From: "Paul E. McKenney" <paulmck@linux.vnet.ibm.com>
To: Peter Zijlstra <peterz@infradead.org>
Cc: Andrea Arcangeli <aarcange@redhat.com>,
Avi Kivity <avi@redhat.com>, Thomas Gleixner <tglx@linutronix.de>,
Rik van Riel <riel@redhat.com>, Ingo Molnar <mingo@elte.hu>,
akpm@linux-foundation.org,
Linus Torvalds <torvalds@linux-foundation.org>,
linux-kernel@vger.kernel.org, linux-arch@vger.kernel.org,
Benjamin Herrenschmidt <benh@kernel.crashing.org>,
David Miller <davem@davemloft.net>,
Hugh Dickins <hugh.dickins@tiscali.co.uk>,
Mel Gorman <mel@csn.ul.ie>, Nick Piggin <npiggin@suse.de>
Subject: Re: [PATCH 13/13] mm: Optimize page_lock_anon_vma
Date: Fri, 9 Apr 2010 12:22:49 -0700 [thread overview]
Message-ID: <20100409192249.GD2421@linux.vnet.ibm.com> (raw)
In-Reply-To: <1270802129.20295.3269.camel@laptop>
On Fri, Apr 09, 2010 at 10:35:29AM +0200, Peter Zijlstra wrote:
> On Thu, 2010-04-08 at 15:18 -0700, Paul E. McKenney wrote:
> > On Thu, Apr 08, 2010 at 09:17:50PM +0200, Peter Zijlstra wrote:
> > > Optimize page_lock_anon_vma() by removing the atomic ref count
> > > ops from the fast path.
> > >
> > > Rather complicates the code a lot, but might be worth it.
> >
> > Some questions and a disclaimer below.
> >
> > > Signed-off-by: Peter Zijlstra <a.p.zijlstra@chello.nl>
> > > ---
> > > mm/rmap.c | 71 ++++++++++++++++++++++++++++++++++++++++++++++++++++++++++----
> > > 1 file changed, 67 insertions(+), 4 deletions(-)
> > >
> > > Index: linux-2.6/mm/rmap.c
> > > ===================================================================
> > > --- linux-2.6.orig/mm/rmap.c
> > > +++ linux-2.6/mm/rmap.c
> > > @@ -78,6 +78,12 @@ static inline struct anon_vma *anon_vma_
> > > void anon_vma_free(struct anon_vma *anon_vma)
> > > {
> > > VM_BUG_ON(atomic_read(&anon_vma->ref));
> > > + /*
> > > + * Sync against the anon_vma->lock, so that we can hold the
> > > + * lock without requiring a reference. See page_lock_anon_vma().
> > > + */
> > > + mutex_lock(&anon_vma->lock);
> >
> > On some systems, the CPU is permitted to pull references into the critical
> > section from either side. So, do we also need an smp_mb() here?
> >
> > > + mutex_unlock(&anon_vma->lock);
> >
> > So, a question...
> >
> > Can the above mutex be contended? If yes, what happens when the
> > competing mutex_lock() acquires the lock at this point? Or, worse yet,
> > after the kmem_cache_free()?
> >
> > If no, what do we accomplish by acquiring the lock?
>
> The thing we gain is that when the holder of the lock finds a !0
> refcount it knows it can't go away because any free will first wait to
> acquire the lock.
OK. Here is the sequence of events that I am concerned about:
1. CPU 0 invokes page_lock_anon_vma() [13/13], and executes the
assignment to anon_vma. It has not yet attempted to
acquire the anon_vma->lock mutex.
2. CPU 1 invokes page_unlock_anon_vma() [13/13], which in turn
calls anon_vma_put() [5/13], which atomically decrements
->ref, finds it zero, invokes anon_vma_free() [13/13], which
finds ->ref still zero, so acquires ->lock and immediately
releases it, and then calls kmem_cache_free().
3. This kmem_cache does have SLAB_DESTROY_BY_RCU, so this
anon_vma structure will remain an anon_vma for as long as
CPU 0 remains in its RCU read-side critical section.
4. CPU 2 allocates an anon_vma, and gets the one that
CPU 0 just freed. It initializes it and makes ->ref
non-zero.
5. CPU 0 continues executing page_lock_anon_vma(), and therefore
invokes mutex_trylock() on a now-reused struct anon_vma.
