Re: [PATCH 01/10] KVM: SVM: Don't use kmap_atomic in nested_svm_map

[Joerg Roedel <joerg.roedel@amd.com>]

On Thu, Feb 18, 2010 at 03:40:56PM +0200, Avi Kivity wrote:
> On 02/18/2010 01:38 PM, Joerg Roedel wrote:
> >Use of kmap_atomic disables preemption but if we run in
> >shadow-shadow mode the vmrun emulation executes kvm_set_cr3
> >which might sleep or fault. So use kmap instead for
> >nested_svm_map.
> >
> >
> >
> >-static void nested_svm_unmap(void *addr, enum km_type idx)
> >+static void nested_svm_unmap(void *addr)
> >  {
> >  	struct page *page;
> >
> >@@ -1443,7 +1443,7 @@ static void nested_svm_unmap(void *addr, enum km_type idx)
> >
> >  	page = kmap_atomic_to_page(addr);
> >
> >-	kunmap_atomic(addr, idx);
> >+	kunmap(addr);
> >  	kvm_release_page_dirty(page);
> >  }
> 
> kunmap() takes a struct page *, not the virtual address (a
> consistent source of bugs).

Ah true, thanks. I'll fix that.

> kmap() is generally an unloved interface, it is slow and possibly
> deadlock prone, but it's better than sleeping in atomic context.  If
> you can hack your way around it, that is preferred.

Best would be to use kvm_read_guest, but I fear that this will have an
performance impact. Maybe I'll try this and measure if it really has a
significant performance impact.

	Joerg