From mboxrd@z Thu Jan 1 00:00:00 1970 Return-Path: Received: from mails.dpdk.org (mails.dpdk.org [217.70.189.124]) by inbox.dpdk.org (Postfix) with ESMTP id 1459448ADA; Tue, 11 Nov 2025 10:47:00 +0100 (CET) Received: from mails.dpdk.org (localhost [127.0.0.1]) by mails.dpdk.org (Postfix) with ESMTP id F31B640649; Tue, 11 Nov 2025 10:46:59 +0100 (CET) Received: from frasgout.his.huawei.com (frasgout.his.huawei.com [185.176.79.56]) by mails.dpdk.org (Postfix) with ESMTP id D585840613 for ; Tue, 11 Nov 2025 10:46:58 +0100 (CET) Received: from mail.maildlp.com (unknown [172.18.186.231]) by frasgout.his.huawei.com (SkyGuard) with ESMTPS id 4d5M9v30bJzJ46jb; Tue, 11 Nov 2025 17:46:27 +0800 (CST) Received: from frapema500002.china.huawei.com (unknown [7.182.19.148]) by mail.maildlp.com (Postfix) with ESMTPS id 1C22514038F; Tue, 11 Nov 2025 17:46:58 +0800 (CST) Received: from frapema500003.china.huawei.com (7.182.19.114) by frapema500002.china.huawei.com (7.182.19.148) with Microsoft SMTP Server (version=TLS1_2, cipher=TLS_ECDHE_RSA_WITH_AES_256_GCM_SHA384) id 15.2.1544.11; Tue, 11 Nov 2025 10:46:57 +0100 Received: from frapema500003.china.huawei.com ([7.182.19.114]) by frapema500003.china.huawei.com ([7.182.19.114]) with mapi id 15.02.1544.011; Tue, 11 Nov 2025 10:46:57 +0100 From: Marat Khalili To: Stephen Hemminger CC: Konstantin Ananyev , "dev@dpdk.org" Subject: RE: [PATCH v5 4/5] bpf: add test for Rx and Tx filtering Thread-Topic: [PATCH v5 4/5] bpf: add test for Rx and Tx filtering Thread-Index: AQHcUbTYYlsbY2d8hESN8gEKhf0QBrTtOn3Q Date: Tue, 11 Nov 2025 09:46:57 +0000 Message-ID: <6a01168a950543e588d2a66a5f70023c@huawei.com> References: <20251030173732.246435-1-stephen@networkplumber.org> <20251109200854.45942-1-stephen@networkplumber.org> <20251109200854.45942-5-stephen@networkplumber.org> In-Reply-To: <20251109200854.45942-5-stephen@networkplumber.org> Accept-Language: en-US Content-Language: en-US X-MS-Has-Attach: X-MS-TNEF-Correlator: x-originating-ip: [10.206.137.70] Content-Type: text/plain; charset="us-ascii" Content-Transfer-Encoding: quoted-printable MIME-Version: 1.0 X-BeenThere: dev@dpdk.org X-Mailman-Version: 2.1.29 Precedence: list List-Id: DPDK patches and discussions List-Unsubscribe: , List-Archive: List-Post: List-Help: List-Subscribe: , Errors-To: dev-bounces@dpdk.org > -----Original Message----- > From: Stephen Hemminger > Sent: Sunday 9 November 2025 20:08 > To: dev@dpdk.org > Cc: Stephen Hemminger ; Konstantin Ananyev > Subject: [PATCH v5 4/5] bpf: add test for Rx and Tx filtering >=20 > New test using null device to test filtering with BPF. >=20 > Signed-off-by: Stephen Hemminger Acked-by: Marat Khalili (Turns out there were some issues with the pre-existing code which I initially blamed on this patch set. Maybe some cleanup is needed.) > --- > app/test/bpf/filter.c | 53 +++++++ > app/test/bpf/meson.build | 1 + > app/test/test_bpf.c | 307 +++++++++++++++++++++++++++++++++++++++ > 3 files changed, 361 insertions(+) > create mode 100644 app/test/bpf/filter.c >=20 > diff --git a/app/test/bpf/filter.c b/app/test/bpf/filter.c > new file mode 100644 > index 0000000000..d47233a47a > --- /dev/null > +++ b/app/test/bpf/filter.c > @@ -0,0 +1,53 @@ > +/* SPDX-License-Identifier: BSD-3-Clause > + * BPF TX filter program for testing rte_bpf_eth_tx_elf_load > + */ > + > +typedef unsigned char uint8_t; > +typedef unsigned short uint16_t; > +typedef unsigned int uint32_t; > +typedef unsigned long uint64_t; > + > +/* > + * Simple TX filter that accepts TCP packets > + * > + * BPF TX programs receive pointer to data and should return: > + * 0 =3D drop packet > + * non-zero =3D rx/tx packet > + * > + * This filter checks: > + * 1. Packet is IPv4 > + * 2. Protocol is TCP (IPPROTO_TCP =3D 6) > + */ > +__attribute__((section("filter"), used)) > +uint64_t > +test_filter(void *pkt) > +{ > + uint8_t *data =3D pkt; > + > + /* Read version and IHL (first byte of IP header) */ > + uint8_t version_ihl =3D data[14]; > + > + /* Check IPv4 version (upper 4 bits should be 4) */ > + if ((version_ihl >> 4) !=3D 4) > + return 0; > + > + /* Protocol field (byte 9 of IP header) must be TCP (6) */ > + uint8_t proto =3D data[14 + 9]; > + return (proto =3D=3D 6); > +} > + > +__attribute__((section("drop"), used)) > +uint64_t > +test_drop(void *pkt) > +{ > + (void)pkt; > + return 0; > +} > + > +__attribute__((section("allow"), used)) > +uint64_t > +test_allow(void *pkt) > +{ > + (void)pkt; > + return 1; > +} > diff --git a/app/test/bpf/meson.build b/app/test/bpf/meson.build > index b4f54aa976..19fec05521 100644 > --- a/app/test/bpf/meson.build > +++ b/app/test/bpf/meson.build > @@ -32,6 +32,7 @@ cflags +=3D '-DTEST_BPF_ELF_LOAD' > # BPF sources to compile > bpf_progs =3D { > 'load' : 'test_bpf_load', > + 'filter' : 'test_bpf_filter', > } >=20 > foreach bpf_src, bpf_hdr : bpf_progs > diff --git a/app/test/test_bpf.c b/app/test/test_bpf.c > index a80f41a543..24d18f9d33 100644 > --- a/app/test/test_bpf.c > +++ b/app/test/test_bpf.c > @@ -16,6 +16,12 @@ > #include > #include > #include > +#include > +#include > +#include > +#include > +#include > +#include >=20 > #include "test.h" >=20 > @@ -3421,6 +3427,291 @@ test_bpf_elf_load(void) > printf("%s: ELF load test passed\n", __func__); > return TEST_SUCCESS; > } > + > +#include "test_bpf_filter.h" > + > +#define BPF_TEST_BURST 128 > +#define BPF_TEST_POOLSIZE 256 /* at least 2x burst */ > +#define BPF_TEST_PKT_LEN 64 /* Ether + IP + TCP */ > + > +static int null_vdev_setup(const char *name, uint16_t *port, struct rte_= mempool *pool) > +{ > + int ret; > + > + /* Make a null device */ > + ret =3D rte_vdev_init(name, NULL); > + TEST_ASSERT(ret =3D=3D 0, "rte_vdev_init(%s) failed: %d", name, ret); > + > + ret =3D rte_eth_dev_get_port_by_name(name, port); > + TEST_ASSERT(ret =3D=3D 0, "failed to get port id for %s: %d", name, ret= ); > + > + struct rte_eth_conf conf =3D { }; > + ret =3D rte_eth_dev_configure(*port, 1, 1, &conf); > + TEST_ASSERT(ret =3D=3D 0, "failed to configure port %u: %d", *port, ret= ); > + > + struct rte_eth_txconf txconf =3D { }; > + ret =3D rte_eth_tx_queue_setup(*port, 0, BPF_TEST_BURST, SOCKET_ID_ANY,= &txconf); > + TEST_ASSERT(ret =3D=3D 0, "failed to setup tx queue port %u: %d", *port= , ret); > + > + struct rte_eth_rxconf rxconf =3D { }; > + ret =3D rte_eth_rx_queue_setup(*port, 0, BPF_TEST_BURST, SOCKET_ID_ANY, > + &rxconf, pool); > + TEST_ASSERT(ret =3D=3D 0, "failed to setup rx queue port %u: %d", *port= , ret); > + > + ret =3D rte_eth_dev_start(*port); > + TEST_ASSERT(ret =3D=3D 0, "failed to start port %u: %d", *port, ret); > + > + return 0; > +} > + > +static unsigned int > +setup_mbufs(struct rte_mbuf *burst[], unsigned int n) > +{ > + struct rte_ether_hdr eh =3D { > + .ether_type =3D rte_cpu_to_be_16(RTE_ETHER_TYPE_IPV4), > + }; > + const struct rte_ipv4_hdr iph =3D { > + .version_ihl =3D RTE_IPV4_VHL_DEF, > + .total_length =3D rte_cpu_to_be_16(BPF_TEST_PKT_LEN - sizeof(eh)), > + .time_to_live =3D IPDEFTTL, > + .src_addr =3D rte_cpu_to_be_32(ip_src_addr), > + .dst_addr =3D rte_cpu_to_be_32(ip_dst_addr), > + }; > + unsigned int tcp_count =3D 0; > + > + rte_eth_random_addr(eh.dst_addr.addr_bytes); > + > + for (unsigned int i =3D 0; i < n; i++) { > + struct rte_mbuf *mb =3D burst[i]; > + > + /* Setup Ethernet header */ > + *rte_pktmbuf_mtod(mb, struct rte_ether_hdr *) =3D eh; > + > + /* Setup IP header */ > + struct rte_ipv4_hdr *ip > + =3D rte_pktmbuf_mtod_offset(mb, struct rte_ipv4_hdr *, sizeof(eh)); > + *ip =3D iph; > + > + if (rte_rand() & 1) { > + struct rte_udp_hdr *udp > + =3D rte_pktmbuf_mtod_offset(mb, struct rte_udp_hdr *, > + sizeof(eh) + sizeof(iph)); > + > + ip->next_proto_id =3D IPPROTO_UDP; > + *udp =3D (struct rte_udp_hdr) { > + .src_port =3D rte_cpu_to_be_16(9), /* discard */ > + .dst_port =3D rte_cpu_to_be_16(9), /* discard */ > + .dgram_len =3D BPF_TEST_PKT_LEN - sizeof(eh) - sizeof(iph), > + }; > + > + } else { > + struct rte_tcp_hdr *tcp > + =3D rte_pktmbuf_mtod_offset(mb, struct rte_tcp_hdr *, > + sizeof(eh) + sizeof(iph)); > + > + ip->next_proto_id =3D IPPROTO_TCP; > + *tcp =3D (struct rte_tcp_hdr) { > + .src_port =3D rte_cpu_to_be_16(9), /* discard */ > + .dst_port =3D rte_cpu_to_be_16(9), /* discard */ > + .tcp_flags =3D RTE_TCP_RST_FLAG, > + }; > + ++tcp_count; > + } > + } > + > + return tcp_count; > +} > + > +static int bpf_tx_test(uint16_t port, const char *tmpfile, struct rte_me= mpool *pool, > + const char *section, uint32_t flags) > +{ > + const struct rte_bpf_prm prm =3D { > + .prog_arg =3D { > + .type =3D RTE_BPF_ARG_PTR, > + .size =3D sizeof(struct rte_mbuf), > + }, > + }; > + int ret; > + > + struct rte_mbuf *pkts[BPF_TEST_BURST] =3D { }; > + ret =3D rte_pktmbuf_alloc_bulk(pool, pkts, BPF_TEST_BURST); > + TEST_ASSERT(ret =3D=3D 0, "failed to allocate mbufs"); > + > + uint16_t expect =3D setup_mbufs(pkts, BPF_TEST_BURST); > + > + /* Try to load BPF TX program from temp file */ > + ret =3D rte_bpf_eth_tx_elf_load(port, 0, &prm, tmpfile, section, flags)= ; > + TEST_ASSERT(ret =3D=3D 0, "failed to load BPF filter from temp file %s:= %d", > + tmpfile, ret); > + > + uint16_t sent =3D rte_eth_tx_burst(port, 0, pkts, BPF_TEST_BURST); > + TEST_ASSERT_EQUAL(sent, expect, "rte_eth_tx_burst returned: %u expected= %u", > + sent, expect); > + > + /* The unsent packets should be dropped */ > + rte_pktmbuf_free_bulk(pkts + sent, BPF_TEST_BURST - sent); > + > + /* Pool should have same number of packets avail */ > + unsigned int avail =3D rte_mempool_avail_count(pool); > + TEST_ASSERT_EQUAL(avail, BPF_TEST_POOLSIZE, > + "Mempool available %u !