From: Peter Zijlstra <peterz@infradead•org>
To: Alexei Starovoitov <alexei.starovoitov@gmail•com>
Cc: Alexei Starovoitov <ast@kernel•org>,
davem@davemloft•net, daniel@iogearbox•net,
jakub.kicinski@netronome•com, netdev@vger•kernel.org,
kernel-team@fb•com, mingo@redhat•com, will.deacon@arm•com,
Paul McKenney <paulmck@linux•vnet.ibm.com>,
jannh@google•com
Subject: Re: bpf memory model. Was: [PATCH v4 bpf-next 1/9] bpf: introduce bpf_spin_lock
Date: Mon, 28 Jan 2019 10:24:08 +0100 [thread overview]
Message-ID: <20190128092408.GD28467@hirez.programming.kicks-ass.net> (raw)
In-Reply-To: <20190126001725.roqqfrpysyljqiqx@ast-mbp.dhcp.thefacebook.com>
On Fri, Jan 25, 2019 at 04:17:26PM -0800, Alexei Starovoitov wrote:
> On Fri, Jan 25, 2019 at 11:23:12AM +0100, Peter Zijlstra wrote:
> > On Thu, Jan 24, 2019 at 03:58:59PM -0800, Alexei Starovoitov wrote:
> > > On Thu, Jan 24, 2019 at 07:01:09PM +0100, Peter Zijlstra wrote:
> >
> > > > And this would again be the moment where I go pester you about the BPF
> > > > memory model :-)
> > >
> > > hehe :)
> > > How do you propose to define it in a way that it applies to all archs
> > > and yet doesn't penalize x86 ?
> > > "Assume minimum execution ordering model" the way kernel does
> > > unfortunately is not usable, since bpf doesn't have a luxury
> > > of using nice #defines that convert into nops on x86.
> >
> > Why not? Surely the JIT can fix it up? That is, suppose you were to have
> > smp_rmb() as a eBPF instruction then the JIT (which knows what
> > architecture it is targeting) can simply avoid emitting instructions for
> > it.
>
> I'm all for adding new instructions that solve real use cases.
> imo bpf_spin_lock() is the first step in helping with concurrency.
> At plumbers conference we agreed to add new sync_fetch_and_add()
> and xchg() instructions. That's a second step.
> smp_rmb/wmb/mb should be added as well.
> JITs will patch them depending on architecture.
>
> What I want to avoid is to define the whole execution ordering model upfront.
> We cannot say that BPF ISA is weakly ordered like alpha.
> Most of the bpf progs are written and running on x86. We shouldn't
> twist bpf developer's arm by artificially relaxing memory model.
> BPF memory model is equal to memory model of underlying architecture.
> What we can do is to make it bpf progs a bit more portable with
> smp_rmb instructions, but we must not force weak execution on the developer.
Well, I agree with only introducing bits you actually need, and my
smp_rmb() example might have been poorly chosen, smp_load_acquire() /
smp_store_release() might have been a far more useful example.
But I disagree with the last part; we have to pick a model now;
otherwise you'll pain yourself into a corner.
Also; Alpha isn't very relevant these days; however ARM64 does seem to
be gaining a lot of attention and that is very much a weak architecture.
Adding strongly ordered assumptions to BPF now, will penalize them in
the long run.
> > Similarly; could something like this not also help with the spinlock
> > thing? Use that generic test-and-set thing for the interpreter, but
> > provide a better implementation in the JIT?
>
> May be eventually. We can add cmpxchg insn, but the verifier still
> doesn't support loops. We made a lot of progress in bounded loop research
> over the last 2 years, but loops in bpf are still a year away.
> We considered adding 'bpf_spin_lock' as a new instruction instead of helper call,
> but that approach has a lot of negatives, so we went with the helper.
Ah, but the loop won't be in the BPF program itself. The BPF program
would only have had the BPF_SPIN_LOCK instruction, the JIT them emits
code similar to queued_spin_lock()/queued_spin_unlock() (or calls to
out-of-line versions of them).
