From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-10.8 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS, INCLUDES_PATCH,MAILING_LIST_MULTI,SPF_HELO_NONE,SPF_PASS,URIBL_BLOCKED autolearn=unavailable autolearn_force=no version=3.4.0 Received: from mail.kernel.org (mail.kernel.org [198.145.29.99]) by smtp.lore.kernel.org (Postfix) with ESMTP id AAB71C433B4 for ; Wed, 21 Apr 2021 20:00:09 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 78ADF613FA for ; Wed, 21 Apr 2021 20:00:09 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S245596AbhDUUAk (ORCPT ); Wed, 21 Apr 2021 16:00:40 -0400 Received: from us-smtp-delivery-124.mimecast.com ([216.205.24.124]:38839 "EHLO us-smtp-delivery-124.mimecast.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S236019AbhDUUAj (ORCPT ); Wed, 21 Apr 2021 16:00:39 -0400 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=redhat.com; s=mimecast20190719; t=1619035205; h=from:from:reply-to:subject:subject:date:date:message-id:message-id: to:to:cc:cc:mime-version:mime-version:content-type:content-type: content-transfer-encoding:content-transfer-encoding: in-reply-to:in-reply-to:references:references; bh=W0n7BeZseHa19F/4+bqG5QNyysdQ8/eAP434P02NiTg=; b=BIQOGE9VRSx7TbidYiNWNaP9VrPlWPDgfqcZVrLpS3IgjVsyH9irWvWPFwF2ESWzLioF2m kOp9CUn3N91nQKXvhuaO0xFD/Qe+POWDa7Xq6lIxJY55jM8JgQE59q6meRs1jCvuxHFsbk hANyxrshHu6gha61hJSJNCMQ3Gb5BJM= Received: from mail-ed1-f71.google.com (mail-ed1-f71.google.com [209.85.208.71]) (Using TLS) by relay.mimecast.com with ESMTP id us-mta-365-SN-Yv-UxNvKFGZXUfBso2A-1; Wed, 21 Apr 2021 15:59:59 -0400 X-MC-Unique: SN-Yv-UxNvKFGZXUfBso2A-1 Received: by mail-ed1-f71.google.com with SMTP id y10-20020a50f1ca0000b0290382d654f75eso15584308edl.1 for ; Wed, 21 Apr 2021 12:59:59 -0700 (PDT) X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:in-reply-to:references:date :message-id:mime-version:content-transfer-encoding; bh=W0n7BeZseHa19F/4+bqG5QNyysdQ8/eAP434P02NiTg=; b=E3czOR2E4nro92ryF4XdWQwfrK1gvjXLljPCwbHFIBG/KxOLpedCxgNUPb01bpNyaL 3Q6rFyHg0qxtaR+GTDZ6Nz0a18iNP+29j1SmW2bnCxC3C4Dv5/TxEd0m1mjC122Ydcy4 +OYwzUPcYkKNHKDoXh31w5bQAGqxtrmfJUhDBcRJKB+G7H0GaPs97JQZ6AxpUm10oGuQ t/9LS4KrccK3ZfNLCF/yu4Aut10IKl7N8RpucTWIjYnDLgsNojXNYzLqjRHDRbwkQRzp GJuJCTdpfRc5hUwKRt/ds3bXtxfBIj7LlX514tmCJmF8xQtn/ZBfAR+tv5vFVIJwSlof K8hg== X-Gm-Message-State: AOAM532kqAidwNkuiTd1P7GhYCHEnjwhkFUTrOEyIzEsgzGaSFxEd4eK VTyl5k5ERrzUZ54U5veqdeeZjZustqq5ie++tHnesOh2o9pIW/12J8MHozLwE85xSf5cm6VJ97k 6C5fyXVBcvSZf X-Received: by 2002:a17:906:ece4:: with SMTP id qt4mr35144678ejb.514.1619035197449; Wed, 21 Apr 2021 12:59:57 -0700 (PDT) X-Google-Smtp-Source: ABdhPJxAgfvzhtPiTu9YLc4erKBJ+vM0qLeSPW/UGdrmwLGpHqpyhD1AsRczVQTlHcqZBRqTt8GUCQ== X-Received: by 2002:a17:906:ece4:: with SMTP id qt4mr35144625ejb.514.1619035196722; Wed, 21 Apr 2021 12:59:56 -0700 (PDT) Received: from alrua-x1.borgediget.toke.dk ([2a0c:4d80:42:443::2]) by smtp.gmail.com with ESMTPSA id 16sm362227ejw.0.2021.04.21.12.59.55 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 21 Apr 2021 12:59:55 -0700 (PDT) Received: by alrua-x1.borgediget.toke.dk (Postfix, from userid 1000) id 4F5671802FE; Wed, 21 Apr 2021 21:59:55 +0200 (CEST) From: Toke =?utf-8?Q?H=C3=B8iland-J=C3=B8rgensen?= To: paulmck@kernel.org Cc: Martin KaFai Lau , Jesper Dangaard Brouer , Hangbin Liu , bpf@vger.kernel.org, netdev@vger.kernel.org, Jiri Benc , Eelco Chaudron , ast@kernel.org, Daniel Borkmann , Lorenzo Bianconi , David Ahern , Andrii Nakryiko , Alexei Starovoitov , John Fastabend , Maciej Fijalkowski , =?utf-8?B?QmrDtnJuIFQ=?