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=-9.9 required=3.0 tests=DKIMWL_WL_HIGH,DKIM_SIGNED, DKIM_VALID,DKIM_VALID_AU,HEADER_FROM_DIFFERENT_DOMAINS,INCLUDES_PATCH, MAILING_LIST_MULTI,SIGNED_OFF_BY,SPF_HELO_NONE,SPF_PASS,USER_AGENT_GIT autolearn=ham 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 5DEEFC433E1 for ; Sun, 17 May 2020 19:58:03 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 2A92620758 for ; Sun, 17 May 2020 19:58:03 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=fb.com header.i=@fb.com header.b="Cix0pjvD" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1726379AbgEQT6C (ORCPT ); Sun, 17 May 2020 15:58:02 -0400 Received: from mx0b-00082601.pphosted.com ([67.231.153.30]:35380 "EHLO mx0b-00082601.pphosted.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1726729AbgEQT56 (ORCPT ); Sun, 17 May 2020 15:57:58 -0400 Received: from pps.filterd (m0148460.ppops.net [127.0.0.1]) by mx0a-00082601.pphosted.com (8.16.0.42/8.16.0.42) with SMTP id 04HJrxQe029470 for ; Sun, 17 May 2020 12:57:56 -0700 DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=fb.com; h=from : to : cc : subject : date : message-id : in-reply-to : references : mime-version : content-transfer-encoding : content-type; s=facebook; bh=KGd8IKZV0sXzyk1o+zFwRc06cXywfi9UewothHj879s=; b=Cix0pjvDACWPU7EnWR3Q2OP1vW8lIKrbtkPZcWGaVwF+n8bF3PIGVgGSiDBWzxKAk/ue vpBJw6OtdUQXV0IDNC2ZVtEQzlpRytzT2i1ma/fWMGD8kJTwaYrpl01HJrXhtwtoSAOz fN+gFKDgG+ZxIa/M3/ZrYpDgpfvx6vYp0YY= Received: from maileast.thefacebook.com ([163.114.130.16]) by mx0a-00082601.pphosted.com with ESMTP id 312dpwtxgt-2 (version=TLSv1.2 cipher=ECDHE-RSA-AES128-GCM-SHA256 bits=128 verify=NOT) for ; Sun, 17 May 2020 12:57:55 -0700 Received: from intmgw002.08.frc2.facebook.com (2620:10d:c0a8:1b::d) by mail.thefacebook.com (2620:10d:c0a8:82::d) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_128_GCM_SHA256) id 15.1.1847.3; Sun, 17 May 2020 12:57:55 -0700 Received: by devbig012.ftw2.facebook.com (Postfix, from userid 137359) id 89A122EC32DC; Sun, 17 May 2020 12:57:50 -0700 (PDT) Smtp-Origin-Hostprefix: devbig From: Andrii Nakryiko Smtp-Origin-Hostname: devbig012.ftw2.facebook.com To: , , , CC: , , Andrii Nakryiko , "Paul E . McKenney" , Jonathan Lemon , Stanislav Fomichev Smtp-Origin-Cluster: ftw2c04 Subject: [PATCH v2 bpf-next 7/7] docs/bpf: add BPF ring buffer design notes Date: Sun, 17 May 2020 12:57:27 -0700 Message-ID: <20200517195727.279322-8-andriin@fb.com> X-Mailer: git-send-email 2.24.1 In-Reply-To: <20200517195727.279322-1-andriin@fb.com> References: <20200517195727.279322-1-andriin@fb.com> MIME-Version: 1.0 Content-Transfer-Encoding: quoted-printable X-FB-Internal: Safe Content-Type: text/plain X-Proofpoint-Virus-Version: vendor=fsecure engine=2.50.10434:6.0.216,18.0.676 definitions=2020-05-17_08:2020-05-15,2020-05-17 signatures=0 X-Proofpoint-Spam-Details: rule=fb_default_notspam policy=fb_default score=0 mlxscore=0 phishscore=0 malwarescore=0 clxscore=1015 impostorscore=0 spamscore=0 mlxlogscore=999 lowpriorityscore=0 suspectscore=25 bulkscore=0 priorityscore=1501 cotscore=-2147483648 adultscore=0 classifier=spam adjust=0 reason=mlx scancount=1 engine=8.12.0-2004280000 definitions=main-2005170182 X-FB-Internal: deliver Sender: bpf-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org Add commit description from patch #1 as a stand-alone documentation under Documentation/bpf, as it might be more convenient format, in long term perspective. Suggested-by: Stanislav Fomichev Signed-off-by: Andrii Nakryiko --- Documentation/bpf/ringbuf.txt | 191 ++++++++++++++++++++++++++++++++++ 1 file changed, 191 insertions(+) create mode 100644 Documentation/bpf/ringbuf.txt diff --git a/Documentation/bpf/ringbuf.txt b/Documentation/bpf/ringbuf.tx= t new file mode 100644 index 000000000000..72f0a2480db3 --- /dev/null +++ b/Documentation/bpf/ringbuf.txt @@ -0,0 +1,191 @@ +BPF ring buffer +=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D=3D + +Motivation +---------- +There are two distinctive motivators for this work, which are not satisf= ied by +existing perf buffer, which prompted creation of a new ring buffer +implementation. + - more efficient memory utilization by sharing ring buffer across CPUs= ; + - preserving ordering of events that happen sequentially in time, even + across multiple CPUs (e.g., fork/exec/exit events for a task). + +These two problems are independent, but perf buffer fails to satisfy bot= h. +Both are a result of a choice to have per-CPU perf ring buffer. Both ca= n be +also solved by having an MPSC implementation of ring buffer. The orderin= g +problem could technically be solved for perf buffer with some in-kernel +counting, but given the first one requires an MPSC buffer, the same solu= tion +would solve the second problem automatically. + +Semantics and APIs +------------------ +Single ring buffer is presented to BPF programs as an instance of BPF ma= p of +type BPF_MAP_TYPE_RINGBUF. Two other alternatives considered, but ultima= tely +rejected. + +One way would be to, similar to BPF_MAP_TYPE_PERF_EVENT_ARRAY, make +BPF_MAP_TYPE_RINGBUF could represent an