Marcelo Tosatti wrote:
> And this is against a changed x86.git -mm tree (with pvops64 patches).
> I'll send the PTE-write-via-hypercall patches soon and will rebase on
> top of that (the CR3 cache needs more testing/tuning apparently).

Oops for sale ;)

Triggered by guests wrmsr, looks like some error checks are missing.

I've passed in a physical address.  The vmx_cr3_cache_msr() function has
a gva_to_page() call which makes me suspect it expects a virtual
address.  First it should not Oops when a invalid virtual address is
passed in, and second I think it better shouldn't expect a virtual
address in the first place.

What is the reason to expect the cr3 cache being page aligned btw?  It
should be enougth to require the struct not cross a page border, right?

cheers,
  Gerd