= =?utf-8?B?w7ZwZWw=?= Subject: Re: [PATCHv7 bpf-next 1/4] bpf: run devmap xdp_prog on flush instead of bulk enqueue In-Reply-To: <20210421145951.GT975577@paulmck-ThinkPad-P17-Gen-1> References: <20210417002301.GO4212@paulmck-ThinkPad-P17-Gen-1> <87h7k5hza0.fsf@toke.dk> <20210419165837.GA975577@paulmck-ThinkPad-P17-Gen-1> <87lf9egn3o.fsf@toke.dk> <20210419183223.GC975577@paulmck-ThinkPad-P17-Gen-1> <877dkygeca.fsf@toke.dk> <20210419214117.GE975577@paulmck-ThinkPad-P17-Gen-1> <87y2ddgbsn.fsf@toke.dk> <20210419223143.GG975577@paulmck-ThinkPad-P17-Gen-1> <87wnsvhg0m.fsf@toke.dk> <20210421145951.GT975577@paulmck-ThinkPad-P17-Gen-1> X-Clacks-Overhead: GNU Terry Pratchett Date: Wed, 21 Apr 2021 21:59:55 +0200 Message-ID: <87r1j3h0hw.fsf@toke.dk> MIME-Version: 1.0 Content-Type: text/plain; charset=utf-8 Content-Transfer-Encoding: quoted-printable Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org "Paul E. McKenney" writes: > On Wed, Apr 21, 2021 at 04:24:41PM +0200, Toke H=C3=B8iland-J=C3=B8rgense= n wrote: >> "Paul E. McKenney" writes: >>=20 >> > On Tue, Apr 20, 2021 at 12:16:40AM +0200, Toke H=C3=B8iland-J=C3=B8rge= nsen wrote: >> >> "Paul E. McKenney" writes: >> >>=20 >> >> > On Mon, Apr 19, 2021 at 11:21:41PM +0200, Toke H=C3=B8iland-J=C3=B8= rgensen wrote: >> >> >> "Paul E. McKenney" writes: >> >> >>=20 >> >> >> > On Mon, Apr 19, 2021 at 08:12:27PM +0200, Toke H=C3=B8iland-J=C3= =B8rgensen wrote: >> >> >> >> "Paul E. McKenney" writes: >> >> >> >>=20 >> >> >> >> > On Sat, Apr 17, 2021 at 02:27:19PM +0200, Toke H=C3=B8iland-J= =C3=B8rgensen wrote: >> >> >> >> >> "Paul E. McKenney" writes: >> >> >> >> >>=20 >> >> >> >> >> > On Fri, Apr 16, 2021 at 11:22:52AM -0700, Martin KaFai Lau= wrote: >> >> >> >> >> >> On Fri, Apr 16, 2021 at 03:45:23PM +0200, Jesper Dangaard= Brouer wrote: >> >> >> >> >> >> > On Thu, 15 Apr 2021 17:39:13 -0700 >> >> >> >> >> >> > Martin KaFai Lau wrote: >> >> >> >> >> >> >=20 >> >> >> >> >> >> > > On Thu, Apr 15, 2021 at 10:29:40PM +0200, Toke H=C3= =B8iland-J=C3=B8rgensen wrote: >> >> >> >> >> >> > > > Jesper Dangaard Brouer writes: >> >> >> >> >> >> > > >=20=20=20 >> >> >> >> >> >> > > > > On Thu, 15 Apr 2021 10:35:51 -0700 >> >> >> >> >> >> > > > > Martin KaFai Lau wrote: >> >> >> >> >> >> > > > >=20=20 >> >> >> >> >> >> > > > >> On Thu, Apr 15, 2021 at 11:22:19AM +0200, Toke H= =C3=B8iland-J=C3=B8rgensen wrote:=20=20 >> >> >> >> >> >> > > > >> > Hangbin Liu writes: >> >> >> >> >> >> > > > >> >=20=20=20=20=20 >> >> >> >> >> >> > > > >> > > On Wed, Apr 14, 2021 at 05:17:11PM -0700, Ma= rtin KaFai Lau wrote:=20=20=20=20 >> >> >> >> >> >> > > > >> > >> > static void bq_xmit_all(struct xdp_dev_b= ulk_queue *bq, u32 flags) >> >> >> >> >> >> > > > >> > >> > { >> >> >> >> >> >> > > > >> > >> > struct net_device *dev =3D bq->dev; >> >> >> >> >> >> > > > >> > >> > - int sent =3D 0, err =3D 0; >> >> >> >> >> >> > > > >> > >> > + int sent =3D 0, drops =3D 0, err =3D 0; >> >> >> >> >> >> > > > >> > >> > + unsigned int cnt =3D bq->count; >> >> >> >> >> >> > > > >> > >> > + int to_send =3D cnt; >> >> >> >> >> >> > > > >> > >> > int i; >> >> >> >> >> >> > > > >> > >> >=20=20 >> >> >> >> >> >> > > > >> > >> > - if (unlikely(!bq->count)) >> >> >> >> >> >> > > > >> > >> > + if (unlikely(!cnt)) >> >> >> >> >> >> > > > >> > >> > return; >> >> >> >> >> >> > > > >> > >> >=20=20 >> >> >> >> >> >> > > > >> > >> > - for (i =3D 0; i < bq->count; i++) { >> >> >> >> >> >> > > > >> > >> > + for (i =3D 0; i < cnt; i++) { >> >> >> >> >> >> > > > >> > >> > struct xdp_frame *xdpf =3D bq->q[i]; >> >> >> >> >> >> > > > >> > >> >=20=20 >> >> >> >> >> >> > > > >> > >> > prefetch(xdpf); >> >> >> >> >> >> > > > >> > >> > } >> >> >> >> >> >> > > > >> > >> >=20=20 >> >> >> >> >> >> > > > >> > >> > - sent =3D dev->netdev_ops->ndo_xdp_xmit(= dev, bq->count, bq->q, flags); >> >> >> >> >> >> > > > >> > >> > + if (bq->xdp_prog) {=20=20=20=20 >> >> >> >> >> >> > > > >> > >> bq->xdp_prog is used here >> >> >> >> >> >> > > > >> > >>=20=20=20=20=20 >> >> >> >> >> >> > > > >> > >> > + to_send =3D dev_map_bpf_prog_run(bq->x= dp_prog, bq->q, cnt, dev); >> >> >> >> >> >> > > > >> > >> > + if (!to_send) >> >> >> >> >> >> > > > >> > >> > + goto out; >> >> >> >> >> >> > > > >> > >> > + >> >> >> >> >> >> > > > >> > >> > + drops =3D cnt - to_send; >> >> >> >> >> >> > > > >> > >> > + } >> >> >> >> >> >> > > > >> > >> > +=20=20=20=20 >> >> >> >> >> >> > > > >> > >>=20 >> >> >> >> >> >> > > > >> > >> [ ... ] >> >> >> >> >> >> > > > >> > >>=20=20=20=20=20 >> >> >> >> >> >> > > > >> > >> > static void bq_enqueue(struct net_device= *dev, struct xdp_frame *xdpf, >> >> >> >> >> >> > > > >> > >> > - struct net_device *dev_rx) >> >> >> >> >> >> > > > >> > >> > + struct net_device *dev_rx, stru= ct bpf_prog *xdp_prog) >> >> >> >> >> >> > > > >> > >> > { >> >> >> >> >> >> > > > >> > >> > struct list_head *flush_list =3D this_c= pu_ptr(&dev_flush_list); >> >> >> >> >> >> > > > >> > >> > struct xdp_dev_bulk_queue *bq =3D this_= cpu_ptr(dev->xdp_bulkq); >> >> >> >> >> >> > > > >> > >> > @@ -412,18 +466,22 @@ static void bq_enqu= eue(struct net_device *dev, struct xdp_frame *xdpf, >> >> >> >> >> >> > > > >> > >> > /* Ingress dev_rx will be the same for = all xdp_frame's in >> >> >> >> >> >> > > > >> > >> > * bulk_queue, because bq stored per-CP= U and must be flushed >> >> >> >> >> >> > > > >> > >> > * from net_device drivers NAPI func en= d. >> >> >> >> >> >> > > > >> > >> > + * >> >> >> >> >> >> > > > >> > >> > + * Do the same with xdp_prog and flush_= list since these fields >> >> >> >> >> >> > > > >> > >> > + * are only ever modified together. >> >> >> >> >> >> > > > >> > >> > */ >> >> >> >> >> >> > > > >> > >> > - if (!bq->dev_rx) >> >> >> >> >> >> > > > >> > >> > + if (!bq->dev_rx) { >> >> >> >> >> >> > > > >> > >> > bq->dev_rx =3D dev_rx; >> >> >> >> >> >> > > > >> > >> > + bq->xdp_prog =3D xdp_prog;=20=20=20=20 >> >> >> >> >> >> > > > >> > >> bp->xdp_prog is assigned here and could be = used later in bq_xmit_all(). >> >> >> >> >> >> > > > >> > >> How is bq->xdp_prog protected? Are they all= under one rcu_read_lock()? >> >> >> >> >> >> > > > >> > >> It is not very obvious after taking a quick= look at xdp_do_flush[_map]. >> >> >> >> >> >> > > > >> > >>=20 >> >> >> >> >> >> > > > >> > >> e.g. what if the devmap elem gets deleted.= =20=20=20=20 >> >> >> >> >> >> > > > >> > > >> >> >> >> >> >> > > > >> > > Jesper knows better than me. From my veiw, b= ased on the description of >> >> >> >> >> >> > > > >> > > __dev_flush(): >> >> >> >> >> >> > > > >> > > >> >> >> >> >> >> > > > >> > > On devmap tear down we ensure the flush list= is empty before completing to >> >> >> >> >> >> > > > >> > > ensure all flush operations have completed. = When drivers update the bpf >> >> >> >> >> >> > > > >> > > program they may need to ensure any flush op= s are also complete.