On Tue, Sep 22, 2020 at 10:25:15AM -0400, Sasha Levin wrote:
> On Mon, Sep 21, 2020 at 04:07:01PM +0100, Mark Brown wrote:

> > This is pretty much a cosmetic change - previously we were silently not
> > reading the register, this just removes the attempt to read it since we
> > added an error message in the core.

> Right, the only reason I took it is that error message - I find that
> bogus error messages have almost the same (bad) impact as real kernel
> bugs.

The point is that the error message isn't present in stable kernels so
there is no impact on any user visible behaviour.