From: Boqun Feng <boqun.feng@gmail.com>
To: Dan Lustig <dlustig@nvidia.com>
Cc: Will Deacon <will@kernel.org>,
Peter Zijlstra <peterz@infradead.org>,
Linus Torvalds <torvalds@linux-foundation.org>,
Alan Stern <stern@rowland.harvard.edu>,
Alexander Shishkin <alexander.shishkin@linux.intel.com>,
Peter Anvin <hpa@zytor.com>,
Andrea Parri <parri.andrea@gmail.com>,
Ingo Molnar <mingo@kernel.org>,
"Paul E. McKenney" <paulmck@kernel.org>,
Vince Weaver <vincent.weaver@maine.edu>,
Thomas Gleixner <tglx@linutronix.de>,
Jiri Olsa <jolsa@redhat.com>,
Arnaldo Carvalho de Melo <acme@redhat.com>,
Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
Stephane Eranian <eranian@google.com>,
linux-tip-commits@vger.kernel.org, palmer@dabbelt.com,
paul.walmsley@sifive.com, mpe@ellerman.id.au
Subject: Re: [tip:locking/core] tools/memory-model: Add extra ordering for locks and remove it for ordinary release/acquire
Date: Fri, 10 Sep 2021 13:37:48 +0800 [thread overview]
Message-ID: <YTrvLHB6lpol79ka@boqun-archlinux> (raw)
In-Reply-To: <YTqgSmX57l2hCMk0@boqun-archlinux>
On Fri, Sep 10, 2021 at 08:01:14AM +0800, Boqun Feng wrote:
> On Thu, Sep 09, 2021 at 01:03:18PM -0400, Dan Lustig wrote:
> > On 9/9/2021 9:35 AM, Will Deacon wrote:
> > > [+Palmer, PaulW, Daniel and Michael]
> > >
> > > On Thu, Sep 09, 2021 at 09:25:30AM +0200, Peter Zijlstra wrote:
> > >> On Wed, Sep 08, 2021 at 09:08:33AM -0700, Linus Torvalds wrote:
> > >>
> > >>> So if this is purely a RISC-V thing,
> > >>
> > >> Just to clarify, I think the current RISC-V thing is stonger than
> > >> PowerPC, but maybe not as strong as say ARM64, but RISC-V memory
> > >> ordering is still somewhat hazy to me.
> > >>
> > >> Specifically, the sequence:
> > >>
> > >> /* critical section s */
> > >> WRITE_ONCE(x, 1);
> > >> FENCE RW, W
> > >> WRITE_ONCE(s.lock, 0); /* store S */
> > >> AMOSWAP %0, 1, r.lock /* store R */
> > >> FENCE R, RW
> > >> WRITE_ONCE(y, 1);
> > >> /* critical section r */
> > >>
> > >> fully separates section s from section r, as in RW->RW ordering
> > >> (possibly not as strong as smp_mb() though), while on PowerPC it would
> > >> only impose TSO ordering between sections.
> > >>
> > >> The AMOSWAP is a RmW and as such matches the W from the RW->W fence,
> > >> similarly it marches the R from the R->RW fence, yielding an:
> > >>
> > >> RW-> W
> > >> RmW
> > >> R ->RW
> > >>
> > >> ordering. It's the stores S and R that can be re-ordered, but not the
> > >> sections themselves (same on PowerPC and many others).
> > >>
> > >> Clarification from a RISC-V enabled person would be appreciated.
> >
> > To first order, RISC-V's memory model is very similar to ARMv8's. It
> > is "other-multi-copy-atomic", unlike Power, and respects dependencies.
> > It also has AMOs and LR/SC with optional RCsc acquire or release
> > semantics. There's no need to worry about RISC-V somehow pushing the
> > boundaries of weak memory ordering in new ways.
> >
> > The tricky part is that unlike ARMv8, RISC-V doesn't have load-acquire
> > or store-release opcodes at all. Only AMOs and LR/SC have acquire or
> > release options. That means that while certain operations like swap
> > can be implemented with native RCsc semantics, others like store-release
> > have to fall back on fences and plain writes.
