Re: [PATCH v3] i386: Add ratelimit for bus locks acquired in guest

[Eduardo Habkost]

+Stefan

I have a question about ratelimit_set_speed() below:

On Fri, Apr 30, 2021 at 06:33:05PM +0800, Chenyi Qiang wrote:
> A bus lock is acquired through either split locked access to writeback
> (WB) memory or any locked access to non-WB memory. It is typically >1000
> cycles slower than an atomic operation within a cache and can also
> disrupts performance on other cores.
> 
> Virtual Machines can exploit bus locks to degrade the performance of
> system. To address this kind of performance DOS attack coming from the
> VMs, bus lock VM exit is introduced in KVM and it can report the bus
> locks detected in guest. If enabled in KVM, it would exit to the
> userspace to let the user enforce throttling policies once bus locks
> acquired in VMs.
> 
> The availability of bus lock VM exit can be detected through the
> KVM_CAP_X86_BUS_LOCK_EXIT. The returned bitmap contains the potential
> policies supported by KVM. The field KVM_BUS_LOCK_DETECTION_EXIT in
> bitmap is the only supported strategy at present. It indicates that KVM
> will exit to userspace to handle the bus locks.
> 
> This patch adds a ratelimit on the bus locks acquired in guest as a
> mitigation policy.
> 
> Introduce a new field "bus_lock_ratelimit" to record the limited speed
> of bus locks in the target VM. The user can specify it through the
> "bus-lock-ratelimit" as a machine property. In current implementation,
> the default value of the speed is 0 per second, which means no
> restrictions on the bus locks
> 
> As for ratelimit on detected bus locks, simply set the ratelimit
> interval to 1s and restrict the quota of bus lock occurence to the value
> of "bus_lock_ratelimit". A potential alternative is to introduce the
> time slice as a property which can help the user achieve more precise
> control.
> 
> The detail of Bus lock VM exit can be found in spec:
> https://software.intel.com/content/www/us/en/develop/download/intel-architecture-instruction-set-extensions-programming-reference.html
> 
> Signed-off-by: Chenyi Qiang <chenyi.qiang@intel.com>
> 
> ---
> Changes from v2:
>   - do some rename work (bus-lock-ratelimit and BUS_LOCK_TIME_SLICE).
>     (Eduardo)
>   - change to register a class property at the x86_machine_class_init()
>     and write the gettter/setter for the bus_lock_ratelimit property.
>     (Eduardo)
>   - add the lock to access the Ratelimit instance to avoid vcpu thread
>     race condition. (Eduardo)
>   - v2: 
> https://lore.kernel.org/qemu-devel/20210420093736.17613-1-chenyi.qiang@intel.com/
> 
> Changes from RFC v1:
>   - Remove the rip info output, as the rip can't reflect the bus lock
>     position correctly. (Xiaoyao)
>   - RFC v1: 
> https://lore.kernel.org/qemu-devel/20210317084709.15605-1-chenyi.qiang@intel.com/
[...]
> diff --git a/include/hw/i386/x86.h b/include/hw/i386/x86.h
> index c09b648dff..49b130a649 100644
> --- a/include/hw/i386/x86.h
> +++ b/include/hw/i386/x86.h
> @@ -74,12 +74,21 @@ struct X86MachineState {
>       * will be translated to MSI messages in the address space.
>       */
>      AddressSpace *ioapic_as;
> +
> +    /*
> +     * Ratelimit enforced on detected bus locks in guest.
> +     * The default value of the bus_lock_ratelimit is 0 per second,
> +     * which means no limitation on the guest's bus locks.
> +     */
> +    uint64_t bus_lock_ratelimit;
> +    RateLimit bus_lock_ratelimit_ctrl;
>  };
>  
>  #define X86_MACHINE_SMM              "smm"
>  #define X86_MACHINE_ACPI             "acpi"
>  #define X86_MACHINE_OEM_ID           "x-oem-id"
>  #define X86_MACHINE_OEM_TABLE_ID     "x-oem-table-id"
> +#define X86_MACHINE_BUS_LOCK_RATELIMIT  "bus-lock-ratelimit"
>  
>  #define TYPE_X86_MACHINE   MACHINE_TYPE_NAME("x86")
>  OBJECT_DECLARE_TYPE(X86MachineState, X86MachineClass, X86_MACHINE)
> diff --git a/target/i386/kvm/kvm.c b/target/i386/kvm/kvm.c
> index 7fe9f52710..19b6c4a7e8 100644
> --- a/target/i386/kvm/kvm.c
> +++ b/target/i386/kvm/kvm.c
> @@ -130,6 +130,9 @@ static bool has_msr_mcg_ext_ctl;
>  static struct kvm_cpuid2 *cpuid_cache;
>  static struct kvm_msr_list *kvm_feature_msrs;
>  
> +#define BUS_LOCK_SLICE_TIME 1000000000ULL /* ns */
> +static QemuMutex bus_lock_ratelimit_lock;
> +
>  int kvm_has_pit_state2(void)
>  {
>      return has_pit_state2;
> @@ -2267,6 +2270,28 @@ int kvm_arch_init(MachineState *ms, KVMState *s)
>          }
>      }
>  
> +    if (object_dynamic_cast(OBJECT(ms), TYPE_X86_MACHINE)) {
> +        X86MachineState *x86ms = X86_MACHINE(ms);
> +
> +        if (x86ms->bus_lock_ratelimit > 0) {
> +            ret = kvm_check_extension(s, KVM_CAP_X86_BUS_LOCK_EXIT);
> +            if (!(ret & KVM_BUS_LOCK_DETECTION_EXIT)) {
> +                error_report("kvm: bus lock detection unsupported");
> +                return -ENOTSUP;
> +            }
> +            ret = kvm_vm_enable_cap(s, KVM_CAP_X86_BUS_LOCK_EXIT, 0,
> +                                    KVM_BUS_LOCK_DETECTION_EXIT);
> +            if (ret < 0) {
> +                error_report("kvm: Failed to enable bus lock detection cap: 
> %s",
> +                             strerror(-ret));
> +                return ret;
> +            }
> +            qemu_mutex_init(&bus_lock_ratelimit_lock);
> +            ratelimit_set_speed(&x86ms->bus_lock_ratelimit_ctrl, 
> x86ms->bus_lock_ratelimit,
> +                                BUS_LOCK_SLICE_TIME);
> +        }
> +    }
> +
>      return 0;
>  }
>  
> @@ -4221,6 +4246,20 @@ void kvm_arch_pre_run(CPUState *cpu, struct kvm_run 
> *run)
>      }
>  }
>  
> +static void kvm_rate_limit_on_bus_lock(void)
> +{
> +    MachineState *ms = MACHINE(qdev_get_machine());

