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=-10.1 required=3.0 tests=BAYES_00,DKIMWL_WL_HIGH, DKIM_SIGNED,DKIM_VALID,DKIM_VALID_AU,INCLUDES_PATCH,MAILING_LIST_MULTI, SIGNED_OFF_BY,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 ED30CC4363A for ; Thu, 29 Oct 2020 18:36:27 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id 7B49020825 for ; Thu, 29 Oct 2020 18:36:27 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1603996587; bh=1zNo5J90tuLOYAACGimlMDt5iKknN7hXSGTCcYmDCiM=; h=References:In-Reply-To:From:Date:Subject:To:Cc:List-ID:From; b=1BFzVLLy8zYwYjnnsIBYYKZg2Kt0cIibuV2kvVUPhApk8bsfoICutNznkHvHHqJOs 6kM9F69AZv6LBoDe4T2itOLoi9V3tf7tE4IWVImaeCQVrQLV7Exgf9r80s6OKIv62W OGQ48VybjlHQcWwzcILKNb15eOUUlemMSeFDBqmE= Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1725870AbgJ2Sg0 (ORCPT ); Thu, 29 Oct 2020 14:36:26 -0400 Received: from mail.kernel.org ([198.145.29.99]:35636 "EHLO mail.kernel.org" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1725769AbgJ2Sg0 (ORCPT ); Thu, 29 Oct 2020 14:36:26 -0400 Received: from mail-lf1-f47.google.com (mail-lf1-f47.google.com [209.85.167.47]) (using TLSv1.2 with cipher ECDHE-RSA-AES128-GCM-SHA256 (128/128 bits)) (No client certificate requested) by mail.kernel.org (Postfix) with ESMTPSA id 2C45120759; Thu, 29 Oct 2020 18:36:24 +0000 (UTC) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/simple; d=kernel.org; s=default; t=1603996584; bh=1zNo5J90tuLOYAACGimlMDt5iKknN7hXSGTCcYmDCiM=; h=References:In-Reply-To:From:Date:Subject:To:Cc:From; b=tT8YuBPg94R1AMuklLBpbjwaTNczGoEM9mqH9LbisHpUYKxQ911wExDiYMrCcw5k/ GycHpts+RxUcZ8bfHu5i0ODCx7goNSFztIiiEjlNcmUUQwgOIWjn+02b/lesWItqz6 qThxZb0IutBpUdTBjVqekP1nMrbQDz49GkHsvbws= Received: by mail-lf1-f47.google.com with SMTP id b1so4550604lfp.11; Thu, 29 Oct 2020 11:36:24 -0700 (PDT) X-Gm-Message-State: AOAM532hSDiDiQdAHiBWJXdVZJNxYxh9iDXn8RXnADtj+Bd6Npg/HAQY UOZqYD8QRimqDZHouyxUEF/H2XueQkscYsSWTvk= X-Google-Smtp-Source: ABdhPJzgHlLM/3uYT1Smzn4i22TwEX9JYWMD85Bsd/BGgFJTEI5Sg28ykPy+82rzy5W5XB0SirLw8b9Bbiz9Ex+LWNw= X-Received: by 2002:a19:804d:: with SMTP id b74mr2038802lfd.55.1603996582303; Thu, 29 Oct 2020 11:36:22 -0700 (PDT) MIME-Version: 1.0 References: <20201029111730.6881-1-david.verbeiren@tessares.net> In-Reply-To: <20201029111730.6881-1-david.verbeiren@tessares.net> From: Song Liu Date: Thu, 29 Oct 2020 11:36:10 -0700 X-Gmail-Original-Message-ID: Message-ID: Subject: Re: [PATCH bpf] selftest/bpf: Validate initial values of per-cpu hash elems To: David Verbeiren Cc: bpf , Networking , Andrii Nakryiko Content-Type: text/plain; charset="UTF-8" Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org On Thu, Oct 29, 2020 at 4:19 AM David Verbeiren wrote: > > Tests that when per-cpu hash map or LRU hash map elements are > re-used as a result of a bpf program inserting elements, the > element values for the other CPUs than the one executing the > BPF code are reset to 0. > > This validates the fix proposed in: > https://lkml.kernel.org/bpf/20201027221324.27894-1-david.verbeiren@tessares.net/ > > Change-Id: I38bc7b3744ed40704a7b2cc6efa179fb344c4bee > Suggested-by: Andrii Nakryiko > Signed-off-by: David Verbeiren > --- > .../selftests/bpf/prog_tests/map_init.c | 204 ++++++++++++++++++ > 1 file changed, 204 insertions(+) > create mode 100644 tools/testing/selftests/bpf/prog_tests/map_init.c > > diff --git a/tools/testing/selftests/bpf/prog_tests/map_init.c b/tools/testing/selftests/bpf/prog_tests/map_init.c > new file mode 100644 > index 000000000000..9640cf925908 > --- /dev/null > +++ b/tools/testing/selftests/bpf/prog_tests/map_init.c > @@ -0,0 +1,204 @@ > +// SPDX-License-Identifier: GPL-2.0-only > +// Copyright (c) 2020 Tessares SA > + > +#include > + > +#define TEST_VALUE 0x1234 > + > +static int nr_cpus; > +static int duration; > +static char bpf_log_buf[BPF_LOG_BUF_SIZE]; > + > +typedef unsigned long long map_key_t; > +typedef unsigned long long map_value_t; > +typedef struct { > + map_value_t v; /* padding */ > +} __bpf_percpu_val_align pcpu_map_value_t; > + > +/* executes bpf program that updates map with key, value */ > +static int bpf_prog_insert_elem(int fd, map_key_t key, map_value_t value) > +{ > + struct bpf_load_program_attr prog; > + struct bpf_insn insns[] = { > + BPF_LD_IMM64(BPF_REG_8, key), > + BPF_LD_IMM64(BPF_REG_9, value), > + > + /* update: R1=fd, R2=&key, R3=&value, R4=flags */ > + BPF_LD_MAP_FD(BPF_REG_1, fd), > + BPF_MOV64_REG(BPF_REG_2, BPF_REG_10), > + BPF_ALU64_IMM(BPF_ADD, BPF_REG_2, -8), > + BPF_STX_MEM(BPF_DW, BPF_REG_2, BPF_REG_8, 0), > + BPF_MOV64_REG(BPF_REG_3, BPF_REG_2), > + BPF_ALU64_IMM(BPF_ADD, BPF_REG_3, -8), > + BPF_STX_MEM(BPF_DW, BPF_REG_3, BPF_REG_9, 0), > + BPF_MOV64_IMM(BPF_REG_4, 0), > + BPF_EMIT_CALL(BPF_FUNC_map_update_elem), > + > + BPF_MOV64_IMM(BPF_REG_0, 0), > + BPF_EXIT_INSN(), > + }; Impressive hand written assembly. ;-) I would recommend using skeleton for future work. For example: BPF program: selftests/bpf/progs/bpf_iter_bpf_map.c Use the program in tests: selftests/bpf/prog_tests/bpf_iter.c:#include "bpf_iter_bpf_map.skel.h" > + char buf[64] = {}; > + int pfd, err; > + __u32 retval = 0; > + > + memset(&prog, 0, sizeof(prog)); > + prog.prog_type = BPF_PROG_TYPE_SCHED_CLS; > + prog.insns = insns; > + prog.insns_cnt = ARRAY_SIZE(insns); > + prog.license = "GPL"; > + > + pfd = bpf_load_program_xattr(&prog, bpf_log_buf, BPF_LOG_BUF_SIZE); > + if (CHECK(pfd < 0, "bpf_load_program_xattr", "failed: %s\n%s\n", > + strerror(errno), bpf_log_buf)) > + return -1; > + > + err = bpf_prog_test_run(pfd, 1, buf, sizeof(buf), NULL, NULL, > + &retval, NULL); > + if (CHECK(err || retval, "bpf_prog_test_run", > + "err=%d retval=%d errno=%d\n", err, retval, errno)) > + err = -1; > + > + close(pfd); > + > + return err; > +} > + > +static int