On 2018-04-26 15:02:03 [-0500], Julia Cartwright wrote:

The mainline implementation of read_seqbegin() orders prior loads w.r.t.
the read-side critical section.  Fixup the RT writer-boosting
implementation to provide the same guarantee.

Also, while we're here, update the usage of ACCESS_ONCE() to use
READ_ONCE().

I'm taking this without the READ_ONCE hunk because ACCESS_ONCE is gone
since v4.15 and I had to move on in v4.16.

Fixes: e69f15cf77c23 ("seqlock: Prevent rt starvation")
Cc: stable-rt@vger.kernel.org
Signed-off-by: Julia Cartwright <redacted>

Sebastian