Re: [Qemu-devel] [PATCH resend] kvmclock: fix clock_is_reliable on migration from QEMU < 2.9

[Paolo Bonzini]

On 06/04/2018 07:34, Michael Chapman wrote:
> When migrating from a pre-2.9 QEMU, no clock_is_reliable flag is
> transferred. We should assume that the source host has an unreliable
> KVM_GET_CLOCK, rather than using whatever was determined locally, to
> ensure that any drift from the TSC-based value calculated by the guest
> is corrected.
> 
> Signed-off-by: Michael Chapman <address@hidden>
> ---
> Resent due to missing CCs.
> 
>  hw/i386/kvm/clock.c | 14 ++++++++++++++
>  1 file changed, 14 insertions(+)
> 
> diff --git a/hw/i386/kvm/clock.c b/hw/i386/kvm/clock.c
> index 1707434db3..7dac319403 100644
> --- a/hw/i386/kvm/clock.c
> +++ b/hw/i386/kvm/clock.c
> @@ -241,6 +241,19 @@ static const VMStateDescription 
> kvmclock_reliable_get_clock = {
>      }
>  };
>  
> +/*
> + * When migrating, assume the source has an unreliable
> + * KVM_GET_CLOCK unless told otherwise.
> + */
> +static int kvmclock_pre_load(void *opaque)
> +{
> +    KVMClockState *s = opaque;
> +
> +    s->clock_is_reliable = false;
> +
> +    return 0;
> +}
> +
>  /*
>   * When migrating, read the clock just before migration,
>   * so that the guest clock counts during the events
> @@ -268,6 +281,7 @@ static const VMStateDescription kvmclock_vmsd = {
>      .name = "kvmclock",
>      .version_id = 1,
>      .minimum_version_id = 1,
> +    .pre_load = kvmclock_pre_load,
>      .pre_save = kvmclock_pre_save,
>      .fields = (VMStateField[]) {
>          VMSTATE_UINT64(clock, KVMClockState),
> 

Queued, thanks.

Paolo