Re: [Xen-devel] [PATCH 5/6] libxl: add libxl_vm_generation_id_set()

[Ian Campbell <Ian.Campbell@xxxxxxxxxx>]

On Tue, 2014-05-27 at 18:31 +0100, David Vrabel wrote:
> Add a function to allow a toolstack to set a domain's VM generation
> ID.
> 
> Although the specification requires that a ACPI Notify event is raised
> if the generation ID is changed, in practice this appears not to be
> necessary if the generation ID is changed while the Windows VM is
> suspended.

Do we (or did Microsoft) envisage scenarios where the ID might change
while the VM was not suspended?

> diff --git a/tools/libxl/libxl.h b/tools/libxl/libxl.h
> index c7aa817..8dfd715 100644
> --- a/tools/libxl/libxl.h
> +++ b/tools/libxl/libxl.h
> @@ -1163,6 +1163,9 @@ int libxl_flask_getenforce(libxl_ctx *ctx);
>  int libxl_flask_setenforce(libxl_ctx *ctx, int mode);
>  int libxl_flask_loadpolicy(libxl_ctx *ctx, void *policy, uint32_t size);
>  
> +int libxl_vm_generation_id_set(libxl_ctx *ctx, uint32_t domid,
> +                               const libxl_uuid *uuid);

This needs a #define LIBXL_HAVE_FOO thing as described in the comment
about compatibility near the top of the file.

Each toolstack needs to be modified to call this new function. Why is
that preferred to a flag in
libxl_domain_config/libxl_domain_restore_params which can do the right
thing by default for all toolstacks?


> +    /*
> +     * Update the ID in guest memory (if available).

Does "if available" here translate to "on restore"? (with on create
being handled by hvmloader, IOW this is the cooperation I was wondering
about on the docs patch)

Ian.


_______________________________________________
Xen-devel mailing list
Xen-devel@xxxxxxxxxxxxx
http://lists.xen.org/xen-devel