Received-SPF: pass (google.com: domain of linux-kernel-owner@vger.kernel.org designates 23.128.96.18 as permitted sender) client-ip=23.128.96.18;
MIME-Version: 1.0
References: <1616868399-82848-1-git-send-email-guoren@kernel.org>
 <1616868399-82848-4-git-send-email-guoren@kernel.org> <YGGGqftfr872/4CU@hirez.programming.kicks-ass.net>
 <CAJF2gTQNV+_txMHJw0cmtS-xcnuaCja-F7XBuOL_J0yN39c+uQ@mail.gmail.com>
 <YGG5c4QGq6q+lKZI@hirez.programming.kicks-ass.net> <CAJF2gTQUe237NY-kh+4_Yk4DTFJmA5_xgNQ5+BMpFZpUDUEYdw@mail.gmail.com>
 <YGHM2/s4FpWZiEQ6@hirez.programming.kicks-ass.net> <CAJF2gTRncV1+GT7nBpYkvfpyaG57o9ecaHBjoR6gEQAkG2ELrg@mail.gmail.com>
 <4d0dbaa0-1f96-470c-0ed0-04f6827ea384@redhat.com>
In-Reply-To: <4d0dbaa0-1f96-470c-0ed0-04f6827ea384@redhat.com>
From:   Guo Ren <guoren@kernel.org>
Date:   Tue, 6 Apr 2021 00:45:58 +0800
Message-ID: <CAJF2gTRN2FLUbJzwsgitMg2j3sMRcq5a1Gm5dTQivuiakmAdOQ@mail.gmail.com>
Subject: Re: [PATCH v4 3/4] locking/qspinlock: Add ARCH_USE_QUEUED_SPINLOCKS_XCHG32
To:     Waiman Long <longman@redhat.com>
Cc:     Peter Zijlstra <peterz@infradead.org>,
        linux-riscv <linux-riscv@lists.infradead.org>,
        Linux Kernel Mailing List <linux-kernel@vger.kernel.org>,
        linux-csky@vger.kernel.org,
        linux-arch <linux-arch@vger.kernel.org>,
        Guo Ren <guoren@linux.alibaba.com>,
        Will Deacon <will@kernel.org>, Ingo Molnar <mingo@redhat.com>,
        Arnd Bergmann <arnd@arndb.de>, Anup Patel <anup@brainfault.org>
Content-Type: text/plain; charset="UTF-8"
Precedence: bulk

On Tue, Mar 30, 2021 at 10:09 PM Waiman Long <longman@redhat.com> wrote:
>
> On 3/29/21 11:13 PM, Guo Ren wrote:
> > On Mon, Mar 29, 2021 at 8:50 PM Peter Zijlstra <peterz@infradead.org> wrote:
> >> On Mon, Mar 29, 2021 at 08:01:41PM +0800, Guo Ren wrote:
> >>> u32 a = 0x55aa66bb;
> >>> u16 *ptr = &a;
> >>>
> >>> CPU0                       CPU1
> >>> =========             =========
> >>> xchg16(ptr, new)     while(1)
> >>>                                      WRITE_ONCE(*(ptr + 1), x);
> >>>
> >>> When we use lr.w/sc.w implement xchg16, it'll cause CPU0 deadlock.
> >> Then I think your LL/SC is broken.
> >>
> >> That also means you really don't want to build super complex locking
> >> primitives on top, because that live-lock will percolate through.
> > Do you mean the below implementation has live-lock risk?
> > +static __always_inline u32 xchg_tail(struct qspinlock *lock, u32 tail)
> > +{
> > +       u32 old, new, val = atomic_read(&lock->val);
> > +
> > +       for (;;) {
> > +               new = (val & _Q_LOCKED_PENDING_MASK) | tail;
> > +               old = atomic_cmpxchg(&lock->val, val, new);
> > +               if (old == val)
> > +                       break;
> > +
> > +               val = old;
> > +       }
> > +       return old;
> > +}
> If there is a continuous stream of incoming spinlock takers, it is
> possible that some cpus may have to wait a long time to set the tail
> right. However, this should only happen on artificial workload. I doubt
> it will happen with real workload or with limit number of cpus.
Yes, I think is ok for LR/SC in riscv, becasue

CPU0 LR
CPU1 LR
CPU0 SC //success
CPU1 SC //fail

or

CPU0 LR
CPU1 LR
CPU1 SC //success
CPU0 SC //fail

So always one store condition would success. I think it's OK.

> >
> >> Step 1 would be to get your architecute fixed such that it can provide
> >> fwd progress guarantees for LL/SC. Otherwise there's absolutely no point
> >> in building complex systems with it.
> > Quote Waiman's comment [1] on xchg16 optimization:
> >
> > "This optimization is needed to make the qspinlock achieve performance
> > parity with ticket spinlock at light load."
> >
> > [1] https://lore.kernel.org/kvm/1429901803-29771-6-git-send-email-Waiman.Long@hp.com/
> >
> > So for a non-xhg16 machine:
> >   - ticket-lock for small numbers of CPUs
> >   - qspinlock for large numbers of CPUs
> >
> > Okay, I'll put all of them into the next patch :P
> >
> It is true that qspinlock may not offer much advantage when the number
> of cpus is small. It shines for systems with many cpus. You may use
> NR_CPUS to determine if the default should be ticket or qspinlock with
> user override. To determine the right NR_CPUS threshold, you may need to
> run on real SMP RISCV systems to find out.
I Agree

-- 
Best Regards
 Guo Ren

ML: https://lore.kernel.org/linux-csky/