From patchwork Mon Feb 11 15:15:03 2019 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Ira Weiny X-Patchwork-Id: 10806099 X-Patchwork-Delegate: jgg@ziepe.ca Return-Path: Received: from mail.wl.linuxfoundation.org (pdx-wl-mail.web.codeaurora.org [172.30.200.125]) by pdx-korg-patchwork-2.web.codeaurora.org (Postfix) with ESMTP id B0BBA922 for ; Mon, 11 Feb 2019 15:15:39 +0000 (UTC) Received: from mail.wl.linuxfoundation.org (localhost [127.0.0.1]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id 9DFC82A97C for ; Mon, 11 Feb 2019 15:15:39 +0000 (UTC) Received: by mail.wl.linuxfoundation.org (Postfix, from userid 486) id 925422A9A9; Mon, 11 Feb 2019 15:15:39 +0000 (UTC) X-Spam-Checker-Version: SpamAssassin 3.3.1 (2010-03-16) on pdx-wl-mail.web.codeaurora.org X-Spam-Level: X-Spam-Status: No, score=-7.9 required=2.0 tests=BAYES_00,MAILING_LIST_MULTI, RCVD_IN_DNSWL_HI autolearn=ham version=3.3.1 Received: from vger.kernel.org (vger.kernel.org [209.132.180.67]) by mail.wl.linuxfoundation.org (Postfix) with ESMTP id DD7192A97C for ; Mon, 11 Feb 2019 15:15:38 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S2391622AbfBKPPc (ORCPT ); Mon, 11 Feb 2019 10:15:32 -0500 Received: from mga01.intel.com ([192.55.52.88]:38268 "EHLO mga01.intel.com" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S2391509AbfBKPPb (ORCPT ); Mon, 11 Feb 2019 10:15:31 -0500 X-Amp-Result: SKIPPED(no attachment in message) X-Amp-File-Uploaded: False Received: from fmsmga008.fm.intel.com ([10.253.24.58]) by fmsmga101.fm.intel.com with ESMTP/TLS/DHE-RSA-AES256-GCM-SHA384; 11 Feb 2019 07:15:31 -0800 X-ExtLoop1: 1 X-IronPort-AV: E=Sophos;i="5.58,359,1544515200"; d="scan'208";a="123595514" Received: from iweiny-desk2.sc.intel.com ([10.3.52.157]) by fmsmga008.fm.intel.com with ESMTP; 11 Feb 2019 07:15:31 -0800 From: ira.weiny@intel.com To: linux-rdma@vger.kernel.org Cc: Hal Rosenstock , Doug Ledford , Jason Gunthorpe , Steven Rostedt , Ingo Molnar , Alexei Starovoitov , Daniel Borkmann , Leon Romanovsky , Ira Weiny Subject: [PATCH for-next V4 6/6] BPF: Add sample code for new ib_umad tracepoint Date: Mon, 11 Feb 2019 07:15:03 -0800 Message-Id: <20190211151503.16434-7-ira.weiny@intel.com> X-Mailer: git-send-email 2.20.1 In-Reply-To: <20190211151503.16434-1-ira.weiny@intel.com> References: <1545074124-18185-1-git-send-email-ira.weiny@intel.com> <20190211151503.16434-1-ira.weiny@intel.com> MIME-Version: 1.0 Sender: linux-rdma-owner@vger.kernel.org Precedence: bulk List-ID: X-Mailing-List: linux-rdma@vger.kernel.org X-Virus-Scanned: ClamAV using ClamSMTP From: Ira Weiny Provide a count of class types for a summary of MAD packets. The example shows one way to filter the trace data based on management class. Signed-off-by: Ira Weiny --- Changes from v3 Change dev_name to dev_index Changes from V2 Change to new umad trace layout Changes from V1 Update MAINTAINERS file MAINTAINERS | 2 + samples/bpf/Makefile | 3 + samples/bpf/ibumad_kern.c | 123 ++++++++++++++++++++++++++++++++++++++ samples/bpf/ibumad_user.c | 120 +++++++++++++++++++++++++++++++++++++ 4 files changed, 248 insertions(+) create mode 100644 samples/bpf/ibumad_kern.c create mode 100644 samples/bpf/ibumad_user.c diff --git a/MAINTAINERS b/MAINTAINERS index 18d42f6521dc..b7f07404b48b 100644 --- a/MAINTAINERS +++ b/MAINTAINERS @@ -7493,6 +7493,8 @@ F: include/uapi/rdma/ F: include/rdma/ F: include/trace/events/ib_mad.h F: include/trace/events/ib_umad.h +F: samples/bpf/ibumad_kern.c +F: samples/bpf/ibumad_user.c INGENIC JZ4780 DMA Driver M: Zubair Lutfullah Kakakhel diff --git a/samples/bpf/Makefile b/samples/bpf/Makefile index db1a91dfa702..a63c366949ab 100644 --- a/samples/bpf/Makefile +++ b/samples/bpf/Makefile @@ -53,6 +53,7 @@ hostprogs-y += xdpsock hostprogs-y += xdp_fwd hostprogs-y += task_fd_query hostprogs-y += xdp_sample_pkts +hostprogs-y += ibumad # Libbpf dependencies LIBBPF = $(TOOLS_PATH)/lib/bpf/libbpf.a @@ -109,6 +110,7 @@ xdpsock-objs := xdpsock_user.o xdp_fwd-objs := xdp_fwd_user.o task_fd_query-objs := bpf_load.o task_fd_query_user.o $(TRACE_HELPERS) xdp_sample_pkts-objs := xdp_sample_pkts_user.o $(TRACE_HELPERS) +ibumad-objs := bpf_load.o ibumad_user.o $(TRACE_HELPERS) # Tell kbuild to always build the programs always := $(hostprogs-y) @@ -167,6 +169,7 @@ always += xdpsock_kern.o always += xdp_fwd_kern.o always += task_fd_query_kern.o always += xdp_sample_pkts_kern.o +always += ibumad_kern.o KBUILD_HOSTCFLAGS += -I$(objtree)/usr/include KBUILD_HOSTCFLAGS += -I$(srctree)/tools/lib/ diff --git a/samples/bpf/ibumad_kern.c b/samples/bpf/ibumad_kern.c new file mode 100644 index 000000000000..2a90780ea5d3 --- /dev/null +++ b/samples/bpf/ibumad_kern.c @@ -0,0 +1,123 @@ +/** + * ibumad BPF sample kernel side + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of version 2 of the GNU General Public + * License as published by the Free Software Foundation. + * + * Copyright(c) 2018 Ira Weiny, Intel Corporation + */ + +#define KBUILD_MODNAME "ibumad_count_pkts_by_class" +#include + +#include "bpf_helpers.h" + + +struct bpf_map_def SEC("maps") read_count = { + .type = BPF_MAP_TYPE_ARRAY, + .key_size = sizeof(u32), /* class; u32 required */ + .value_size = sizeof(u64), /* count of mads read */ + .max_entries = 256, /* Room for all Classes */ +}; + +struct bpf_map_def SEC("maps") write_count = { + .type = BPF_MAP_TYPE_ARRAY, + .key_size = sizeof(u32), /* class; u32 required */ + .value_size = sizeof(u64), /* count of mads written */ + .max_entries = 256, /* Room for all Classes */ +}; + +#undef DEBUG +#ifdef DEBUG +#define bpf_debug(fmt, ...) \ +({ \ + char ____fmt[] = fmt; \ + bpf_trace_printk(____fmt, sizeof(____fmt), \ + ##__VA_ARGS__); \ +}) +#else +#define bpf_debug(fmt, ...) +#endif + +/* Taken from the current format defined in + * include/trace/events/ib_umad.h + * and + * /sys/kernel/debug/tracing/events/ib_umad/ib_umad_read/format + * /sys/kernel/debug/tracing/events/ib_umad/ib_umad_write/format + */ +struct ib_umad_rw_args { + u64 pad; + u8 port_num; + u8 sl; + u8 path_bits; + u8 grh_present; + u32 id; + u32 status; + u32 timeout_ms; + u32 retires; + u32 length; + u32 qpn; + u32 qkey; + u8 gid_index; + u8 hop_limit; + u16 lid; + u16 attr_id; + u16 pkey_index; + u8 base_version; + u8 mgmt_class; + u8 class_version; + u8 method; + u32 flow_label; + u16 mad_status; + u16 class_specific; + u32 attr_mod; + u64 tid; + u8 gid[16]; + u32 dev_index; + u8 traffic_class; +}; +SEC("tracepoint/ib_umad/ib_umad_read") +int on_ib_umad_read(struct ib_umad_rw_args *ctx) +{ + u64 zero = 0, *val; + u8 class = ctx->mgmt_class; + + bpf_debug("ib_umad