=20=20=20=20 >> >> >> >> >> >> > > > >> >> >> >> >> >> >> > > > >> AFAICT, the bq->xdp_prog is not from the dev. It= is from a devmap's elem. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > The bq->xdp_prog comes form the devmap "dev" element, a= nd it is stored >> >> >> >> >> >> > in temporarily in the "bq" structure that is only valid= for this >> >> >> >> >> >> > softirq NAPI-cycle. I'm slightly worried that we copie= d this pointer >> >> >> >> >> >> > the the xdp_prog here, more below (and Q for Paul). >> >> >> >> >> >> >=20 >> >> >> >> >> >> > > > >> >=20 >> >> >> >> >> >> > > > >> > Yeah, drivers call xdp_do_flush() before exiti= ng their NAPI poll loop, >> >> >> >> >> >> > > > >> > which also runs under one big rcu_read_lock().= So the storage in the >> >> >> >> >> >> > > > >> > bulk queue is quite temporary, it's just used = for bulking to increase >> >> >> >> >> >> > > > >> > performance :)=20=20=20=20 >> >> >> >> >> >> > > > >> >> >> >> >> >> >> > > > >> I am missing the one big rcu_read_lock() part. = For example, in i40e_txrx.c, >> >> >> >> >> >> > > > >> i40e_run_xdp() has its own rcu_read_lock/unlock(= ). dst->xdp_prog used to run >> >> >> >> >> >> > > > >> in i40e_run_xdp() and it is fine. >> >> >> >> >> >> > > > >>=20 >> >> >> >> >> >> > > > >> In this patch, dst->xdp_prog is run outside of i= 40e_run_xdp() where the >> >> >> >> >> >> > > > >> rcu_read_unlock() has already done. It is now r= un in xdp_do_flush_map(). >> >> >> >> >> >> > > > >> or I missed the big rcu_read_lock() in i40e_napi= _poll()? >> >> >> >> >> >> > > > >> >> >> >> >> >> >> > > > >> I do see the big rcu_read_lock() in mlx5e_napi_p= oll().=20=20 >> >> >> >> >> >> > > > > >> >> >> >> >> >> > > > > I believed/assumed xdp_do_flush_map() was already= protected under an >> >> >> >> >> >> > > > > rcu_read_lock. As the devmap and cpumap, which g= et called via >> >> >> >> >> >> > > > > __dev_flush() and __cpu_map_flush(), have multipl= e RCU objects that we >> >> >> >> >> >> > > > > are operating on.=20=20 >> >> >> >> >> >> > > >> >> >> >> >> >> > > What other rcu objects it is using during flush? >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Look at code: >> >> >> >> >> >> > kernel/bpf/cpumap.c >> >> >> >> >> >> > kernel/bpf/devmap.c >> >> >> >> >> >> >=20 >> >> >> >> >> >> > The devmap is filled with RCU code and complicated take= -down steps.=20=20 >> >> >> >> >> >> > The devmap's elements are also RCU objects and the BPF = xdp_prog is >> >> >> >> >> >> > embedded in this object (struct bpf_dtab_netdev). The = call_rcu >> >> >> >> >> >> > function is __dev_map_entry_free(). >> >> >> >> >> >> >=20 >> >> >> >> >> >> >=20 >> >> >> >> >> >> > > > > Perhaps it is a bug in i40e?=20=20 >> >> >> >> >> >> > > >> >> >> >> >> >> > > A quick look into ixgbe falls into the same bucket. >> >> >> >> >> >> > > didn't look at other drivers though. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Intel driver are very much in copy-paste mode. >> >> >> >> >> >> >=20=20 >> >> >> >> >> >> > > > > >> >> >> >> >> >> > > > > We are running in softirq in NAPI context, when x= dp_do_flush_map() is >> >> >> >> >> >> > > > > call, which I think means that this CPU will not = go-through a RCU grace >> >> >> >> >> >> > > > > period before we exit softirq, so in-practice it = should be safe.