> >
> > That's where the complexity came up last time this was discussed, at
> > least as it relates to RISC-V: how to make sure the combination of RCsc
> > atomics and plain operations+fences gives the semantics everyone is
> > asking for here. And to be clear there, I'm not asking for LKMM to
> > weaken anything about critical section ordering just for RISC-V's sake.
> > TSO/RCsc ordering between critical sections is a perfectly reasonable
> > model in my opinion. I just want to make sure RISC-V gets it right
> > given whatever the decision is.
> >
> > >>> then I think it's entirely reasonable to
> > >>>
> > >>> spin_unlock(&r);
> > >>> spin_lock(&s);
> > >>>
> > >>> cannot be reordered.
> > >>
> > >> I'm obviously completely in favour of that :-)
> > >
> > > I don't think we should require the accesses to the actual lockwords to
> > > be ordered here, as it becomes pretty onerous for relaxed LL/SC
> > > architectures where you'd end up with an extra barrier either after the
> > > unlock() or before the lock() operation. However, I remain absolutely in
> > > favour of strengthening the ordering of the _critical sections_ guarded by
> > > the locks to be RCsc.
> >
> > I agree with Will here. If the AMOSWAP above is actually implemented with
> > a RISC-V AMO, then the two critical sections will be separated as if RW,RW,
> > as Peter described. If instead it's implemented using LR/SC, then RISC-V
>
> Just out of curiosity, in the following code, can the store S and load L
> be reordered?
>
> WRITE_ONCE(x, 1); // store S
> FENCE RW, W
> WRITE_ONCE(s.lock, 0); // unlock(s)
> AMOSWAP %0, 1, s.lock // lock(s)
> FENCE R, RW
> r1 = READ_ONCE(y); // load L
>
> I think they can, because neither "FENCE RW, W" nor "FENCE R, RW" order
> them. Note that the reordering is allowed in LKMM, because unlock-lock
> only need to be as strong as RCtso.
>
> Moreover, how about the outcome of the following case:
>
> {
> r1, r2 are registers (variables) on each CPU, X, Y are memory
> locations, and initialized as 0
> }
>
> CPU 0
> =====
> AMOSWAP r1, 1, X
> FENCE R, RW
> r2 = READ_ONCE(Y);
>
> CPU 1
> =====
> WRITE_ONCE(Y, 1);
> FENCE RW, RW
> r2 = READ_ONCE(X);
>
> can we observe the result where r2 on CPU0 is 0 while r2 on CPU1 is 1?
>
As reminded by Andrea, what I meant to ask here is:
can we observer the result where r2 on CPU0 is 0 while r2 on CPU1 is 0?
Regards,
Boqun
> Regards,
> Boqun
>
> > gives only TSO (R->R, R->W, W->W), because the two pieces of the AMO are
> > split, and that breaks the chain. Getting full RW->RW between the critical
> > sections would therefore require an extra fence. Also, the accesses to the
> > lockwords themselves would not be ordered without an extra fence.
> >
> > > Last time this came up, I think the RISC-V folks were generally happy to
> > > implement whatever was necessary for Linux [1]. The thing that was stopping
> > > us was Power (see CONFIG_ARCH_WEAK_RELEASE_ACQUIRE), wasn't it? I think
> > > Michael saw quite a bit of variety in the impact on benchmarks [2] across
> > > different machines. So the question is whether newer Power machines are less
> > > affected to the degree that we could consider making this change again.
> >
> > Yes, as I said above, RISC-V will implement what is needed to make this work.
