Re: PATCH? rcu_do_batch: fix a pure theoretical memory ordering race

[Oleg Nesterov <oleg@tv-sign.ru>]

On 12/03, Eric Dumazet wrote:
>
> Oleg Nesterov a ?crit :
> >On top of rcu-add-a-prefetch-in-rcu_do_batch.patch
> >
> >rcu_do_batch:
> >
> >	struct rcu_head *next, *list;
> >
> >	while (list) {
> >		next = list->next;	<------ [1]
> >		list->func(list);
> >		list = next;
> >	}
> >
> >We can't trust *list after list->func() call, that is why we load 
> >list->next
> >beforehand. However I suspect in theory this is not enough, suppose that
> >
> >	- [1] is stalled
> >
> >	- list->func() marks *list as unused in some way
> >
> >	- another CPU re-uses this rcu_head and dirties it
> >
> >	- [1] completes and gets a wrong result
> >
> >This means we need a barrier in between. mb() looks more suitable, but I 
> >think
> >rmb() should suffice.
> >
> 
> Well, hopefully the "list->func()" MUST do the right thing [*], so your 
> patch is not necessary.

Yes, I don't claim it is necessary, note the "pure theoretical".

> For example, most structures are freed with kfree()/kmem_cache_free() and 
> these functions MUST imply an smp_mb() [if/when exchanging data with other 
> cpus], or else many uses in the kernel should be corrected as well.

Yes, mb() is enough (wmb() isn't) and kfree()/kmem_cache_free() are ok.
And I don't know any example of "unsafe" code in that sense.

However I believe it is easy to make the code which is correct from the
RCU's API pov, but unsafe.

Oleg.