Messages in this thread | | | Subject | Re: [Xen-devel] [RFC PATCH v1] Consider void entries in the P2M as 1-1 mapping. | From | Ian Campbell <> | Date | Wed, 22 Dec 2010 08:36:55 +0000 |
| |
On Tue, 2010-12-21 at 21:37 +0000, Konrad Rzeszutek Wilk wrote: > In the past we used to think of those regions as "missing" and under > the ownership of the balloon code. But the balloon code only operates > on a specific region. This region is in lastE820 RAM page (basically > any region past nr_pages is considered balloon type page).
That is true at start of day but once the system is up and running the balloon driver can make a hole for anything which can be returned by alloc_page.
The following descriptions seem to consider this correctly but I just wanted to clarify.
I don't think it's necessarily the last E820 RAM page either, that's just what the tools today happen to build. In principal the tools could push down a holey e820 (e.g. with PCI holes prepunched etc) and boot the domain ballooned down such that the N-2, N-3 e820 RAM regions are above nr_pages too.
> This patchset considers the void entries as "identity" and for balloon > pages you have to set the PFNs to be "missing". This means that the > void entries are now considered 1-1, so for PFNs which exist in large > gaps of the P2M space will return the same PFN.
I would naively have expected that a missing entry indicated an invalid/missing entry rather than an identity region, it just seems like the safer default since we are (maybe) more likely to catch an INVALID_P2M_ENTRY before handing it to the hypervisor and getting ourselves shot.
In that case the identity regions would need to be explicitly registered, is that harder to do?
I guess we could register any hole or explicit non-RAM region in the e820 as identity but do we sometimes see I/O memory above the top of the e820 or is there some other problem I'm not thinking of?
> The xen/mmu.c code where it deals with _PAGE_IOMAP can be removed, but > to guard against regressions or bugs lets take it one patchset at a > time.
Could we have a WARN_ON(_PAGE_IOMAP && !PAGE_IDENTITY) (or whatever the predicates really are) in some relevant places in mmu.c?
Ian.
| |