPC2_VSX),
>  GEN_HANDLER_E(stxvd2x, 0x1F, 0xC, 0x1E, 0, PPC_NONE, PPC2_VSX),
> +GEN_HANDLER_E(stxvw4x, 0x1F, 0xC, 0x1C, 0, PPC_NONE, PPC2_VSX),
> 
>  #undef GEN_XX3FORM_DM
>  #define GEN_XX3FORM_DM(name, opc2, opc3) \

Reviewed-by: Paolo Bonzini <pbonzini@redhat.com>