It finds ->ref nonzero, so increments it and continues using
it, despite its having been reallocated, possibly to some
other process.
Or am I missing a step? (Extremely possible, as I am not as familiar
with this code as I might be.)
> > If the above mutex can be contended, can we fix by substituting
> > synchronize_rcu_expedited()? Which will soon require some scalability
> > attention if it gets used here, but what else is new? ;-)
>
> No, synchronize_rcu_expedited() will not work here, there is no RCU read
> side that covers the full usage of the anon_vma (there can't be, it
> needs to sleep).
Got it, apologies for my confusion.
> > > kmem_cache_free(anon_vma_cachep, anon_vma);
> > > }
> > >
> > > @@ -291,7 +297,7 @@ void __init anon_vma_init(void)
> > >
> > > /*
> > > * Getting a lock on a stable anon_vma from a page off the LRU is
> > > - * tricky: page_lock_anon_vma relies on RCU to guard against the races.
> > > + * tricky: anon_vma_get relies on RCU to guard against the races.
> > > */
> > > struct anon_vma *anon_vma_get(struct page *page)
> > > {
> > > @@ -320,12 +326,70 @@ out:
> > > return anon_vma;
> > > }
> > >
> > > +/*
> > > + * Similar to anon_vma_get(), however it relies on the anon_vma->lock
> > > + * to pin the object. However since we cannot wait for the mutex
> > > + * acquisition inside the RCU read lock, we use the ref count
> > > + * in the slow path.
> > > + */
> > > struct anon_vma *page_lock_anon_vma(struct page *page)
> > > {
> > > - struct anon_vma *anon_vma = anon_vma_get(page);
> > > + struct anon_vma *anon_vma = NULL;
> > > + unsigned long anon_mapping;
> > > +
> > > +again:
> > > + rcu_read_lock();
> >
> > This is interesting. You have an RCU read-side critical section with
> > no rcu_dereference().
> >
> > This strange state of affairs is actually legal (assuming that
> > anon_mapping is the RCU-protected structure) because all dereferences
> > of the anon_vma variable are atomic operations that guarantee ordering
> > (the mutex_trylock() and the atomic_inc_not_zero().
> >
> > The other dereferences (the atomic_read()s) are under the lock, so
> > are also OK assuming that the lock is held when initializing and
> > updating these fields, and even more OK due to the smp_rmb() below.
> >
> > But see below.
>
> Right so the only thing rcu_read_lock() does here is create the
> guarantee that anon_vma is safe to dereference (it lives on a
> SLAB_DESTROY_BY_RCU slab).
>
> But yes, I suppose that page->mapping read that now uses ACCESS_ONCE()
> would actually want to be an rcu_dereference(), since that both provides
> the ACCESS_ONCE() as the read-depend barrier that I thing would be
> needed.
Ah, I was getting the wrong access. Now that I see it, yes, this is
tied to the access of page->mapping that is assigned to anon_mapping.
> > > + anon_mapping = (unsigned long) ACCESS_ONCE(page->mapping);
> > > + if ((anon_mapping & PAGE_MAPPING_FLAGS) != PAGE_MAPPING_ANON)
> > > + goto unlock;
> > > + if (!page_mapped(page))
> > > + goto unlock;
> > > +
> > > + anon_vma = (struct anon_vma *) (anon_mapping - PAGE_MAPPING_ANON);
> > > + if (!mutex_trylock(&anon_vma->lock)) {
> > > + /*
> > > + * We failed to acquire the lock, take a ref so we can
> > > + * drop the RCU read lock and sleep on it.
> > > + */
> > > + if (!atomic_inc_not_zero(&anon_vma->ref)) {
> > > + /*
> > > + * Failed to get a ref, we're dead, bail.
> > > + */
> > > + anon_vma = NULL;
> > > + goto unlock;
> > > + }
> > > + rcu_read_unlock();
> > >
> > > - if (anon_vma)
> > > mutex_lock(&anon_vma->lock);
> > > + /*
> > > + * We got the lock, drop the temp. ref, if it was the last
> > > + * one free it and bail.
> > > + */
> > > + if (atomic_dec_and_test(&anon_vma->ref)) {
> > > + mutex_unlock(&anon_vma->lock);
> > > + anon_vma_free(anon_vma);
> > > + anon_vma = NULL;
> > > + }
> > > + goto out;
> > > + }
> > > + /*
> > > + * Got the lock, check we're still alive. Seeing a ref
> > > + * here guarantees the object will stay alive due to
> > > + * anon_vma_free() syncing against the lock we now hold.
