>>> On 02.06.15 at 14:47, <ross.lagerw...@citrix.com> wrote:
> +void arch_hvm_load_post(struct domain *d)
> +{
> + /* Re-enable cache flushes and flush the cache. */
> + this_cpu(memory_type_changed_ignore) = 0;
> + memory_type_changed(d);
> +}
Does this really need to be done unconditionally? I.e. couldn't this be
a tristate, with memory_type_changed() e.g. flipping its sign when it
gets actually called?
Also, are we certain that memory_type_changed() will never be
called from asynchronous code (i.e. this doesn't introduce a latent,
hard to debug issue)?
Jan
_______________________________________________
Xen-devel mailing list
Xen-devel@lists.xen.org
http://lists.xen.org/xen-devel
