qemu-devel.nongnu.org archive mirror
 help / color / mirror / Atom feed
From: Igor Mammedov <imammedo@redhat.com>
To: Igor Mammedov <imammedo@redhat.com>
Cc: "Eduardo Habkost" <ehabkost@redhat.com>,
	"Andreas Färber" <afaerber@suse.de>,
	qemu-devel@nongnu.org
Subject: Re: [Qemu-devel] [RFC 7/9] target-i386: CPU subclass for -cpu "host"
Date: Wed, 2 Jan 2013 21:07:26 +0100	[thread overview]
Message-ID: <20130102210726.3aae1692@thinkpad.mammed.net> (raw)
In-Reply-To: <20130102200026.3ab5eee4@thinkpad.mammed.net>

On Wed, 2 Jan 2013 20:00:26 +0100
Igor Mammedov <imammedo@redhat.com> wrote:

> On Fri, 28 Dec 2012 18:34:04 -0200
> Eduardo Habkost <ehabkost@redhat.com> wrote:
> 
> > Note that we are initializing the CPU features inside instance_init (and
> > not storing any CPU feature information inside the class struct) because
> > kvm_cpu_fill_host() needs KVM to be initialized, and we can't guarantee
> > that KVM will be initialized when class_init is called.
> initializing defaults in initfn will be broken after we convert features into
> static properties due to all initfn()s are called before static properties
> defaults are set.
Never mind, It was my wish-full thinking;
currently device_initfn() sets static property defaults, then sets global
properties and only then children initfn()s are called.

> 
> Is it possible to initialize kvm first before calling class_init().
> 
> > 
> > Signed-off-by: Eduardo Habkost <ehabkost@redhat.com>
> > ---
> >  target-i386/cpu.c | 45 ++++++++++++++++++++++++++++++++++++++++++---
> >  1 file changed, 42 insertions(+), 3 deletions(-)
> > 
> > diff --git a/target-i386/cpu.c b/target-i386/cpu.c
> > index c824c08..2b6cc3b 100644
> > --- a/target-i386/cpu.c
> > +++ b/target-i386/cpu.c
> > @@ -330,6 +330,14 @@ typedef struct x86_def_t {
> >  #define TCG_SVM_FEATURES 0
> >  #define TCG_7_0_EBX_FEATURES (CPUID_7_0_EBX_SMEP | CPUID_7_0_EBX_SMAP)
> >  
> > +
> > +/* CPU class name definitions: */
> > +
> > +#define CPU_CLASS_NAME(name) (name "-" TYPE_X86_CPU)
> > +
> > +#define TYPE_X86_HOST_CPU CPU_CLASS_NAME("host")
> > +
> > +
> >  /* maintains list of cpu model definitions
> >   */
> >  static x86_def_t *x86_defs = {NULL};
> > @@ -1221,9 +1229,7 @@ static X86CPU *x86_cpu_create_from_name(const char *name, Error **errp)
> >  
> >      if (kvm_enabled() && name && strcmp(name, "host") == 0) {
> >  #ifdef CONFIG_KVM
> > -        cpu = X86_CPU(object_new(TYPE_X86_CPU));
> > -        kvm_cpu_fill_host(x86_cpu_def);
> > -        cpudef_2_x86_cpu(cpu, x86_cpu_def, &error);
> > +        cpu = X86_CPU(object_new(TYPE_X86_HOST_CPU));
> >  #endif
> >      } else {
> >          x86_def_t *def;
> > @@ -2168,9 +2174,42 @@ static const TypeInfo x86_cpu_type_info = {
> >      .class_init = x86_cpu_common_class_init,
> >  };
> >  
> > +#ifdef CONFIG_KVM
> > +
> > +static void x86_host_cpu_initfn(Object *obj)
> > +{
> > +    X86CPU *cpu = X86_CPU(obj);
> > +    Error *err = NULL;
> > +    x86_def_t cpudef;
> > +
> > +    memset(&cpudef, 0, sizeof(cpudef));
> > +    kvm_cpu_fill_host(&cpudef);
> > +    cpudef_2_x86_cpu(cpu, &cpudef, &err);
> > +
> > +    if (err) {
> > +        error_report("unexpected cpu init error: %s", error_get_pretty(err));
> > +        exit(1);
> > +    }
> > +}
> > +
> > +static const TypeInfo x86_host_cpu_type_info = {
> > +    .name = TYPE_X86_HOST_CPU,
> > +    .parent = TYPE_X86_CPU,
> > +    .instance_size = sizeof(X86CPU),
> > +    .instance_init = x86_host_cpu_initfn,
> > +    .abstract = false,
> > +    .class_size = sizeof(X86CPUClass),
> > +};
> > +
> > +#endif /* CONFIG_KVM */
> > +
> > +
> >  static void x86_cpu_register_types(void)
> >  {
> >      type_register_static(&x86_cpu_type_info);
> > +#ifdef CONFIG_KVM
> > +    type_register_static(&x86_host_cpu_type_info);
> > +#endif
> >  }
> >  
> >  type_init(x86_cpu_register_types)
> > -- 
> > 1.7.11.7
> > 
> 
> 
> -- 
> Regards,
>   Igor
> 


-- 
Regards,
  Igor

  reply	other threads:[~2013-01-02 20:07 UTC|newest]

Thread overview: 14+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2012-12-28 20:33 [Qemu-devel] [PATCH 0/9] x86 CPU subclasses Eduardo Habkost
2012-12-28 20:33 ` [Qemu-devel] [PATCH 1/9] target-i386: Move CPU object creation to cpu.c Eduardo Habkost
2012-12-28 20:33 ` [Qemu-devel] [PATCH 2/9] target-i386: Make cpu_x86_create() get Error argument Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [PATCH 3/9] target-i386: Simplify cpu_x86_find_by_name() logic Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 4/9] target-i386: Set feature string parsing results directly on CPU object Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 5/9] target-i386: Move kvm_features/hypervisor initialization to cpu_x86_find_by_name() Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 6/9] target-i386: Move CPU creation code to model name lookup function Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 7/9] target-i386: CPU subclass for -cpu "host" Eduardo Habkost
2013-01-02 19:00   ` Igor Mammedov
2013-01-02 20:07     ` Igor Mammedov [this message]
2013-01-02 20:20       ` Eduardo Habkost
2013-01-02 20:16     ` Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 8/9] target-i386: CPU subclasses for predefined CPU models Eduardo Habkost
2012-12-28 20:34 ` [Qemu-devel] [RFC 9/9] target-i386: Unify CPU object creation on x86_cpu_create_from_name() Eduardo Habkost

Reply instructions:

You may reply publicly to this message via plain-text email
using any one of the following methods:

* Save the following mbox file, import it into your mail client,
  and reply-to-all from there: mbox

  Avoid top-posting and favor interleaved quoting:
  https://en.wikipedia.org/wiki/Posting_style#Interleaved_style

* Reply using the --to, --cc, and --in-reply-to
  switches of git-send-email(1):

  git send-email \
    --in-reply-to=20130102210726.3aae1692@thinkpad.mammed.net \
    --to=imammedo@redhat.com \
    --cc=afaerber@suse.de \
    --cc=ehabkost@redhat.com \
    --cc=qemu-devel@nongnu.org \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).