On 12/04/2022 09.41, Nico Boehr wrote:
On Mon, 2022-04-11 at 17:30 +0200, Thomas Huth wrote:
Thanks for tackling this!
Having written powerpc/sprs.c in the past, I've got one question /
request:
Could we turn this into a "real" test immediately? E.g. write a sane
value
to all control registers that are currently not in use by the k-u-t
before
migration, and then check whether the values are still in there after
migration? Maybe also some vector registers and the "guarded storage
control"?
... or is this rather something for a later update?
My plan was to first add the infrastructure for migration tests
including the selftest and then later one by one add "real" tests.
But if you think it is preferable, I can extend the scope and add some
inital "real" tests in this series.
I think a simple test that checks some register values should not be too
hard to implement, so I'd prefer that instead of this simple selftest ...
but if you're too short in time right now, I also won't insist.
Thomas
