[Qemu-devel] Re: [PATCH v3] qemu, qmp: convert do_inject_nmi() to QObject, QError

[Luiz Capitulino <lcapitulino@redhat.com>]

On Wed, 15 Dec 2010 18:45:09 +0100
Markus Armbruster <armbru@redhat.com> wrote:

> Luiz Capitulino <lcapitulino@redhat.com> writes:
> 
> > On Wed, 15 Dec 2010 19:18:32 +0200
> > Avi Kivity <avi@redhat.com> wrote:
> >
> >> On 12/15/2010 07:09 PM, Luiz Capitulino wrote:
> >> > On Wed, 15 Dec 2010 17:49:27 +0800
> >> > Lai Jiangshan<laijs@cn.fujitsu.com>  wrote:
> >> >
> >> > >
> >> > >  Convert do_inject_nmi() to QObject, QError, we need to use it(via libvirt).
> >> > >
> >> > >  changed from v1
> >> > >  Add document.
> >> > >  Add error handling when the cpu index is invalid.
> >> > >
> >> > >  changed from v2
> >> > >  use QERR_INVALID_PARAMETER_VALUE as Markus suggest.
> >> > >
> >> > >  Signed-off-by:  Lai Jiangshan<laijs@cn.fujitsu.com>
> >> > >  ---
> >> > >  diff --git a/hmp-commands.hx b/hmp-commands.hx
> >> > >  index 23024ba..f86d9fe 100644
> >> > >  --- a/hmp-commands.hx
> >> > >  +++ b/hmp-commands.hx
> >> > >  @@ -724,7 +724,8 @@ ETEXI
> >> > >           .args_type  = "cpu_index:i",
> >> > >           .params     = "cpu",
> >> > >           .help       = "inject an NMI on the given CPU",
> >> > >  -        .mhandler.cmd = do_inject_nmi,
> >> > >  +        .user_print = monitor_user_noop,
> >> > >  +        .mhandler.cmd_new = do_inject_nmi,
> >> > >       },
> >> > >   #endif
> >> > >   STEXI
> >> > >  diff --git a/monitor.c b/monitor.c
> >> > >  index ec31eac..3e33a96 100644
> >> > >  --- a/monitor.c
> >> > >  +++ b/monitor.c
> >> > >  @@ -2119,7 +2119,7 @@ static void do_wav_capture(Monitor *mon, const QDict *qdict)
> >> > >   #endif
> >> > >
> >> > >   #if defined(TARGET_I386)
> >> > >  -static void do_inject_nmi(Monitor *mon, const QDict *qdict)
> >> > >  +static int do_inject_nmi(Monitor *mon, const QDict *qdict, QObject **ret_data)
> >> > >   {
> >> > >       CPUState *env;
> >> > >       int cpu_index = qdict_get_int(qdict, "cpu_index");
> >> > >  @@ -2127,8 +2127,12 @@ static void do_inject_nmi(Monitor *mon, const QDict *qdict)
> >> > >       for (env = first_cpu; env != NULL; env = env->next_cpu)
> >> > >           if (env->cpu_index == cpu_index) {
> >> > >               cpu_interrupt(env, CPU_INTERRUPT_NMI);
> >> > >  -            break;
> >> > >  +            return 0;
> >> > >           }
> >> > >  +
> >> > >  +    qerror_report(QERR_INVALID_PARAMETER_VALUE, "cpu_index",
> >> > >  +                  "a CPU number");
> >> > >  +    return -1;
> >> > >   }
> >> > >   #endif
> >> > >
> >> > >  diff --git a/qmp-commands.hx b/qmp-commands.hx
> >> > >  index e5f157f..fcb6bf2 100644
> >> > >  --- a/qmp-commands.hx
> >> > >  +++ b/qmp-commands.hx
> >> > >  @@ -429,6 +429,33 @@ Example:
> >> > >
> >> > >   EQMP
> >> > >
> >> > >  +#if defined(TARGET_I386)
> >> > >  +    {
> >> > >  +        .name       = "inject_nmi",
> >> > >  +        .args_type  = "cpu_index:i",
> >> > >  +        .params     = "cpu",
> >> > >  +        .help       = "inject an NMI on the given CPU",
> >> > >  +        .user_print = monitor_user_noop,
> >> > >  +        .mhandler.cmd_new = do_inject_nmi,
> >> > >  +    },
> >> > >  +#endif
> >> > >  +SQMP
> >> > >  +inject_nmi
> >> > >  +----------
> >> > >  +
> >> > >  +Inject an NMI on the given CPU (x86 only).
> >> > >  +
> >> > >  +Arguments:
> >> > >  +
> >> > >  +- "cpu_index": the index of the CPU to be injected NMI (json-int)
> >> >
> >> > Please, use cpu-index, that's what we're using for the human-monitor-command.
> >> >
> >> > Avi, Anthony, can you please ACK this new command?
> >> >
> >> 
> >> I'd like to see cpu-index made optional; if not present, nmi all cpus 
> >> (that's what the nmi button on many machines does, or at least I think 
> >> that's what it does).
> >
> > Looks like a GUI feature to me,
> 
> Really?  Can't see how you can build "NMI to all CPUs" from "NMI this
> CPU".  Or am I misunderstanding you?

I guess so. Avi referred to 'nmi button on many machines', I assumed he
meant a virtual machine GUI, am I wrong?

> >                                 _might_ turn out to be an undesirable
> > side effect to client writers.
> 
> They seem to be coping fine with optional arguments elsewhere.

Which we might want to review.

> >                                I guess I prefer a to-all-cpus argument.
> 
> How would that look like?  "cpu-index": "all"?

Like this:

{ "execute": "inject-nmi", "arguments": { "to-all-cpus": true } }

But this looks like an optimization to me, because it's also easy to do:

for cpu in query-cpus; do
  inject-nmi cpu

Unless we want to do this in an "atomic" way, due to side effects I'm
not aware about.

> I find "optional json-int" a simpler schema than "either a json-int or
> the json-string "all".
>