From: Julia Cartwright <julia@ni.com>
Date: Thu, 26 Apr 2018 15:02:03 -0500
Subject: seqlock: provide the same ordering semantics as mainline
Git-repo: git://git.kernel.org/pub/scm/linux/kernel/git/rt/linux-rt-devel.git
Git-commit: e28115a132eb76fc0b95cb4928e7e227372fdff4
Patch-mainline: Queued in subsystem maintainer repository
References: SLE Realtime Extension

[ Upstream commit afa4c06b89a3c0fb7784ff900ccd707bef519cb7 ]

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().

Fixes: e69f15cf77c23 ("seqlock: Prevent rt starvation")
Cc: stable-rt@vger.kernel.org
Signed-off-by: Julia Cartwright <julia@ni.com>
Signed-off-by: Sebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: Steven Rostedt (VMware) <rostedt@goodmis.org>
Signed-off-by: Mike Galbraith <mgalbraith@suse.de>
---
 include/linux/seqlock.h |    1 +
 1 file changed, 1 insertion(+)

--- a/include/linux/seqlock.h
+++ b/include/linux/seqlock.h
@@ -461,6 +461,7 @@ static inline unsigned read_seqbegin(seq
 		spin_unlock_wait(&sl->lock);
 		goto repeat;
 	}
+	smp_rmb();
 	return ret;
 }
 #endif