Re: [PATCH-for-9.1 17/18] target/i386: Remove X86CPU::kvm_no_smi_migration field

[Thomas Huth <thuth@redhat.com>]

On 05/03/2024 14.42, Philippe Mathieu-Daudé wrote:
> X86CPU::kvm_no_smi_migration was only used by the
> pc-i440fx-2.3 machine, which got removed. Remove it
> and simplify kvm_put_vcpu_events().
> 
> Signed-off-by: Philippe Mathieu-Daudé <philmd@linaro.org>
> ---
>   target/i386/cpu.h     | 3 ---
>   target/i386/cpu.c     | 2 --
>   target/i386/kvm/kvm.c | 6 ------
>   3 files changed, 11 deletions(-)
> 
> diff --git a/target/i386/cpu.h b/target/i386/cpu.h
> index 952174bb6f..bdc640e844 100644
> --- a/target/i386/cpu.h
> +++ b/target/i386/cpu.h
> @@ -2018,9 +2018,6 @@ struct ArchCPU {
>       /* if set, limit maximum value for phys_bits when host_phys_bits is true */
>       uint8_t host_phys_bits_limit;
>   
> -    /* Stop SMI delivery for migration compatibility with old machines */
> -    bool kvm_no_smi_migration;
> -
>       /* Forcefully disable KVM PV features not exposed in guest CPUIDs */
>       bool kvm_pv_enforce_cpuid;
>   
> diff --git a/target/i386/cpu.c b/target/i386/cpu.c
> index 2666ef3808..0e3ad8db2b 100644
> --- a/target/i386/cpu.c
> +++ b/target/i386/cpu.c
> @@ -7905,8 +7905,6 @@ static Property x86_cpu_properties[] = {
>       DEFINE_PROP_BOOL("x-vendor-cpuid-only", X86CPU, vendor_cpuid_only, true),
>       DEFINE_PROP_BOOL("lmce", X86CPU, enable_lmce, false),
>       DEFINE_PROP_BOOL("l3-cache", X86CPU, enable_l3_cache, true),
> -    DEFINE_PROP_BOOL("kvm-no-smi-migration", X86CPU, kvm_no_smi_migration,
> -                     false),
>       DEFINE_PROP_BOOL("kvm-pv-enforce-cpuid", X86CPU, kvm_pv_enforce_cpuid,
>                        false),
>       DEFINE_PROP_BOOL("vmware-cpuid-freq", X86CPU, vmware_cpuid_freq, true),
> diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c
> index 42970ab046..571cbbf1fc 100644
> --- a/target/i386/kvm/kvm.c
> +++ b/target/i386/kvm/kvm.c
> @@ -4344,12 +4344,6 @@ static int kvm_put_vcpu_events(X86CPU *cpu, int level)
>               events.smi.pending = 0;
>               events.smi.latched_init = 0;
>           }
> -        /* Stop SMI delivery on old machine types to avoid a reboot
> -         * on an inward migration of an old VM.
> -         */
> -        if (!cpu->kvm_no_smi_migration) {
> -            events.flags |= KVM_VCPUEVENT_VALID_SMM;
> -        }

Shouldn't it be the other way round, i.e. that the flag is now always set?

pc_compat_2_3[] had:

     { TYPE_X86_CPU, "kvm-no-smi-migration", "on" },

... so I think kvm_no_smi_migration was set to true for the old machines?

  Thomas