check_values_one_cpu(pcpu_map_value_t *value, map_value_t expected) > +{ > + int i, nzCnt = 0; > + map_value_t val; > + > + for (i = 0; i < nr_cpus; i++) { > + val = bpf_percpu(value, i); > + if (val) { > + if (val != expected) { > + PRINT_FAIL("Unexpected value (cpu %d): 0x%llx\n", > + i, val); I guess we can also use CHECK() here? > + return -1; > + } [...] > + > + /* delete key=1 element so it will later be re-used*/ > + key = 1; > + err = bpf_map_delete_elem(map_fd, &key); > + if (CHECK(err, "bpf_map_delete_elem", "failed: %s\n", strerror(errno))) > + goto error_map; > + > + /* run bpf prog that inserts new elem, re-using the slot just freed */ > + err = bpf_prog_insert_elem(map_fd, key, TEST_VALUE); > + if (!ASSERT_OK(err, "bpf_prog_insert_elem")) > + goto error_map; What's the reason to use ASSERT_OK() instead of CHECK()? > + > + /* check that key=1 was re-created by bpf prog */ > + err = bpf_map_lookup_elem(map_fd, &key, value); > + if (CHECK(err, "bpf_map_lookup_elem", "failed: %s\n", strerror(errno))) > + goto error_map; > + > + /* and has expected value for just a single CPU, 0 for all others */ > + check_values_one_cpu(value, TEST_VALUE); > + > +error_map: > + close(map_fd); > +} > + > +/* Add key=1 and key=2 elems with values set for all CPUs > + * Run bpf prog that inserts new key=3 elem > + * (only for current cpu; other cpus should have initial value = 0) > + * Lookup Key=1 and check value is as expected for all CPUs > + */ > +static void test_pcpu_lru_map_init(void) > +{ > + pcpu_map_value_t value[nr_cpus]; > + int map_fd, err; > + map_key_t key; > + > + /* Set up LRU map with 2 elements, values filled for all CPUs. > + * With these 2 elements, the LRU map is full > + */ > + map_fd = map_setup(BPF_MAP_TYPE_LRU_PERCPU_HASH, 2, 2); > + if (CHECK(map_fd < 0, "map_setup", "failed\n")) > + return; > + > + /* run bpf prog that inserts new key=3 element, re-using LRU slot */ > + key = 3; > + err = bpf_prog_insert_elem(map_fd, key, TEST_VALUE); > + if (!ASSERT_OK(err, "bpf_prog_insert_elem")) > + goto error_map; ditto > + > + /* check that key=3 present */ > + err = bpf_map_lookup_elem(map_fd, &key, value); > + if (CHECK(err, "bpf_map_lookup_elem", "failed: %s\n", strerror(errno))) > + goto error_map; > + > + /* and has expected value for just a single CPU, 0 for all others */ > + check_values_one_cpu(value, TEST_VALUE); > + > +error_map: > + close(map_fd); > +} > + > +void test_map_init(void) > +{ > + nr_cpus = bpf_num_possible_cpus(); > + if (CHECK(nr_cpus <= 1, "nr_cpus", "> 1 needed for this test")) > + return; Instead of failing the test, let's skip the tests with something like: printf("%s:SKIP: >1 cpu needed for this test\n", __func__); test__skip(); > + > + if (test__start_subtest("pcpu_map_init")) > + test_pcpu_map_init(); > + if (test__start_subtest("pcpu_lru_map_init")) > + test_pcpu_lru_map_init(); > +} > -- > 2.29.0 >