From mboxrd@z Thu Jan 1 00:00:00 1970 Received: from eggs.gnu.org ([2001:4830:134:3::10]:33363) by lists.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XI2Ff-0002F8-SF for qemu-devel@nongnu.org; Thu, 14 Aug 2014 17:08:10 -0400 Received: from Debian-exim by eggs.gnu.org with spam-scanned (Exim 4.71) (envelope-from ) id 1XI2FZ-0008WG-LR for qemu-devel@nongnu.org; Thu, 14 Aug 2014 17:08:03 -0400 Received: from mx1.redhat.com ([209.132.183.28]:9494) by eggs.gnu.org with esmtp (Exim 4.71) (envelope-from ) id 1XI2FZ-0008V2-CQ for qemu-devel@nongnu.org; Thu, 14 Aug 2014 17:07:57 -0400 Date: Thu, 14 Aug 2014 23:08:30 +0200 From: "Michael S. Tsirkin" Message-ID: <20140814210830.GA12649@redhat.com> References: <1408044362-11621-1-git-send-email-ehabkost@redhat.com> <1408044362-11621-28-git-send-email-ehabkost@redhat.com> MIME-Version: 1.0 Content-Type: text/plain; charset=us-ascii Content-Disposition: inline In-Reply-To: <1408044362-11621-28-git-send-email-ehabkost@redhat.com> Subject: Re: [Qemu-devel] [PATCH v4 27/33] target-i386: Register X86CPU "feat-kvmclock" feature List-Id: List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , To: Eduardo Habkost Cc: Marcel Apfelbaum , Alexander Graf , Don Slutz , qemu-devel@nongnu.org, Igor Mammedov , Andreas =?iso-8859-1?Q?F=E4rber?= On Thu, Aug 14, 2014 at 04:25:56PM -0300, Eduardo Habkost wrote: > The "kvmclock" feature is special because it affects two bits in the KVM > CPUID leaf, so it has to be handled differently from the other feature > properties that will be added. > > Signed-off-by: Eduardo Habkost > --- > target-i386/cpu.c | 61 +++++++++++++++++++++++++++++++++++++++++++++++++++++++ > 1 file changed, 61 insertions(+) > > diff --git a/target-i386/cpu.c b/target-i386/cpu.c > index b005b0d..0eb401b 100644 > --- a/target-i386/cpu.c > +++ b/target-i386/cpu.c > @@ -2774,6 +2774,61 @@ uint32_t x86_cpu_apic_id_from_index(unsigned int cpu_index) > } > } > > +typedef struct FeatureProperty { > + FeatureWord word; > + uint32_t mask; > +} FeatureProperty; > + > + > +static void x86_cpu_get_feature_prop(Object *obj, > + struct Visitor *v, > + void *opaque, > + const char *name, > + Error **errp) > +{ > + X86CPU *cpu = X86_CPU(obj); > + CPUX86State *env = &cpu->env; > + FeatureProperty *fp = opaque; > + bool value = (env->features[fp->word] & fp->mask) == fp->mask; > + visit_type_bool(v, &value, name, errp); > +} > + > +static void x86_cpu_set_feature_prop(Object *obj, > + struct Visitor *v, > + void *opaque, > + const char *name, > + Error **errp) > +{ > + X86CPU *cpu = X86_CPU(obj); > + CPUX86State *env = &cpu->env; > + FeatureProperty *fp = opaque; > + bool value; > + visit_type_bool(v, &value, name, errp); > + if (value) { > + env->features[fp->word] |= fp->mask; > + } else { > + env->features[fp->word] &= ~fp->mask; > + } > +} > + > +/* Register a boolean feature-bits property. > + * If mask has multiple bits, all must be set for the property to return true. > + */ > +static void x86_cpu_register_feature_prop(X86CPU *cpu, > + const char *prop_name, > + FeatureWord w, > + uint32_t mask) > +{ > + FeatureProperty *fp; > + fp = g_new0(FeatureProperty, 1); > + fp->word = w; > + fp->mask = mask; > + object_property_add(OBJECT(cpu), prop_name, "bool", > + x86_cpu_set_feature_prop, > + x86_cpu_get_feature_prop, > + NULL, fp, &error_abort); > +} > + This looks similar to what what DEFINE_PROP_BIT does. Can't this be reused in some way? > static void x86_cpu_initfn(Object *obj) > { > CPUState *cs = CPU(obj); > @@ -2819,6 +2874,12 @@ static void x86_cpu_initfn(Object *obj) > x86_cpu_get_feature_words, > NULL, NULL, (void *)cpu->filtered_features, NULL); > > + /* "feat-kvmclock" will affect both kvmclock feature bits */ > + x86_cpu_register_feature_prop(cpu, "feat-kvmclock", FEAT_KVM, > + (1UL << KVM_FEATURE_CLOCKSOURCE) | > + (1UL << KVM_FEATURE_CLOCKSOURCE2)); > + > + > cpu->hyperv_spinlock_attempts = HYPERV_SPINLOCK_NEVER_RETRY; > env->cpuid_apic_id = x86_cpu_apic_id_from_index(cs->cpu_index); > > -- > 1.9.3