Re: [RFC][PATCH] spin loop arch primitives for busy waiting

[Nicholas Piggin <npiggin@gmail.com>]

On Thu, 6 Apr 2017 15:13:53 +0100
Will Deacon <will.deacon@arm.com> wrote:

> Hi Nick,
> 
> On Thu, Apr 06, 2017 at 10:59:58AM +1000, Nicholas Piggin wrote:
> > On Wed, 05 Apr 2017 07:01:57 -0700 (PDT)
> > David Miller <davem@davemloft.net> wrote:
> >   
> > > From: Nicholas Piggin <npiggin@gmail.com>
> > > Date: Tue, 4 Apr 2017 13:02:33 +1000
> > >   
> > > > On Mon, 3 Apr 2017 17:43:05 -0700
> > > > Linus Torvalds <torvalds@linux-foundation.org> wrote:
> > > >     
> > > >> But that depends on architectures having some pattern that we *can*
> > > >> abstract. Would some "begin/in-loop/end" pattern like the above be
> > > >> sufficient?    
> > > > 
> > > > Yes. begin/in/end would be sufficient for powerpc SMT priority, and
> > > > for x86, and it looks like sparc64 too. So we could do that if you
> > > > prefer.    
> > > 
> > > Sparc64 has two cases, on older chips we can induce a cpu thread yield
> > > with a special sequence of instructions, and on newer chips we have
> > > a bonafide pause instruction.
> > > 
> > > So cpu_relax() all by itself pretty much works for us.
> > >   
> > 
> > Thanks for taking a look. The default spin primitives should just
> > continue to do the right thing for you in that case.
> > 
> > Arm has a yield instruction, ia64 has a pause... No unusual
> > requirements that I can see.  
> 
> Yield tends to be implemented as a NOP in practice, since it's in the
> architecture for SMT CPUs and most ARM CPUs are single-threaded. We do have
> the WFE instruction (wait for event) which is used in our implementation of
> smp_cond_load_acquire, but I don't think we'd be able to use it with the
> proposals here.
> 
> WFE can stop the clock for the CPU until an "event" is signalled by
> another CPU. This could be done by an explicit SEV (send event) instruction,
> but that tends to require heavy barriers on the signalling side. Instead,
> the preferred way to generate an event is to clear the exclusive monitor
> reservation for the CPU executing the WFE. That means that the waiter
> does something like:
> 
> 	LDXR x0, [some_address]	// Load exclusive from some_address
> 	CMP  x0, some value	// If the value matches what I want
> 	B.EQ out		// then we're done
> 	WFE			// otherwise, wait
> 
> at this point, the waiter will stop on the WFE until its monitor is cleared,
> which happens if another CPU writes to some_address.
> 
> We've wrapped this up in the arm64 code as __cmpwait, and we use that
> to build smp_cond_load_acquire. It would be nice to use the same machinery
> for the conditional spinning here, unless you anticipate that we're only
> going to be spinning for a handful of iterations anyway?

So I do want to look at adding spin loop primitives as well as the
begin/in/end primitives to help with powerpc's SMT priorities.

So we'd have:

  spin_begin();
  spin_do {
    if (blah) {
        spin_end();
        return;
    }
  } spin_until(!locked);
  spin_end();

So you could implement your monitor with that. There's a handful of core
places. mutex, bit spinlock, seqlock, polling idle, etc. So I think if it
is beneficial for you in smp_cond_load_acquire, it should be useful in
those too.