array of ring buffers, but not e= nforce +"same CPU only" rule. This would be more familiar interface compatible w= ith +existing perf buffer use in BPF, but would fail if application needed mo= re +advanced logic to lookup ring buffer by arbitrary key. HASH_OF_MAPS addr= esses +this with current approach. Additionally, given the performance of BPF +ringbuf, many use cases would just opt into a simple single ring buffer = shared +among all CPUs, for which current approach would be an overkill. + +Another approach could introduce a new concept, alongside BPF map, to +represent generic "container" object, which doesn't necessarily have key= /value +interface with lookup/update/delete operations. This approach would add = a lot +of extra infrastructure that has to be built for observability and verif= ier +support. It would also add another concept that BPF developers would hav= e to +familiarize themselves with, new syntax in libbpf, etc. But then would r= eally +provide no additional benefits over the approach of using a map. +BPF_MAP_TYPE_RINGBUF doesn't support lookup/update/delete operations, bu= t so +doesn't few other map types (e.g., queue and stack; array doesn't suppor= t +delete, etc). + +The approach chosen has an advantage of re-using existing BPF map +infrastructure (introspection APIs in kernel, libbpf support, etc), bein= g +familiar concept (no need to teach users a new type of object in BPF pro= gram), +and utilizing existing tooling (bpftool). For common scenario of using +a single ring buffer for all CPUs, it's as simple and straightforward, a= s +would be with a dedicated "container" object. On the other hand, by bein= g +a map, it can be combined with ARRAY_OF_MAPS and HASH_OF_MAPS map-in-map= s to +implement a wide variety of topologies, from one ring buffer for each CP= U +(e.g., as a replacement for perf buffer use cases), to a complicated +application hashing/sharding of ring buffers (e.g., having a small pool = of +ring buffers with hashed task's tgid being a look up key to preserve ord= er, +but reduce contention). + +Key and value sizes are enforced to be zero. max_entries is used to spec= ify +the size of ring buffer and has to be a power of 2 value. + +There are a bunch of similarities between perf buffer +(BPF_MAP_TYPE_PERF_EVENT_ARRAY) and new BPF ring buffer semantics: + - variable-length records; + - if there is no more space left in ring buffer, reservation fails, no + blocking; + - memory-mappable data area for user-space applications for ease of + consumption and high performance; + - epoll notifications for new incoming data; + - but still the ability to do busy polling for new data to achieve the + lowest latency, if necessary. + +BPF ringbuf provides two sets of APIs to BPF programs: + - bpf_ringbuf_output() allows to *copy* data from one place to a ring + buffer, similarly to bpf_perf_event_output(); + - bpf_ringbuf_reserve()/bpf_ringbuf_commit()/bpf_ringbuf_discard() API= s + split the whole process into two steps. First, a fixed amount of spa= ce is + reserved. If successful, a pointer to a data inside ring buffer data= area + is returned, which BPF programs can use similarly to a data inside + array/hash maps. Once ready, this piece of memory is either committe= d or + discarded. Discard is similar to commit, but makes consumer ignore t= he + record. + +bpf_ringbuf_output() has disadvantage of incurring extra memory copy, be= cause +record has to be prepared in some other place first. But it allows to su= bmit +records of the length that's not known to verifier beforehand. It also c= losely +matches bpf_perf_event_output(), so will simplify migration significantl= y. + +bpf_ringbuf_reserve() avoids the extra copy of memory by providing a mem= ory +pointer directly to ring buffer memory. In a lot of cases records are la= rger +than BPF stack space allows, so many programs have use extra per-CPU arr= ay as +a temporary heap for preparing sample. bpf_ringbuf_reserve() avoid this = needs +completely. But in exchange, it only allows a known constant size of mem= ory to +be reserved, such that verifier can verify that BPF program can't access +memory outside its reserved record space. bpf_ringbuf_output(), while sl= ightly +slower due to extra memory copy, covers some use cases that are not suit= able +for bpf_ringbuf_reserve(). + +The difference between commit and discard is very small. Discard just ma= rks +a record as discarded, and such records are supposed to be ignored by co= nsumer +code. Discard is useful for some advanced use-cases, such as ensuring +all-or-nothing multi-record submission, or emulating temporary malloc()/= free() +within single BPF program invocation. + +Each reserved record is tracked by verifier through existing +reference-tracking logic, similar to socket ref-tracking. It is thus +impossible to reserve a record, but forget to submit (or discard) it. + +bpf_ringbuf_query() helper allows to query various properties of ring bu= ffer. +Currently 4 are supported: + - BPF_RB_AVAIL_DATA returns amount of unconsumed data in ring buffer; + - BPF_RB_RING_SIZE returns the size of ring