> >
> > Dan
> >
> > > Will
> > >
> > > [1] https://lore.kernel.org/lkml/11b27d32-4a8a-3f84-0f25-723095ef1076@nvidia.com/
> > > [2] https://lore.kernel.org/lkml/87tvp3xonl.fsf@concordia.ellerman.id.au/
next prev parent reply other threads:[~2021-09-10 5:39 UTC|newest]
Thread overview: 41+ messages / expand[flat|nested] mbox.gz Atom feed top
2018-09-26 18:28 [PATCH memory-model 0/5] Updates to the formal memory model Paul E. McKenney
2018-09-26 18:29 ` [PATCH memory-model 1/5] tools/memory-model: Add litmus-test naming scheme Paul E. McKenney
2018-10-02 10:10 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-09-26 18:29 ` [PATCH memory-model 2/5] tools/memory-model: Add extra ordering for locks and remove it for ordinary release/acquire Paul E. McKenney
2018-10-02 10:11 ` [tip:locking/core] " tip-bot for Alan Stern
2021-09-08 11:00 ` Peter Zijlstra
2021-09-08 11:44 ` Peter Zijlstra
2021-09-08 14:42 ` Alan Stern
2021-09-08 15:12 ` Peter Zijlstra
2021-09-08 16:08 ` Linus Torvalds
2021-09-09 7:25 ` Peter Zijlstra
2021-09-09 13:35 ` Will Deacon
2021-09-09 17:02 ` Linus Torvalds
2021-09-09 18:59 ` Alan Stern
2021-09-09 17:03 ` Dan Lustig
2021-09-09 18:00 ` Paul E. McKenney
2021-09-10 14:20 ` Boqun Feng
2021-09-10 15:33 ` Palmer Dabbelt
2021-09-10 16:36 ` Alan Stern
2021-09-10 17:12 ` Peter Zijlstra
2021-09-10 17:56 ` Alan Stern
2021-09-10 17:17 ` Peter Zijlstra
2021-09-12 0:26 ` Boqun Feng
2021-09-10 0:01 ` Boqun Feng
2021-09-10 5:37 ` Boqun Feng [this message]
2021-09-10 9:33 ` Peter Zijlstra
2021-09-10 10:04 ` Boqun Feng
2021-09-10 13:48 ` Dan Lustig
2021-09-10 14:15 ` Boqun Feng
2021-09-09 17:46 ` Paul E. McKenney
2021-09-10 11:08 ` Will Deacon
2021-09-17 3:21 ` Nicholas Piggin
2021-09-17 5:31 ` Nicholas Piggin
2021-09-17 14:36 ` Michael Ellerman
2018-09-26 18:29 ` [PATCH memory-model 3/5] tools/memory-model: Fix a README typo Paul E. McKenney
2018-10-02 10:11 ` [tip:locking/core] " tip-bot for SeongJae Park
2018-09-26 18:29 ` [PATCH memory-model 4/5] tools/memory-model: Add more LKMM limitations Paul E. McKenney
2018-10-02 10:12 ` [tip:locking/core] " tip-bot for Paul E. McKenney
2018-09-26 18:29 ` [PATCH memory-model 5/5] doc: Replace smp_cond_acquire() with smp_cond_load_acquire() Paul E. McKenney
2018-10-02 10:12 ` [tip:locking/core] locking/memory-barriers: " tip-bot for Andrea Parri
2018-10-02 8:28 ` [PATCH memory-model 0/5] Updates to the formal memory model Ingo Molnar
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=YTrvLHB6lpol79ka@boqun-archlinux \
--to=boqun.feng@gmail.com \
--cc=acme@redhat.com \
--cc=alexander.shishkin@linux.intel.com \
--cc=dlustig@nvidia.com \
--cc=eranian@google.com \
--cc=hpa@zytor.com \
--cc=jolsa@redhat.com \
--cc=linux-kernel@vger.kernel.org \
--cc=linux-tip-commits@vger.kernel.org \
--cc=mingo@kernel.org \
--cc=mpe@ellerman.id.au \
--cc=palmer@dabbelt.com \
--cc=parri.andrea@gmail.com \
--cc=paul.walmsley@sifive.com \
--cc=paulmck@kernel.org \
--cc=peterz@infradead.org \
--cc=stern@rowland.harvard.edu \
--cc=tglx@linutronix.de \
--cc=torvalds@linux-foundation.org \
--cc=vincent.weaver@maine.edu \
--cc=will@kernel.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).