On Fri, Nov 02, 2018 at 11:02:35AM -0700, Linus Torvalds wrote:

On Fri, Nov 2, 2018 at 10:21 AM Michael S. Tsirkin [off-list ref] wrote:

quoted

it seems that it depends on current not on the active mm.

Yes, see my other suggestion to just fix "use_mm()" to update the address range.

Would you mind testing that?

Sure, I'll test it.

Because that would seem to be the *much* less error-prone model..

           Linus

I agree, it's always been bothering me.

-- 
MST