> > > + */
> > > + smp_rmb(); /* Order against anon_vma_put() */
> >
> > This is ordering the fetch into anon_vma against the atomic_read() below?
> > If so, smp_read_barrier_depends() will cover it more cheaply. Alternatively,
> > use rcu_dereference() when fetching into anon_vma.
> >
> > Or am I misunderstanding the purpose of this barrier?
>
> Yes, it is:
>
> atomic_dec_and_test(&anon_vma->ref) /* implies mb */
>
> smp_rmb();
> atomic_read(&anon_vma->ref);
>
> > (Disclaimer: I have not yet found anon_vma_put(), so I am assuming that
> > anon_vma_free() plays the role of a grace period.)
>
> Yes, that lives in one of the other patches (does not exist in
> mainline).
Thank you -- and yes, I should have thought to search the patch set.
Thanx, Paul
next prev parent reply other threads:[~2010-04-09 19:23 UTC|newest]
Thread overview: 96+ messages / expand[flat|nested] mbox.gz Atom feed top
2010-04-08 19:17 [PATCH 00/13] mm: preemptibility -v2 Peter Zijlstra
2010-04-08 19:17 ` [PATCH 01/13] powerpc: Add rcu_read_lock() to gup_fast() implementation Peter Zijlstra
2010-04-08 20:31 ` Rik van Riel
2010-04-09 3:11 ` Nick Piggin
2010-04-13 1:05 ` Benjamin Herrenschmidt
2010-04-13 3:43 ` Paul E. McKenney
2010-04-14 13:51 ` Peter Zijlstra
2010-04-15 14:28 ` Paul E. McKenney
2010-04-16 6:54 ` Benjamin Herrenschmidt
2010-04-16 13:43 ` Paul E. McKenney
2010-04-16 23:25 ` Benjamin Herrenschmidt
2010-04-16 13:51 ` Peter Zijlstra
2010-04-16 14:17 ` Paul E. McKenney
2010-04-16 14:23 ` Peter Zijlstra
2010-04-16 14:32 ` Paul E. McKenney
2010-04-16 14:56 ` Peter Zijlstra
2010-04-16 15:09 ` Paul E. McKenney
2010-04-16 15:14 ` Peter Zijlstra
2010-04-16 16:45 ` Paul E. McKenney
2010-04-16 19:37 ` Peter Zijlstra
2010-04-16 20:28 ` Paul E. McKenney
2010-04-18 3:06 ` James Bottomley
2010-04-18 13:55 ` Paul E. McKenney
2010-04-18 18:55 ` James Bottomley
2010-04-16 6:51 ` Benjamin Herrenschmidt
2010-04-16 8:18 ` Nick Piggin
2010-04-16 8:29 ` Benjamin Herrenschmidt
2010-04-16 9:22 ` Nick Piggin
2010-04-08 19:17 ` [PATCH 02/13] mm: Revalidate anon_vma in page_lock_anon_vma() Peter Zijlstra
2010-04-08 20:50 ` Rik van Riel
2010-04-08 21:20 ` Andrew Morton
2010-04-08 21:54 ` Peter Zijlstra
2010-04-09 2:19 ` KOSAKI Motohiro
2010-04-09 2:19 ` Minchan Kim
2010-04-09 3:16 ` Nick Piggin
2010-04-09 4:56 ` KAMEZAWA Hiroyuki
2010-04-09 6:34 ` KOSAKI Motohiro
2010-04-09 6:47 ` KAMEZAWA Hiroyuki
2010-04-09 7:29 ` KOSAKI Motohiro
2010-04-09 7:57 ` KAMEZAWA Hiroyuki
2010-04-09 8:03 ` KAMEZAWA Hiroyuki
2010-04-09 8:24 ` KAMEZAWA Hiroyuki
2010-04-09 8:01 ` Minchan Kim
2010-04-09 8:17 ` KOSAKI Motohiro
2010-04-09 14:41 ` mlock and pageout race? Minchan Kim
2010-04-09 8:44 ` [PATCH 02/13] mm: Revalidate anon_vma in page_lock_anon_vma() Peter Zijlstra
2010-05-24 19:32 ` Andrew Morton
2010-05-25 9:01 ` Peter Zijlstra
2010-04-09 12:57 ` Peter Zijlstra
