Perhaps not entirely unexpected the work there has turned up a few
further items which want dealing with. Most if not all of these
aren't interdependent, so could probably be looked at (and go in)
in (about) any order.

There's one more change I'd like to make, but I didn't get around
to making (trying to make) yet: PGC_page_table being a shadow-only
concept, it can apparently simply be 0 when !SHADOW_PAGING.

1: shadow: drop shadow_prepare_page_type_change()'s 3rd parameter
2: shadow: properly handle get_page() failing
3: mm: enforce return value checking on get_page()
4: shadow: exclude HVM-only code from sh_remove_shadows() when !HVM
5: shadow: don't open-code shadow_remove_all_shadows()
6: shadow: drop CONFIG_HVM conditionals from sh_update_cr3()
7: mm: adjust type check around _get_page_type()'s TLB flush
8: mm: re-arrange type check around _get_page_type()'s TLB flush

Jan

At 17:58 +0200 on 26 Jul (1658858332), Jan Beulich wrote:
> Perhaps not entirely unexpected the work there has turned up a few
> further items which want dealing with. Most if not all of these
> aren't interdependent, so could probably be looked at (and go in)
> in (about) any order.

Shadow parts Acked-by: Tim Deegan <tim@xen.org>

Cheers,

Tim.