qdev_get_machine() seems thread safe except for the first call,
but it's not documented as such.

Until it is documented as thread safe (which could take a while,
considering that there are ongoing attempts to clean it up), I
would avoid calling without the BQL, just in case.

> +    X86MachineState *x86ms = X86_MACHINE(ms);
> +
> +    qemu_mutex_lock(&bus_lock_ratelimit_lock);
> +    uint64_t delay_ns = 
> ratelimit_calculate_delay(&x86ms->bus_lock_ratelimit_ctrl, 1);
> +    qemu_mutex_unlock(&bus_lock_ratelimit_lock);

Stefan, ratelimit_calculate_delay() is supposed to be thread
safe, correct?

In that case, bus_lock_ratelimit_lock would be completely unnecessary.

I normally prefer to avoid static variables, but in this case a

   static RateLimit bus_lock_ratelimit_ctrl;

variable could be the simplest solution here.


> +
> +    if (delay_ns) {
> +        g_usleep(delay_ns / SCALE_US);
> +    }
> +}
> +
>  MemTxAttrs kvm_arch_post_run(CPUState *cpu, struct kvm_run *run)
>  {
>      X86CPU *x86_cpu = X86_CPU(cpu);
> @@ -4236,6 +4275,9 @@ MemTxAttrs kvm_arch_post_run(CPUState *cpu, struct 
> kvm_run *run)
>      } else {
>          env->eflags &= ~IF_MASK;
>      }
> +    if (run->flags & KVM_RUN_X86_BUS_LOCK) {
> +        kvm_rate_limit_on_bus_lock();
> +    }
>  
>      /* We need to protect the apic state against concurrent accesses from
>       * different threads in case the userspace irqchip is used. */
> @@ -4594,6 +4636,10 @@ int kvm_arch_handle_exit(CPUState *cs, struct kvm_run 
> *run)
>          ioapic_eoi_broadcast(run->eoi.vector);
>          ret = 0;
>          break;
> +    case KVM_EXIT_X86_BUS_LOCK:
> +        /* already handled in kvm_arch_post_run */
> +        ret = 0;
> +        break;
>      default:
>          fprintf(stderr, "KVM: unknown exit reason %d\n", run->exit_reason);
>          ret = -1;
> -- 
> 2.17.1
> 

-- 
Eduardo