Re: [PATCH 2/5] rhashtable: don't hold lock on first table throughout insertion.

From: NeilBrown
Date: Mon Mar 11 2019 - 17:50:22 EST

Next message: Petr Vorel: "Re: [PATCH v3 0/7] selftests/ima: add kexec and kernel module tests"
Previous message: Sam Ravnborg: "Re: [PATCH] drm/atomic-helper: Validate pointer before dereference"
In reply to: Paul E. McKenney: "Re: [PATCH 2/5] rhashtable: don't hold lock on first table throughout insertion."
Next in thread: Paul E. McKenney: "Re: [PATCH 2/5] rhashtable: don't hold lock on first table throughout insertion."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]

On Mon, Mar 11 2019, Paul E. McKenney wrote:

> On Tue, Jul 31, 2018 at 07:44:29AM -0700, Paul E. McKenney wrote:
>> On Tue, Jul 31, 2018 at 03:04:48PM +1000, NeilBrown wrote:
>> > On Mon, Jul 30 2018, Paul E. McKenney wrote:
>> >
>> > > On Tue, Jul 31, 2018 at 10:45:45AM +1000, NeilBrown wrote:
>> > >> On Fri, Jul 27 2018, Paul E. McKenney wrote:
>> > >>
>> > >> > On Thu, Jul 26, 2018 at 08:18:15PM -0700, Paul E. McKenney wrote:
>> > >> >> On Fri, Jul 27, 2018 at 11:04:37AM +1000, NeilBrown wrote:
>> > >> >> > On Wed, Jul 25 2018, Paul E. McKenney wrote:
>> > >> >> > >>
>> > >> >> > >> Looks good ... except ... naming is hard.
>> > >> >> > >>
>> > >> >> > >> is_after_call_rcu_init() asserts where in the lifecycle we are,
>> > >> >> > >> is_after_call_rcu() tests where in the lifecycle we are.
>> > >> >> > >>
>> > >> >> > >> The names are similar but the purpose is quite different.
>> > >> >> > >> Maybe s/is_after_call_rcu_init/call_rcu_init/ ??
>> > >> >> > >
>> > >> >> > > How about rcu_head_init() and rcu_head_after_call_rcu()?
>> > >> >>
>> > >> >> Very well, I will pull this change in on my next rebase.
>> > >> >
>> > >> > Like this?
>> > >>
>> > >> Hard to say - unwinding white-space damage in my head is too challenging
>> > >> when newlines have been deleted :-(
>> > >
>> > > What??? Don't you like block-structured code?
>> > >
>> > > All kidding aside, how about the following more conventionally formatted
>> > > version?
>> >
>> > Wow - it's like I just got new glasses!
>> > Yes - nice an clear and now flaws to be found. Thanks a lot.
>>
>> Now that flaws are to be found, please feel free to report them. ;-)
>
> Hello, Neil,
>
> Any plans to use these functions? There are still no upstream uses.
> On the other hand, if they proved unuseful, I will remove them. If I
> don't hear otherwise from you, I will pull them in v5.2.

Hi Paul,
yes, I do still have plans for them. I've got quite a few things I
want to add to rhashtables including this, but got stalled late last
year and I haven't managed to get back to it.
Thanks for your prompting - I'll make an effort to post some patches
soon, particularly the one that makes use of this new functionality.

