Groups | Search | Server Info | Keyboard shortcuts | Login | Register [http] [https] [nntp] [nntps]
Groups > linux.kernel > #1323910
| From | Boqun Feng <boqun.feng@gmail.com> |
|---|---|
| Newsgroups | linux.kernel |
| Subject | Re: [RFC][PATCH] mips: Fix arch_spin_unlock() |
| Date | 2016-02-02 10:40 +0100 |
| Message-ID | <qXFBh-xF-23@gated-at.bofh.it> (permalink) |
| References | (5 earlier) <qXAie-4OK-9@gated-at.bofh.it> <qXBxE-5VB-7@gated-at.bofh.it> <qXCWK-6RP-9@gated-at.bofh.it> <qXEc9-80M-7@gated-at.bofh.it> <qXElQ-84M-3@gated-at.bofh.it> |
| Organization | linux.* mail to news gateway |
[Multipart message — attachments visible in raw view] - view raw
Hello Linus,
On Tue, Feb 02, 2016 at 12:19:04AM -0800, Linus Torvalds wrote:
> On Tue, Feb 2, 2016 at 12:07 AM, Linus Torvalds
> <torvalds@linux-foundation.org> wrote:
> >
> > So we *absolutely* should say that *OF COURSE* these things work:
> >
> > - CPU A:
> >
> > .. initialize data structure -> smp_wmb() -> WRITE_ONCE(ptr);
> >
> > - CPU B:
> >
> > smp_load_acquire(ptr) - we can rely on things behind "ptr" being initialized
>
> That's a bad example, btw. I shouldn't have made it be a "pointer",
> because then we get the whole address dependency chain ordering
> anyway.
>
> So instead of "ptr", read "state flag". It might just be an "int" that
> says "data has been initialized".
>
> So
>
> .. initialize memory ..
> smp_wmb();
> WRITE_ONCE(&is_initialized, 1);
>
> should pair with
>
> if (smp_load_acquire(&is_initialized))
> ... we can read and write the data, knowing it has been initialized ..
>
> exactly because "smp_wmb()" (cheap write barrier) might be cheaper
> than "smp_store_release()" (expensive full barrier) and thus
> preferred.
>
Just to be clear, what Will, Paul and I are discussing here is about
local transitivity, which refers to something like this following
example:
(a, b and is_initialized are all initially zero)
P0:
WRITE_ONCE(a, 1);
smp_store_release(&is_initialized, 1);
P1:
r1 = smp_load_acquire(&is_initialized);
smp_store_release(&b, 1);
P2:
r2 = smp_load_acquire(&b);
r3 = READ_ONCE(a);
, in which case, r1 == 1 && r2 == 1 && r3 == 0 can not happen because
RELEASE+ACQUIRE pairs guarantee local transitivity.
More on local transitvity:
http://article.gmane.org/gmane.linux.kernel.virtualization/26856
And what I'm asking here is something like this following example:
(a, b and is_initialized are all initially zero)
P0:
WRITE_ONCE(a, 1);
smp_store_release(&is_initialized, 1);
P1:
if (r1 = READ_ONCE(is_initialized))
smp_store_release(&b, 1);
P2:
if (r2 = READ_ONCE(b)) {
smp_rmb();
r3 = READ_ONCE(a);
}
, in which case, can r1 == 1 && r2 == 1 && r3 == 0 happen?
Please note this example is about two questions on local transitivity:
1. Could "READ_ONCE(); if" extend a locally transitive chain.
2. Could "READ_ONCE(); if; smp_rmb()" at least be a locally
transitive chain termination?
> So mixing ordering metaphors actually does make sense, and should be
> entirely well-defined.
>
I think Paul does agree that smp_{r,w}mb() with applicative memory
operations around could pair with smp_store_release() or
smp_load_acquire().
Hope I didn't misunderstand any of you or make you misunderstood with
each other..
Regards,
Boqun
> There's likely less reason to do it the other way (ie
> "smp_store_release()" on one side pairing with "LOAD_ONCE() +
> smp_rmb()" on the other) since there likely isn't the same kind of
> performance reason for that pairing. But even if we would never
> necessarily want to do it, I think our memory ordering rules would be
> *much* better for strongly stating that it has to work, than being
> timid and trying to make the rules weak.
>
> Memory ordering is confusing enough as it is. We should not make
> people worry more than they already have to. Strong rules are good.
>
> Linus
Back to linux.kernel | Previous | Next — Previous in thread | Next in thread | Find similar | Unroll thread
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-01-28 01:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-01-28 11:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-01-28 23:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-01-29 11:10 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-01-29 23:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-01 15:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 05:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Boqun Feng <boqun.feng@gmail.com> - 2016-02-02 06:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 07:50 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 09:10 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 09:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Boqun Feng <boqun.feng@gmail.com> - 2016-02-02 10:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 18:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-02 19:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 19:10 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-02 20:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 21:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-03 20:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Ingo Molnar <mingo@kernel.org> - 2016-02-03 09:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-03 14:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-03 20:10 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 13:10 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Linus Torvalds <torvalds@linux-foundation.org> - 2016-02-02 19:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 23:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Ralf Baechle <ralf@linux-mips.org> - 2016-02-02 15:50 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Måns Rullgård <mans@mansr.com> - 2016-02-02 16:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Ralf Baechle <ralf@linux-mips.org> - 2016-02-02 16:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Måns Rullgård <mans@mansr.com> - 2016-02-02 17:00 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Peter Zijlstra <peterz@infradead.org> - 2016-02-02 18:30 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 23:40 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-02 12:50 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Boqun Feng <boqun.feng@gmail.com> - 2016-02-02 13:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-02 13:30 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Boqun Feng <boqun.feng@gmail.com> - 2016-02-02 14:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() "Paul E. McKenney" <paulmck@linux.vnet.ibm.com> - 2016-02-02 18:20 +0100
Re: [RFC][PATCH] mips: Fix arch_spin_unlock() Will Deacon <will.deacon@arm.com> - 2016-02-02 18:40 +0100
csiph-web