=20=20 >> >> >> >> >> >> > > >=20 >> >> >> >> >> >> > > > Yup, this seems to be correct: rcu_softirq_qs() is = only called between >> >> >> >> >> >> > > > full invocations of the softirq handler, which for = networking is >> >> >> >> >> >> > > > net_rx_action(), and so translates into full NAPI p= oll cycles.=20=20 >> >> >> >> >> >> > > >> >> >> >> >> >> > > I don't know enough to comment on the rcu/softirq par= t, may be someone >> >> >> >> >> >> > > can chime in. There is also a recent napi_threaded_p= oll(). >> >> >> >> >> >> >=20 >> >> >> >> >> >> > CC added Paul. (link to patch[1][2] for context) >> >> >> >> >> >> Updated Paul's email address. >> >> >> >> >> >>=20 >> >> >> >> >> >> >=20 >> >> >> >> >> >> > > If it is the case, then some of the existing rcu_read= _lock() is unnecessary? >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Well, in many cases, especially depending on how kernel= is compiled, >> >> >> >> >> >> > that is true. But we want to keep these, as they also = document the >> >> >> >> >> >> > intend of the programmer. And allow us to make the ker= nel even more >> >> >> >> >> >> > preempt-able in the future. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > > At least, it sounds incorrect to only make an excepti= on here while keeping >> >> >> >> >> >> > > other rcu_read_lock() as-is. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Let me be clear: I think you have spotted a problem, a= nd we need to >> >> >> >> >> >> > add rcu_read_lock() at least around the invocation of >> >> >> >> >> >> > bpf_prog_run_xdp() or before around if-statement that c= all >> >> >> >> >> >> > dev_map_bpf_prog_run(). (Hangbin please do this in V8). >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Thank you Martin for reviewing the code carefully enoug= h to find this >> >> >> >> >> >> > issue, that some drivers don't have a RCU-section aroun= d the full XDP >> >> >> >> >> >> > code path in their NAPI-loop. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > Question to Paul. (I will attempt to describe in gener= ic terms what >> >> >> >> >> >> > happens, but ref real-function names). >> >> >> >> >> >> >=20 >> >> >> >> >> >> > We are running in softirq/NAPI context, the driver will= call a >> >> >> >> >> >> > bq_enqueue() function for every packet (if calling xdp_= do_redirect) , >> >> >> >> >> >> > some driver wrap this with a rcu_read_lock/unlock() sec= tion (other have >> >> >> >> >> >> > a large RCU-read section, that include the flush operat= ion). >> >> >> >> >> >> >=20 >> >> >> >> >> >> > In the bq_enqueue() function we have a per_cpu_ptr (tha= t store the >> >> >> >> >> >> > xdp_frame packets) that will get flushed/send in the ca= ll >> >> >> >> >> >> > xdp_do_flush() (that end-up calling bq_xmit_all()). Th= is flush will >> >> >> >> >> >> > happen before we end our softirq/NAPI context. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > The extension is that the per_cpu_ptr data structure (a= fter this patch) >> >> >> >> >> >> > store a pointer to an xdp_prog (which is a RCU object).