Re: [RFC] pte equality tests

[Jeff Dike]

> > Your analysis is correct (it can/does happen), however I don't
> > think it happens often enough to worry about it.
> ^^^^^^^^^^^^^^^^^^^^^^
> ^^^^^^^^^^^^^^^^^^^^^^^^^^^ i hope the above attitude never becomes
> popular in linux as it seems to be better suited to many MS products 

wrt correctness, you're right, but I was suggesting a slight optimization, for 
which "how often does it happen?" is the right question.

As Andrea pointed out, the more complicated test that I was suggesting would 
slow down the common case more than it helped in the rare case, making it a 
pessimization.

				Jeff