read : class 0x%x\n", class); + + val = bpf_map_lookup_elem(&read_count, &class); + if (!val) { + bpf_map_update_elem(&read_count, &class, &zero, BPF_NOEXIST); + val = bpf_map_lookup_elem(&read_count, &class); + if (!val) + return 0; + } + + (*val) += 1; + + return 0; +} + +SEC("tracepoint/ib_umad/ib_umad_write") +int on_ib_umad_write(struct ib_umad_rw_args *ctx) +{ + u64 zero = 0, *val; + u8 class = ctx->mgmt_class; + + bpf_debug("ib_umad write : class 0x%x\n", class); + + val = bpf_map_lookup_elem(&write_count, &class); + if (!val) { + bpf_map_update_elem(&write_count, &class, &zero, BPF_NOEXIST); + val = bpf_map_lookup_elem(&write_count, &class); + if (!val) + return 0; + } + + (*val) += 1; + + return 0; +} + +char _license[] SEC("license") = "GPL"; + diff --git a/samples/bpf/ibumad_user.c b/samples/bpf/ibumad_user.c new file mode 100644 index 000000000000..c36cd99383a0 --- /dev/null +++ b/samples/bpf/ibumad_user.c @@ -0,0 +1,120 @@ +/** + * ibumad BPF sample user side + * + * This program is free software; you can redistribute it and/or + * modify it under the terms of version 2 of the GNU General Public + * License as published by the Free Software Foundation. + * + * Copyright(c) 2018 Ira Weiny, Intel Corporation + */ + +#include +#include +#include +#include +#include +#include +#include +#include + +#include +#include +#include + +#include "bpf_load.h" +#include "bpf_util.h" +#include "bpf/libbpf.h" + +static void dump_counts(int fd) +{ + __u32 key; + __u64 value; + + for (key = 0; key < 256; key++) { + if (bpf_map_lookup_elem(fd, &key, &value)) { + printf("failed to read key %u\n", key); + continue; + } + if (value) + printf("0x%02x : %llu\n", key, value); + } +} + +static void dump_all_counts(void) +{ + printf("Read 'Class : count'\n"); + dump_counts(map_fd[0]); + printf("Write 'Class : count'\n"); + dump_counts(map_fd[1]); +} + +static void dump_exit(int sig) +{ + dump_all_counts(); + exit(0); +} + +static const struct option long_options[] = { + {"help", no_argument, NULL, 'h'}, + {"delay", required_argument, NULL, 'd'}, +}; + +static void usage(char *cmd) +{ + printf("eBPF test program to count packets from various IP addresses\n" + "Usage: %s \n" + " --help, -h this menu\n" + " --delay, -d wait sec between prints [1 - 1000000]\n" + , cmd + ); +} + +int main(int argc, char **argv) +{ + unsigned long delay = 5; + int longindex = 0; + int opt; + char bpf_file[256]; + + /* Create the eBPF kernel code path name. + * This follows the pattern of all of the other bpf samples + */ + snprintf(bpf_file, sizeof(bpf_file), "%s_kern.o", argv[0]); + + /* Do one final dump when exiting */ + signal(SIGINT, dump_exit); + signal(SIGTERM, dump_exit); + + while ((opt = getopt_long(argc, argv, "hd:rSw", + long_options, &longindex)) != -1) { + switch (opt) { + case 'd': + delay = strtoul(optarg, NULL, 0); + if (delay == ULONG_MAX || delay < 0 || delay > 1000000) { + fprintf(stderr, "ERROR: invalid delay : %s\n", + optarg); + usage(argv[0]); + return 1; + } + break; + default: + case 'h': + usage(argv[0]); + return 1; + break; + } + } + + if (load_bpf_file(bpf_file)) { + fprintf(stderr, "ERROR: failed to load eBPF from file : %s\n", + bpf_file); + return 1; + } + + while (1) { + sleep(delay); + dump_all_counts(); + } + + return 0; +}