= In the flush >> >> >> >> >> >> > operation (which we will wrap with RCU-read section), w= e will use this >> >> >> >> >> >> > xdp_prog pointer. I can see that it is in-principle w= rong to pass >> >> >> >> >> >> > this-pointer between RCU-read sections, but I consider = this safe as we >> >> >> >> >> >> > are running under softirq/NAPI and the per_cpu_ptr is o= nly valid in >> >> >> >> >> >> > this short interval. >> >> >> >> >> >> >=20 >> >> >> >> >> >> > I claim a grace/quiescent RCU cannot happen between the= se two RCU-read >> >> >> >> >> >> > sections, but I might be wrong? (especially in the futu= re or for RT). >> >> >> >> >> > >> >> >> >> >> > If I am reading this correctly (ha!), a very high-level su= mmary of the >> >> >> >> >> > code in question is something like this: >> >> >> >> >> > >> >> >> >> >> > void foo(void) >> >> >> >> >> > { >> >> >> >> >> > local_bh_disable(); >> >> >> >> >> > >> >> >> >> >> > rcu_read_lock(); >> >> >> >> >> > p =3D rcu_dereference(gp); >> >> >> >> >> > do_something_with(p); >> >> >> >> >> > rcu_read_unlock(); >> >> >> >> >> > >> >> >> >> >> > do_something_else(); >> >> >> >> >> > >> >> >> >> >> > rcu_read_lock(); >> >> >> >> >> > do_some_other_thing(p); >> >> >> >> >> > rcu_read_unlock(); >> >> >> >> >> > >> >> >> >> >> > local_bh_enable(); >> >> >> >> >> > } >> >> >> >> >> > >> >> >> >> >> > void bar(struct blat *new_gp) >> >> >> >> >> > { >> >> >> >> >> > struct blat *old_gp; >> >> >> >> >> > >> >> >> >> >> > spin_lock(my_lock); >> >> >> >> >> > old_gp =3D rcu_dereference_protected(gp, lock_held(my_lo= ck)); >> >> >> >> >> > rcu_assign_pointer(gp, new_gp); >> >> >> >> >> > spin_unlock(my_lock); >> >> >> >> >> > synchronize_rcu(); >> >> >> >> >> > kfree(old_gp); >> >> >> >> >> > } >> >> >> >> >>=20 >> >> >> >> >> Yeah, something like that (the object is freed using call_rc= u() - but I >> >> >> >> >> think that's equivalent, right?). And the question is whethe= r we need to >> >> >> >> >> extend foo() so that is has one big rcu_read_lock() that cov= ers the >> >> >> >> >> whole lifetime of p. >> >> >> >> > >> >> >> >> > Yes, use of call_rcu() is an asynchronous version of synchron= ize_rcu(). >> >> >> >> > In fact, synchronize_rcu() is implemented in terms of call_rc= u(). ;-) >> >> >> >>=20 >> >> >> >> Right, gotcha! >> >> >> >>=20 >> >> >> >> >> > I need to check up on -rt. >> >> >> >> >> > >> >> >> >> >> > But first... In recent mainline kernels, the local_bh_disa= ble() region >> >> >> >> >> > will look like one big RCU read-side critical section. Bu= t don't try >> >> >> >> >> > this prior to v4.20!!! In v4.19 and earlier, you would ne= ed to use >> >> >> >> >> > both synchronize_rcu() and synchronize_rcu_bh() to make th= is work, or, >> >> >> >> >> > for less latency, synchronize_rcu_mult(call_rcu, call_rcu_= bh). >> >> >> >> >>=20 >> >> >> >> >> OK. Variants of this code has been around since before then,= but I >> >> >> >> >> honestly have no idea what it looked like back then exactly.= .. >> >> >> >> > >> >> >> >> > I know that feeling... >> >> >> >> > >> >> >> >> >> > Except that in that case, why not just drop the inner rcu_= read_unlock() >> >> >> >> >> > and rcu_read_lock() pair? Awkward function boundaries or = some such? >> >> >> >> >>=20 >> >> >> >> >> Well if we can just treat such a local_bh_disable()/enable()= pair as the >> >> >> >> >> equivalent of rcu_read_lock()/unlock() then I suppose we cou= ld just get >> >> >> >> >> rid of the inner ones. What about tools like lockdep; do the= y understand >> >> >> >> >> this, or are we likely to get complaints if we remove it? >> >> >> >> > >> >> >> >> > If you just got rid of the first rcu_read_unlock() and the se= cond >> >> >> >> > rcu_read_lock() in