From: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx> [ Upstream commit 2840cf02fae627860156737e83326df354ee4ec6 ] When the prev and next task's mm change, switch_mm() provides the core serializing guarantees before returning to usermode. The only case where an explicit core serialization is needed is when the scheduler keeps the same mm for prev and next. Suggested-by: Oleg Nesterov <oleg@xxxxxxxxxx> Signed-off-by: Mathieu Desnoyers <mathieu.desnoyers@xxxxxxxxxxxx> Signed-off-by: Peter Zijlstra (Intel) <peterz@xxxxxxxxxxxxx> Cc: Chris Metcalf <cmetcalf@xxxxxxxxxx> Cc: Christoph Lameter <cl@xxxxxxxxx> Cc: Eric W. Biederman <ebiederm@xxxxxxxxxxxx> Cc: Kirill Tkhai <tkhai@xxxxxxxxx> Cc: Linus Torvalds <torvalds@xxxxxxxxxxxxxxxxxxxx> Cc: Mike Galbraith <efault@xxxxxx> Cc: Paul E. McKenney <paulmck@xxxxxxxxxxxxx> Cc: Peter Zijlstra <peterz@xxxxxxxxxxxxx> Cc: Russell King - ARM Linux admin <linux@xxxxxxxxxxxxxxx> Cc: Thomas Gleixner <tglx@xxxxxxxxxxxxx> Link: https://lkml.kernel.org/r/20190919173705.2181-4-mathieu.desnoyers@xxxxxxxxxxxx Signed-off-by: Ingo Molnar <mingo@xxxxxxxxxx> Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx> --- include/linux/sched/mm.h | 2 ++ 1 file changed, 2 insertions(+) diff --git a/include/linux/sched/mm.h b/include/linux/sched/mm.h index 4a7944078cc35..8557ec6642130 100644 --- a/include/linux/sched/mm.h +++ b/include/linux/sched/mm.h @@ -362,6 +362,8 @@ enum { static inline void membarrier_mm_sync_core_before_usermode(struct mm_struct *mm) { + if (current->mm != mm) + return; if (likely(!(atomic_read(&mm->membarrier_state) & MEMBARRIER_STATE_PRIVATE_EXPEDITED_SYNC_CORE))) return; -- 2.20.1
