Re: [PATCH] monitor: Support specified vCPU registers

[Darren Kenny]

On Tuesday, 2022-07-19 at 15:55:44 +08, zhenwei pi wrote:
> Originally we have to get all the vCPU registers and parse the
> specified one. To improve the performance of this usage, allow user
> specified vCPU id to query registers.
>
> Run a VM with 16 vCPU, use bcc tool to track the latency of
> 'hmp_info_registers':
> 'info registers -a' uses about 3ms;
> 'info register 12' uses about 150us.

TYPO: I'm assuming that should be 'info registers 12'?

>
> Signed-off-by: zhenwei pi <pizhenwei@bytedance.com>
> ---
>  hmp-commands-info.hx |  6 +++---
>  monitor/misc.c       | 19 +++++++++++++++++++
>  2 files changed, 22 insertions(+), 3 deletions(-)
>
> diff --git a/hmp-commands-info.hx b/hmp-commands-info.hx
> index 3ffa24bd67..6023e2b5c5 100644
> --- a/hmp-commands-info.hx
> +++ b/hmp-commands-info.hx
> @@ -100,9 +100,9 @@ ERST
>  
>      {
>          .name       = "registers",
> -        .args_type  = "cpustate_all:-a",
> -        .params     = "[-a]",
> -        .help       = "show the cpu registers (-a: all - show register info 
> for all cpus)",
> +        .args_type  = "cpustate_all:-a,vcpu:i?",
> +        .params     = "[-a] [vcpu]",

>From what I can see in the code, only one of these may be specified at a
time - or at least, '-a' will take precedence.

Maybe it would read more correctly as '[-a|vcpu]' ?

> +        .help       = "show the cpu registers (-a: all - show register info 
> for all cpus; vcpu: vCPU to query)",

Possibly also it would be worth saying "...; vcpu: specific vCPU to query"?

>          .cmd        = hmp_info_registers,
>      },
>  
> diff --git a/monitor/misc.c b/monitor/misc.c
> index 3d2312ba8d..b12309faad 100644
> --- a/monitor/misc.c
> +++ b/monitor/misc.c
> @@ -307,6 +307,7 @@ int monitor_get_cpu_index(Monitor *mon)
>  static void hmp_info_registers(Monitor *mon, const QDict *qdict)
>  {
>      bool all_cpus = qdict_get_try_bool(qdict, "cpustate_all", false);
> +    int vcpu = qdict_get_try_int(qdict, "vcpu", -1);
>      CPUState *cs;
>  
>      if (all_cpus) {
> @@ -314,6 +315,24 @@ static void hmp_info_registers(Monitor *mon, const QDict 
> *qdict)
>              monitor_printf(mon, "\nCPU#%d\n", cs->cpu_index);
>              cpu_dump_state(cs, NULL, CPU_DUMP_FPU);
>          }
> +    } else if (vcpu >= 0) {
> +        CPUState *target_cs = NULL;
> +
> +        CPU_FOREACH(cs) {
> +            if (cs->cpu_index == vcpu) {
> +                target_cs = cs;
> +                break;
> +            }
> +        }
> +
> +        if (!target_cs) {
> +            monitor_printf(mon, "CPU#%d not available\n", vcpu);
> +            return;
> +        }
> +
> +        monitor_printf(mon, "\nCPU#%d\n", target_cs->cpu_index);
> +        cpu_dump_state(target_cs, NULL, CPU_DUMP_FPU);
> +        return;

This return call seems unnecessary here.

Thanks,

Darren.