George Dunlap
2010-Jan-06 12:19 UTC
[Xen-devel] [PATCH] Handle PoD case in hvm_hap_nested_page_fault()
The new combined nested page fault handling doesn''t consider the case where the gfn_to_mfn() translation caused the page to be transparently populated. Attached patch is one solution; a more robust solution would be to take the access type (read or write) and check it against the type returned by gfn_to_mfn() to handle spurious NPFs as well (if such things ever happen). Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com> _______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel
Keir Fraser
2010-Jan-06 12:49 UTC
Re: [Xen-devel] [PATCH] Handle PoD case in hvm_hap_nested_page_fault()
I checked in a different version as c/s 20763: take a look. It''s questionable whether returning 0 from the function is ever a sensible thing. I suppose at least this way we have to have considered all cases properly. At least I think we won''t change it during the 4.0.0 freeze. -- Keir On 06/01/2010 12:19, "George Dunlap" <dunlapg@umich.edu> wrote:> The new combined nested page fault handling doesn''t consider the case > where the gfn_to_mfn() translation caused the page to be transparently > populated. Attached patch is one solution; a more robust solution > would be to take the access type (read or write) and check it against > the type returned by gfn_to_mfn() to handle spurious NPFs as well (if > such things ever happen). > > Signed-off-by: George Dunlap <george.dunlap@eu.citrix.com>_______________________________________________ Xen-devel mailing list Xen-devel@lists.xensource.com http://lists.xensource.com/xen-devel