2010-04-08 19:17 ` [PATCH 03/13] x86: Remove last traces of quicklist usage Peter Zijlstra
2010-04-08 20:51 ` Rik van Riel
2010-04-08 19:17 ` [PATCH 04/13] mm: Move anon_vma ref out from under CONFIG_KSM Peter Zijlstra
2010-04-09 12:35 ` Rik van Riel
2010-04-08 19:17 ` [PATCH 05/13] mm: Make use of the anon_vma ref count Peter Zijlstra
2010-04-09 7:04 ` Christian Ehrhardt
2010-04-09 9:57 ` Peter Zijlstra
2010-04-08 19:17 ` [PATCH 06/13] mm: Preemptible mmu_gather Peter Zijlstra
2010-04-09 3:25 ` Nick Piggin
2010-04-09 8:18 ` Peter Zijlstra
2010-04-09 20:36 ` Peter Zijlstra
2010-04-19 19:16 ` Peter Zijlstra
2010-04-08 19:17 ` [PATCH 07/13] powerpc: " Peter Zijlstra
2010-04-09 4:07 ` Nick Piggin
2010-04-09 8:14 ` Peter Zijlstra
2010-04-09 8:46 ` Nick Piggin
2010-04-09 9:22 ` Peter Zijlstra
2010-04-13 2:06 ` Benjamin Herrenschmidt
2010-04-13 1:56 ` Benjamin Herrenschmidt
2010-04-13 1:23 ` Benjamin Herrenschmidt
2010-04-13 10:22 ` Peter Zijlstra
2010-04-14 13:34 ` Peter Zijlstra
2010-04-14 13:51 ` Peter Zijlstra
2010-04-08 19:17 ` [PATCH 08/13] sparc: " Peter Zijlstra
2010-04-08 19:17 ` [PATCH 09/13] mm, powerpc: Move the RCU page-table freeing into generic code Peter Zijlstra
2010-04-09 3:35 ` Nick Piggin
2010-04-09 8:08 ` Peter Zijlstra
2010-04-08 19:17 ` [PATCH 10/13] lockdep, mutex: Provide mutex_lock_nest_lock Peter Zijlstra
2010-04-09 15:36 ` Rik van Riel
2010-04-08 19:17 ` [PATCH 11/13] mutex: Provide mutex_is_contended Peter Zijlstra
2010-04-09 15:37 ` Rik van Riel
2010-04-08 19:17 ` [PATCH 12/13] mm: Convert i_mmap_lock and anon_vma->lock to mutexes Peter Zijlstra
2010-04-08 19:17 ` [PATCH 13/13] mm: Optimize page_lock_anon_vma Peter Zijlstra
2010-04-08 22:18 ` Paul E. McKenney
2010-04-09 8:35 ` Peter Zijlstra
2010-04-09 19:22 ` Paul E. McKenney [this message]
2010-04-08 20:29 ` [PATCH 00/13] mm: preemptibility -v2 David Miller
2010-04-08 20:35 ` Peter Zijlstra
2010-04-09 1:00 ` David Miller
2010-04-09 4:14 ` Nick Piggin
2010-04-09 8:35 ` Peter Zijlstra
2010-04-09 8:50 ` Nick Piggin
2010-04-09 8:58 ` Peter Zijlstra
2010-04-09 8:58 ` Martin Schwidefsky
2010-04-09 9:53 ` Peter Zijlstra
2010-04-09 9:03 ` David Howells
2010-04-09 9:22 ` Peter Zijlstra
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20100409192249.GD2421@linux.vnet.ibm.com \
--to=paulmck@linux.vnet.ibm.com \
--cc=aarcange@redhat.com \
--cc=akpm@linux-foundation.org \
--cc=avi@redhat.com \
--cc=benh@kernel.crashing.org \
--cc=davem@davemloft.net \
--cc=hugh.dickins@tiscali.co.uk \
--cc=linux-arch@vger.kernel.org \
--cc=linux-kernel@vger.kernel.org \
--cc=mel@csn.ul.ie \
--cc=mingo@elte.hu \
--cc=npiggin@suse.de \
--cc=peterz@infradead.org \
--cc=riel@redhat.com \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
* If your mail client supports setting the In-Reply-To header
via mailto: links, try the mailto: link
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox;
as well as URLs for NNTP newsgroup(s).