Re: [PATCH v2 4/6] target/openrisc: Use cpu_unwind_state_data for mfspr

[Claudio Fontana]

On 10/27/22 12:02, Richard Henderson wrote:
> Since we do not plan to exit, use cpu_unwind_state_data
> and extract exactly the data requested.
> 
> Signed-off-by: Richard Henderson <richard.hender...@linaro.org>
> ---
>  target/openrisc/sys_helper.c | 11 +++++++++--
>  1 file changed, 9 insertions(+), 2 deletions(-)
> 
> diff --git a/target/openrisc/sys_helper.c b/target/openrisc/sys_helper.c
> index a3508e421d..dde2fa1623 100644
> --- a/target/openrisc/sys_helper.c
> +++ b/target/openrisc/sys_helper.c
> @@ -199,6 +199,7 @@ target_ulong HELPER(mfspr)(CPUOpenRISCState *env, 
> target_ulong rd,
>                             target_ulong spr)
>  {
>  #ifndef CONFIG_USER_ONLY
> +    uint64_t data[TARGET_INSN_START_WORDS];
>      MachineState *ms = MACHINE(qdev_get_machine());
>      OpenRISCCPU *cpu = env_archcpu(env);
>      CPUState *cs = env_cpu(env);
> @@ -232,14 +233,20 @@ target_ulong HELPER(mfspr)(CPUOpenRISCState *env, 
> target_ulong rd,
>          return env->evbar;
>  
>      case TO_SPR(0, 16): /* NPC (equals PC) */
> -        cpu_restore_state(cs, GETPC(), false);
> +        if (cpu_unwind_state_data(cs, GETPC(), data)) {
> +            return data[0];
> +        }
>          return env->pc;
>  
>      case TO_SPR(0, 17): /* SR */
>          return cpu_get_sr(env);
>  
>      case TO_SPR(0, 18): /* PPC */
> -        cpu_restore_state(cs, GETPC(), false);
> +        if (cpu_unwind_state_data(cs, GETPC(), data)) {
> +            if (data[1] & 2) {
> +                return data[0] - 4;
> +            }
> +        }
>          return env->ppc;
>  
>      case TO_SPR(0, 32): /* EPCR */

I am struggling to understand if the fact that we are not setting 
cpu->env.dflag anymore in the mfspr helper is fine;

here I am unfamiliar with the arch, also Ccing Philippe in case he wants to 
step in to review this bit.

Thanks,

CLaudio