Re: [PATCH] whpx: Added support for saving/restoring VM state

[Paolo Bonzini]

On 5/14/22 03:29, Ivan Shcherbakov wrote:
> +/*
> + * As of Windows 10 21H1, the layout of the XSAVE data returned by the WHPX
> API
> + * does not match the layout used by QEMU.
> + *
> + * Specifically, trying to pass the state returned by
> x86_cpu_xsave_all_areas()
> + * to WHvSetVirtualProcessorXsaveState() causes it to return an error.
> + *
> + * As a result, we do not reflect the captured XSAVE state in the X86CPU
> + * structure, and instead manually save it to the snapshots via the
> + * whpx_xsave_xxxx() callbacks.
> + *
> + * Note that unlike the device drivers that can use the new
> VMStateDescription
> + * mechanism via 'DeviceClass::vmsd' field, AccelClass objects cannot
> easily
> + * do it. Hence, we rely on the legacy state management API.
> + */
> +static SaveVMHandlers savevm_whpx = {
> +    .save_state = whpx_xsave_save,
> +    .load_state = whpx_xsave_load,
> +};
> +

What are the differences?  Is it using the XSAVEC/XSAVES ("compacted") 
format?

Paolo