Thanks,
NeilBrown

>
> Thanx, Paul
>
>> > NeilBrown
>> >
>> > >
>> > > Thanx, Paul
>> > >
>> > > ------------------------------------------------------------------------
>> > >
>> > > commit e3408141ed7d702995b2fdc94703af88aadd226b
>> > > Author: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
>> > > Date: Tue Jul 24 15:28:09 2018 -0700
>> > >
>> > > rcu: Provide functions for determining if call_rcu() has been invoked
>> > >
>> > > This commit adds rcu_head_init() and rcu_head_after_call_rcu() functions
>> > > to help RCU users detect when another CPU has passed the specified
>> > > rcu_head structure and function to call_rcu(). The rcu_head_init()
>> > > should be invoked before making the structure visible to RCU readers,
>> > > and then the rcu_head_after_call_rcu() may be invoked from within
>> > > an RCU read-side critical section on an rcu_head structure that
>> > > was obtained during a traversal of the data structure in question.
>> > > The rcu_head_after_call_rcu() function will return true if the rcu_head
>> > > structure has already been passed (with the specified function) to
>> > > call_rcu(), otherwise it will return false.
>> > >
>> > > If rcu_head_init() has not been invoked on the rcu_head structure
>> > > or if the rcu_head (AKA callback) has already been invoked, then
>> > > rcu_head_after_call_rcu() will do WARN_ON_ONCE().
>> > >
>> > > Reported-by: NeilBrown <neilb@xxxxxxxx>
>> > > Signed-off-by: Paul E. McKenney <paulmck@xxxxxxxxxxxxxxxxxx>
>> > > [ paulmck: Apply neilb naming feedback. ]
>> > >
>> > > diff --git a/include/linux/rcupdate.h b/include/linux/rcupdate.h
>> > > index e4f821165d0b..4db8bcacc51a 100644
>> > > --- a/include/linux/rcupdate.h
>> > > +++ b/include/linux/rcupdate.h
>> > > @@ -857,6 +857,46 @@ static inline notrace void rcu_read_unlock_sched_notrace(void)
>> > > #endif /* #else #ifdef CONFIG_ARCH_WEAK_RELEASE_ACQUIRE */
>> > >
>> > >
>> > > +/* Has the specified rcu_head structure been handed to call_rcu()? */
>> > > +
>> > > +/*
>> > > + * rcu_head_init - Initialize rcu_head for rcu_head_after_call_rcu()
>> > > + * @rhp: The rcu_head structure to initialize.
>> > > + *
>> > > + * If you intend to invoke rcu_head_after_call_rcu() to test whether a
>> > > + * given rcu_head structure has already been passed to call_rcu(), then
>> > > + * you must also invoke this rcu_head_init() function on it just after
>> > > + * allocating that structure. Calls to this function must not race with
>> > > + * calls to call_rcu(), rcu_head_after_call_rcu(), or callback invocation.
>> > > + */
>> > > +static inline void rcu_head_init(struct rcu_head *rhp)
>> > > +{
>> > > + rhp->func = (rcu_callback_t)~0L;
>> > > +}
>> > > +
>> > > +/*
>> > > + * rcu_head_after_call_rcu - Has this rcu_head been passed to call_rcu()?
>> > > + * @rhp: The rcu_head structure to test.
>> > > + * @func: The function passed to call_rcu() along with @rhp.
>> > > + *
>> > > + * Returns @true if the @rhp has been passed to call_rcu() with @func,
>> > > + * and @false otherwise. Emits a warning in any other case, including
>> > > + * the case where @rhp has already been invoked after a grace period.
>> > > + * Calls to this function must not race with callback invocation. One way
>> > > + * to avoid such races is to enclose the call to rcu_head_after_call_rcu()
>> > > + * in an RCU read-side critical section that includes a read-side fetch
>> > > + * of the pointer to the structure containing @rhp.
>> > > + */
>> > > +static inline bool
>> > > +rcu_head_after_call_rcu(struct rcu_head *rhp, rcu_callback_t f)
>> > > +{
>> > > + if (READ_ONCE(rhp->func) == f)
>> > > + return true;
>> > > + WARN_ON_ONCE(READ_ONCE(rhp->func) != (rcu_callback_t)~0L);
>> > > + return false;
>> > > +}
>> > > +
>> > > +
>> > > /* Transitional pre-consolidation compatibility definitions. */
>> > >
>> > > static inline void synchronize_rcu_bh(void)
>> > > diff --git a/kernel/rcu/rcu.h b/kernel/rcu/rcu.h
>> > > index 5dec94509a7e..4c56c1d98fb3 100644
>> > > --- a/kernel/rcu/rcu.h
>> > > +++ b/kernel/rcu/rcu.h
>> > > @@ -224,6 +224,7 @@ void kfree(const void *);
>> > > */
>> > > static inline bool __rcu_reclaim(const char *rn, struct rcu_head *head)
>> > > {
>> > > + rcu_callback_t f;
>> > > unsigned long offset = (unsigned long)head->func;
>> > >
>> > > rcu_lock_acquire(&rcu_callback_map);
>> > > @@ -234,7 +235,9 @@ static inline bool __rcu_reclaim(const char *rn, struct rcu_head *head)
>> > > return true;
>> > > } else {
>> > > RCU_TRACE(trace_rcu_invoke_callback(rn, head);)
>> > > - head->func(head);
>> > > + f = head->func;
>> > > + WRITE_ONCE(head->func, (rcu_callback_t)0L);
>> > > + f(head);
>> > > rcu_lock_release(&rcu_callback_map);
>> > > return false;
>> > > }
>>
>>

Attachment: signature.asc
Description: PGP signature

Next message: Petr Vorel: "Re: [PATCH v3 0/7] selftests/ima: add kexec and kernel module tests"
Previous message: Sam Ravnborg: "Re: [PATCH] drm/atomic-helper: Validate pointer before dereference"
In reply to: Paul E. McKenney: "Re: [PATCH 2/5] rhashtable: don't hold lock on first table throughout insertion."
Next in thread: Paul E. McKenney: "Re: [PATCH 2/5] rhashtable: don't hold lock on first table throughout insertion."
Messages sorted by: [ date ] [ thread ] [ subject ] [ author ]