There isn't anything that mandates the JIT uses the exact same locking
routines the interpreter does, is there?
next prev parent reply other threads:[~2019-01-28 9:24 UTC|newest]
Thread overview: 62+ messages / expand[flat|nested] mbox.gz Atom feed top
2019-01-24 4:13 [PATCH v4 bpf-next 0/9] introduce bpf_spin_lock Alexei Starovoitov
2019-01-24 4:13 ` [PATCH v4 bpf-next 1/9] bpf: " Alexei Starovoitov
2019-01-24 18:01 ` Peter Zijlstra
2019-01-24 18:56 ` Peter Zijlstra
2019-01-24 23:42 ` Paul E. McKenney
2019-01-25 0:05 ` Alexei Starovoitov
2019-01-25 1:22 ` Paul E. McKenney
2019-01-25 1:46 ` Jann Horn
2019-01-25 2:38 ` Alexei Starovoitov
2019-01-25 4:27 ` Alexei Starovoitov
2019-01-25 4:31 ` Paul E. McKenney
2019-01-25 4:47 ` Alexei Starovoitov
2019-01-25 16:02 ` Paul E. McKenney
2019-01-25 4:11 ` Paul E. McKenney
2019-01-25 16:18 ` Jann Horn
2019-01-25 22:51 ` Paul E. McKenney
2019-01-25 23:44 ` Alexei Starovoitov
2019-01-26 0:43 ` Jann Horn
2019-01-26 0:59 ` Jann Horn
2019-01-24 23:58 ` Alexei Starovoitov
2019-01-25 0:18 ` Jann Horn
2019-01-25 2:49 ` Alexei Starovoitov
2019-01-25 2:29 ` Eric Dumazet
2019-01-25 2:34 ` Alexei Starovoitov
2019-01-25 2:44 ` Eric Dumazet
2019-01-25 2:57 ` Alexei Starovoitov
2019-01-25 8:38 ` Peter Zijlstra
2019-01-25 9:10 ` Peter Zijlstra
2019-01-25 23:42 ` Alexei Starovoitov
2019-01-28 8:24 ` Peter Zijlstra
2019-01-28 8:31 ` Peter Zijlstra
2019-01-28 8:35 ` Peter Zijlstra
2019-01-28 20:49 ` Alexei Starovoitov
2019-01-28 8:43 ` Peter Zijlstra
2019-01-28 21:37 ` Alexei Starovoitov
2019-01-29 8:59 ` Peter Zijlstra
2019-01-30 2:20 ` Alexei Starovoitov
2019-01-25 9:59 ` Peter Zijlstra
2019-01-25 10:09 ` Peter Zijlstra
2019-01-25 10:23 ` Peter Zijlstra
2019-01-26 0:17 ` bpf memory model. Was: " Alexei Starovoitov
2019-01-28 9:24 ` Peter Zijlstra [this message]
2019-01-28 21:56 ` Alexei Starovoitov
2019-01-29 9:16 ` Peter Zijlstra
2019-01-30 2:32 ` Alexei Starovoitov
2019-01-30 8:58 ` Peter Zijlstra
2019-01-30 19:36 ` Alexei Starovoitov
2019-01-30 18:11 ` Will Deacon
2019-01-30 18:36 ` Paul E. McKenney
2019-01-30 19:51 ` Alexei Starovoitov
2019-01-30 21:05 ` Paul E. McKenney
2019-01-30 22:57 ` Alexei Starovoitov
2019-01-31 14:01 ` Paul E. McKenney
2019-01-31 18:47 ` Alexei Starovoitov
2019-02-01 14:05 ` Paul E. McKenney
2019-01-30 19:50 ` Alexei Starovoitov
2019-01-24 4:13 ` [PATCH v4 bpf-next 2/9] bpf: add support for bpf_spin_lock to cgroup local storage Alexei Starovoitov
2019-01-24 4:13 ` [PATCH v4 bpf-next 3/9] tools/bpf: sync include/uapi/linux/bpf.h Alexei Starovoitov
2019-01-24 4:13 ` [PATCH v4 bpf-next 4/9] selftests/bpf: add bpf_spin_lock tests Alexei Starovoitov
2019-01-24 4:13 ` [PATCH v4 bpf-next 5/9] selftests/bpf: add bpf_spin_lock C test Alexei Starovoitov
2019-01-24 4:14 ` [PATCH v4 bpf-next 6/9] bpf: introduce BPF_F_LOCK flag Alexei Starovoitov
2019-01-24 4:14 ` [PATCH v4 bpf-next 7/9] tools/bpf: sync uapi/bpf.h Alexei Starovoitov
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=20190128092408.GD28467@hirez.programming.kicks-ass.net \
--to=peterz@infradead$(echo .)org \
--cc=alexei.starovoitov@gmail$(echo .)com \
--cc=ast@kernel$(echo .)org \
--cc=daniel@iogearbox$(echo .)net \
--cc=davem@davemloft$(echo .)net \
--cc=jakub.kicinski@netronome$(echo .)com \
--cc=jannh@google$(echo .)com \
--cc=kernel-team@fb$(echo .)com \
--cc=mingo@redhat$(echo .)com \
--cc=netdev@vger$(echo .)kernel.org \
--cc=paulmck@linux$(echo .)vnet.ibm.com \
--cc=will.deacon@arm$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
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