Re: [PATCH v3] target/i386: Set maximum APIC ID to KVM prior to vCPU creation

["Michael S. Tsirkin" <mst@redhat.com>]

On Fri, Oct 14, 2022 at 09:01:02AM +0800, Zeng Guang wrote:
> PING again !
> This QEMU patch is to optimize max APIC ID set for current VM session
> introduced since linux v6.0. It's also compatible with previous linux
> version.
> 
> Thanks.
> 
> On 9/5/2022 9:27 AM, Zeng Guang wrote:
> > Kindly PING!
> > 
> > On 8/25/2022 10:52 AM, Zeng Guang wrote:
> > > Specify maximum possible APIC ID assigned for current VM session to KVM
> > > prior to the creation of vCPUs. By this setting, KVM can set up VM-scoped
> > > data structure indexed by the APIC ID, e.g. Posted-Interrupt Descriptor
> > > pointer table to support Intel IPI virtualization, with the most optimal
> > > memory footprint.
> > > 
> > > It can be achieved by calling KVM_ENABLE_CAP for KVM_CAP_MAX_VCPU_ID
> > > capability once KVM has enabled it. Ignoring the return error if KVM
> > > doesn't support this capability yet.
> > > 
> > > Signed-off-by: Zeng Guang <guang.zeng@intel.com>
> > > ---
> > >    hw/i386/x86.c              | 4 ++++
> > >    target/i386/kvm/kvm-stub.c | 5 +++++
> > >    target/i386/kvm/kvm.c      | 5 +++++
> > >    target/i386/kvm/kvm_i386.h | 1 +
> > >    4 files changed, 15 insertions(+)
> > > 
> > > diff --git a/hw/i386/x86.c b/hw/i386/x86.c
> > > index 050eedc0c8..4831193c86 100644
> > > --- a/hw/i386/x86.c
> > > +++ b/hw/i386/x86.c
> > > @@ -139,6 +139,10 @@ void x86_cpus_init(X86MachineState *x86ms, int default_cpu_version)
> > >            exit(EXIT_FAILURE);
> > >        }
> > > +    if (kvm_enabled()) {
> > > +        kvm_set_max_apic_id(x86ms->apic_id_limit);
> > > +    }
> > > +
> > >        possible_cpus = mc->possible_cpu_arch_ids(ms);
> > >        for (i = 0; i < ms->smp.cpus; i++) {
> > >            x86_cpu_new(x86ms, possible_cpus->cpus[i].arch_id, &error_fatal);
> > > diff --git a/target/i386/kvm/kvm-stub.c b/target/i386/kvm/kvm-stub.c
> > > index f6e7e4466e..e052f1c7b0 100644
> > > --- a/target/i386/kvm/kvm-stub.c
> > > +++ b/target/i386/kvm/kvm-stub.c
> > > @@ -44,3 +44,8 @@ bool kvm_hyperv_expand_features(X86CPU *cpu, Error **errp)
> > >    {
> > >        abort();
> > >    }
> > > +
> > > +void kvm_set_max_apic_id(uint32_t max_apic_id)
> > > +{
> > > +    return;
> > > +}
> > > diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c
> > > index f148a6d52f..af4ef1e8f0 100644
> > > --- a/target/i386/kvm/kvm.c
> > > +++ b/target/i386/kvm/kvm.c
> > > @@ -5428,3 +5428,8 @@ void kvm_request_xsave_components(X86CPU *cpu, uint64_t mask)
> > >            mask &= ~BIT_ULL(bit);
> > >        }
> > >    }
> > > +
> > > +void kvm_set_max_apic_id(uint32_t max_apic_id)
> > > +{
> > > +    kvm_vm_enable_cap(kvm_state, KVM_CAP_MAX_VCPU_ID, 0, max_apic_id);
> > > +}
> > > diff --git a/target/i386/kvm/kvm_i386.h b/target/i386/kvm/kvm_i386.h
> > > index 4124912c20..c133b32a58 100644
> > > --- a/target/i386/kvm/kvm_i386.h
> > > +++ b/target/i386/kvm/kvm_i386.h
> > > @@ -54,4 +54,5 @@ uint64_t kvm_swizzle_msi_ext_dest_id(uint64_t address);
> > >    bool kvm_enable_sgx_provisioning(KVMState *s);
> > >    void kvm_request_xsave_components(X86CPU *cpu, uint64_t mask);
> > > +void kvm_set_max_apic_id(uint32_t max_apic_id);
> > >    #endif


Looks ok on the surface, but this is Paolo's area.

Acked-by: Michael S. Tsirkin <mst@redhat.com>

-- 
MST