the code above, lockdep will understand. >> >> >> >>=20 >> >> >> >> Right, but doing so entails going through all the drivers, whic= h is what >> >> >> >> we're trying to avoid :) >> >> >> > >> >> >> > I was afraid of that... ;-) >> >> >> > >> >> >> >> > However, if you instead get rid of -all- of the rcu_read_lock= () and >> >> >> >> > rcu_read_unlock() invocations in the code above, you would ne= ed to let >> >> >> >> > lockdep know by adding rcu_read_lock_bh_held(). So instead o= f this: >> >> >> >> > >> >> >> >> > p =3D rcu_dereference(gp); >> >> >> >> > >> >> >> >> > You would do this: >> >> >> >> > >> >> >> >> > p =3D rcu_dereference_check(gp, rcu_read_lock_bh_held()); >> >> >> >> > >> >> >> >> > This would be needed for mainline, regardless of -rt. >> >> >> >>=20 >> >> >> >> OK. And as far as I can tell this is harmless for code paths th= at call >> >> >> >> the same function but from a regular rcu_read_lock()-protected = section >> >> >> >> instead from a bh-disabled section, right? >> >> >> > >> >> >> > That is correct. That rcu_dereference_check() invocation will m= ake >> >> >> > lockdep be OK with rcu_read_lock() or with softirq being disable= d. >> >> >> > Or both, for that matter. >> >> >>=20 >> >> >> OK, great, thank you for confirming my understanding! >> >> >>=20 >> >> >> >> What happens, BTW, if we *don't* get rid of all the existing >> >> >> >> rcu_read_lock() sections? Going back to your foo() example abov= e, what >> >> >> >> we're discussing is whether to add that second rcu_read_lock() = around >> >> >> >> do_some_other_thing(p). I.e., the first one around the rcu_dere= ference() >> >> >> >> is already there (in the particular driver we're discussing), a= nd the >> >> >> >> local_bh_disable/enable() pair is already there. AFAICT from our >> >> >> >> discussion, there really is not much point in adding that second >> >> >> >> rcu_read_lock/unlock(), is there? >> >> >> > >> >> >> > From an algorithmic point of view, the second rcu_read_lock() >> >> >> > and rcu_read_unlock() are redundant. Of course, there are also >> >> >> > software-engineering considerations, including copy-pasta issues. >> >> >> > >> >> >> >> And because that first rcu_read_lock() around the rcu_dereferen= ce() is >> >> >> >> already there, lockdep is not likely to complain either, so we'= re >> >> >> >> basically fine? Except that the code is somewhat confusing as-i= s, of >> >> >> >> course; i.e., we should probably fix it but it's not terribly u= rgent. Or? >> >> >> > >> >> >> > I am concerned about copy-pasta-induced bugs. Someone looks jus= t at >> >> >> > the code, fails to note the fact that softirq is disabled throug= hout, >> >> >> > and decides that leaking a pointer from one RCU read-side critic= al >> >> >> > section to a later one is just fine. :-/ >> >> >>=20 >> >> >> Yup, totally agreed that we need to fix this for the sake of the h= umans >> >> >> reading the code; just wanted to make sure my understanding was co= rrect >> >> >> that we don't strictly need to do anything as far as the machines >> >> >> executing it are concerned :) >> >> >>=20 >> >> >> >> Hmm, looking at it now, it seems not all the lookup code is act= ually >> >> >> >> doing rcu_dereference() at all, but rather just a plain READ_ON= CE() with >> >> >> >> a comment above it saying that RCU ensures objects won't disapp= ear[0]; >> >> >> >> so I suppose we're at least safe