buffer; + - BPF_RB_CONS_POS/BPF_RB_PROD_POS returns current logical possition of + consumer/producer, respectively. +Returned values are momentarily snapshots of ring buffer state and could= be +off by the time helper returns, so this should be used only for +debugging/reporting reasons or for implementing various heuristics, that= take +into account highly-changeable nature of some of those characteristics. + +One such heuristic might involve more fine-grained control over poll/epo= ll +notifications about new data availability in ring buffer. Together with +BPF_RB_NO_WAKEUP/BPF_RB_FORCE_WAKEUP flags for output/commit/discard hel= pers, +it allows BPF program a high degree of control and, e.g., more efficient +batched notifications. Default self-balancing strategy, though, should b= e +adequate for most applications and will work reliable and efficiently al= ready. + +Design and implementation +------------------------- +This reserve/commit schema allows a natural way for multiple producers, = either +on different CPUs or even on the same CPU/in the same BPF program, to re= serve +independent records and work with them without blocking other producers.= This +means that if BPF program was interruped by another BPF program sharing = the +same ring buffer, they will both get a record reserved (provided there i= s +enough space left) and can work with it and submit it independently. Thi= s +applies to NMI context as well, except that due to using a spinlock duri= ng +reservation, in NMI context, bpf_ringbuf_reserve() might fail to get a l= ock, +in which case reservation will fail even if ring buffer is not full. + +The ring buffer itself internally is implemented as a power-of-2 sized +circular buffer, with two logical and ever-increasing counters (which mi= ght +wrap around on 32-bit architectures, that's not a problem): + - consumer counter shows up to which logical position consumer consume= d the + data; + - producer counter denotes amount of data reserved by all producers. + +Each time a record is reserved, producer that "owns" the record will +successfully advance producer counter. At that point, data is still not = yet +ready to be consumed, though. Each record has 8 byte header, which conta= ins +the length of reserved record, as well as two extra bits: busy bit to de= note +that record is still being worked on, and discard bit, which might be se= t at +commit time if record is discarded. In the latter case, consumer is supp= osed +to skip the record and move on to the next one. Record header also encod= es +record's relative offset from the beginning of ring buffer data area (in +pages). This allows bpf_ringbuf_commit()/bpf_ringbuf_discard() to accept= only +the pointer to the record itself, without requiring also the pointer to = ring +buffer itself. Ring buffer memory location will be restored from record +metadata header. This significantly simplifies verifier, as well as impr= oving +API usability. + +Producer counter increments are serialized under spinlock, so there is +a strict ordering between reservations. Commits, on the other hand, are +completely lockless and independent. All records become available to con= sumer +in the order of reservations, but only after all previous records where +already committed. It is thus possible for slow producers to temporarily= hold +off submitted records, that were reserved later. + +Reservation/commit/consumer protocol is verified by litmus tests in +tools/memory-model/litmus-tests, see mpsc-rb*.litmus files. + +One interesting implementation bit, that significantly simplifies (and t= hus +speeds up as well) implementation of both producers and consumers is how= data +area is mapped twice contiguously back-to-back in the virtual memory. Th= is +allows to not take any special measures for samples that have to wrap ar= ound +at the end of the circular buffer data area, because the next page after= the +last data page would be first data page again, and thus the sample will = still +appear completely contiguous in virtual memory. See comment and a simple= ASCII +diagram showing this visually in bpf_ringbuf_area_alloc(). + +Another feature that distinguishes BPF ringbuf from perf ring buffer is +a self-pacing notifications of new data being availability. +bpf_ringbuf_commit() implementation will send a notification of new reco= rd +being available after commit only if consumer has already caught up righ= t up +to the record being committed. If not, consumer still has to catch up an= d thus +will see new data anyways without needing an extra poll notification. +Benchmarks (see tools/testing/selftests/bpf/benchs/bench_ringbuf.c) show= that +this allows to achieve a very high throughput without having to resort t= o +tricks like "notify only every Nth sample", which are necessary with per= f +buffer. For extreme cases, when BPF program wants more manual control of +notifications, commit/discard/output helpers accept BPF_RB_NO_WAKEUP and +BPF_RB_FORCE_WAKEUP flags, which give full control over notifications of= data +availability, but require extra caution and diligence in using this API. --=20 2.24.1