> All the discussion here[1] for example is about having ordering and > doing an smp_cond_load_acquire() on a variable which is sometimes > protected by a CPU's rq->lock and sometimes not? Isn't that one of the > key use cases for this whole discussion?
Not a "pure" one: http://lkml.kernel.org/r/1530629639-27767-1-git-send-email-andrea.pa...@amarulasolutions.com we also need "W->R ordering" in schedule()! so there better be an smp_mb__after_spinlock() or a barrier providing similar ordering. Nick was suggesting a "weaker version" of this barrier back in: 362a61ad61199e ("fix SMP data race in pagetable setup vs walking") c.f., the comment in mm/memory.c:__pte_alloc(), but that does not math our pattern (UNLOCK+LOCK), AFAICT. Andrea > > [1] https://lkml.org/lkml/2015/10/6/805 > > Dan