from lockdep in that sense :P = - but we >> >> >> >> should definitely clean this up. >> >> >> >>=20 >> >> >> >> [0] Exhibit A: https://elixir.bootlin.com/linux/latest/source/k= ernel/bpf/devmap.c#L391 >> >> >> > >> >> >> > That use of READ_ONCE() will definitely avoid lockdep complaints, >> >> >> > including those complaints that point out bugs. It also might g= et you >> >> >> > sparse complaints if the RCU-protected pointer is marked with __= rcu. >> >> >>=20 >> >> >> It's not; it's the netdev_map member of this struct: >> >> >>=20 >> >> >> struct bpf_dtab { >> >> >> struct bpf_map map; >> >> >> struct bpf_dtab_netdev **netdev_map; /* DEVMAP type only */ >> >> >> struct list_head list; >> >> >>=20 >> >> >> /* these are only used for DEVMAP_HASH type maps */ >> >> >> struct hlist_head *dev_index_head; >> >> >> spinlock_t index_lock; >> >> >> unsigned int items; >> >> >> u32 n_buckets; >> >> >> }; >> >> >>=20 >> >> >> Will adding __rcu to such a dynamic array member do the right thin= g when >> >> >> paired with rcu_dereference() on array members (i.e., in place of = the >> >> >> READ_ONCE in the code linked above)? >> >> > >> >> > The only thing __rcu will do is provide information to the sparse s= tatic >> >> > analysis tool. Which will then gripe at you for applying READ_ONCE= () >> >> > to a __rcu pointer. But it is already griping at you for applying >> >> > rcu_dereference() to something not marked __rcu, so... ;-) >> >>=20 >> >> Right, hence the need for a cleanup ;) >> >>=20 >> >> My question was more if it understood arrays, though. I.e., that >> >> 'netdev_map' is an array of RCU pointers, not an RCU pointer to an >> >> array... Or am I maybe thinking that tool is way smarter than it is, = and >> >> it just complains for any access to that field that doesn't use >> >> rcu_dereference()? >> > >> > I believe that sparse will know about the pointers being __rcu, but >> > not the array. Unless you mark both levels. >>=20 >> Hi Paul >>=20 >> One more question, since I started adding the annotations: We are >> currently swapping out the pointers using xchg(): >> https://elixir.bootlin.com/linux/latest/source/kernel/bpf/devmap.c#L555 >>=20 >> and even cmpxchg(): >> https://elixir.bootlin.com/linux/latest/source/kernel/bpf/devmap.c#L831 >>=20 >> Sparse complains about these if I add the __rcu annotation to the >> definition (which otherwise works just fine with the double-pointer, >> BTW). Is there a way to fix that? Some kind of rcu_ macro version of the >> atomic swaps or something? Or do we just keep the regular xchg() and >> ignore those particular sparse warnings? > > Sounds like I need to supply a unrcu_pointer() macro or some such. > This would operate something like the current open-coded casts > in __rcu_dereference_protected(). So with that, I would turn the existing: dev =3D READ_ONCE(dtab->netdev_map[i]); if (!dev || netdev !=3D dev->dev) continue; odev =3D cmpxchg(&dtab->netdev_map[i], dev, NULL); into: dev =3D rcu_dereference(dtab->netdev_map[i]); if (!dev || netdev !=3D dev->dev) continue; odev =3D cmpxchg(unrcu_pointer(&dtab->netdev_map[i]), dev, NULL); and with a _check version: old_dev =3D xchg(unrcu_pointer_check(&dtab->netdev_map[k], rcu_read_lock_b= h_held()), NULL); right? Or would it be: odev =3D cmpxchg(&unrcu_pointer(dtab->netdev_map[i]), dev, NULL); ? > Would something like that work for you? Yeah, I believe it would :) -Toke