=3D %u leaks?", avail, BPF_TEST_POOLSIZE); > + > + rte_bpf_eth_tx_unload(port, 0); > + return TEST_SUCCESS; > +} > + > +/* Test loading a transmit filter which only allows IPv4 packets */ > +static int > +test_bpf_elf_tx_load(void) > +{ > + static const char null_dev[] =3D "net_null_bpf0"; > + char *tmpfile =3D NULL; > + struct rte_mempool *mb_pool =3D NULL; > + uint16_t port =3D UINT16_MAX; > + int ret; > + > + printf("%s start\n", __func__); > + > + /* Make a pool for packets */ > + mb_pool =3D rte_pktmbuf_pool_create("bpf_tx_test_pool", BPF_TEST_POOLSI= ZE, > + 0, 0, RTE_MBUF_DEFAULT_BUF_SIZE, > + SOCKET_ID_ANY); > + > + ret =3D null_vdev_setup(null_dev, &port, mb_pool); > + if (ret !=3D 0) > + goto fail; > + > + /* Create temp file from embedded BPF object */ > + tmpfile =3D create_temp_bpf_file(app_test_bpf_filter_o, app_test_bpf_fi= lter_o_len, "tx"); > + if (tmpfile =3D=3D NULL) > + goto fail; > + > + /* Do test with VM */ > + ret =3D bpf_tx_test(port, tmpfile, mb_pool, "filter", 0); > + if (ret !=3D 0) > + goto fail; > + > + /* Repeat with JIT */ > + ret =3D bpf_tx_test(port, tmpfile, mb_pool, "filter", RTE_BPF_ETH_F_JIT= ); > + if (ret =3D=3D 0) > + printf("%s: TX ELF load test passed\n", __func__); > + > +fail: > + if (tmpfile) { > + unlink(tmpfile); > + free(tmpfile); > + } > + > + if (port !=3D UINT16_MAX) > + rte_vdev_uninit(null_dev); > + > + rte_mempool_free(mb_pool); > + > + return ret =3D=3D 0 ? TEST_SUCCESS : TEST_FAILED; > +} > + > +/* Test loading a receive filter */ > +static int bpf_rx_test(uint16_t port, const char *tmpfile, struct rte_me= mpool *pool, > + const char *section, uint32_t flags, uint16_t expected) > +{ > + struct rte_mbuf *pkts[BPF_TEST_BURST]; > + const struct rte_bpf_prm prm =3D { > + .prog_arg =3D { > + .type =3D RTE_BPF_ARG_PTR, > + .size =3D sizeof(struct rte_mbuf), > + }, > + }; > + int ret; > + > + /* Load BPF program to drop all packets */ > + ret =3D rte_bpf_eth_rx_elf_load(port, 0, &prm, tmpfile, section, flags)= ; > + TEST_ASSERT(ret =3D=3D 0, "failed to load BPF filter from temp file %s:= %d", > + tmpfile, ret); > + > + uint16_t rcvd =3D rte_eth_rx_burst(port, 0, pkts, BPF_TEST_BURST); > + TEST_ASSERT_EQUAL(rcvd, expected, > + "rte_eth_rx_burst returned: %u expect: %u", rcvd, expected); > + > + /* Drop the received packets */ > + rte_pktmbuf_free_bulk(pkts, rcvd); > + > + rte_bpf_eth_rx_unload(port, 0); > + > + /* Pool should now be full */ > + unsigned int avail =3D rte_mempool_avail_count(pool); > + TEST_ASSERT_EQUAL(avail, BPF_TEST_POOLSIZE, > + "Mempool available %u !=3D %u leaks?", avail, BPF_TEST_POOLSIZE); > + > + return TEST_SUCCESS; > +} > + > +/* Test loading a receive filters, first with drop all and then with all= ow all packets */ > +static int > +test_bpf_elf_rx_load(void) > +{ > + static const char null_dev[] =3D "net_null_bpf0"; > + struct rte_mempool *pool =3D NULL; > + char *tmpfile =3D NULL; > + uint16_t port; > + int ret; > + > + printf("%s start\n", __func__); > + > + /* Make a pool for packets */ > + pool =3D rte_pktmbuf_pool_create("bpf_rx_test_pool", 2 * BPF_TEST_BURST= , > + 0, 0, RTE_MBUF_DEFAULT_BUF_SIZE, > + SOCKET_ID_ANY); > + TEST_ASSERT(pool !=3D NULL, "failed to create mempool"); > + > + ret =3D null_vdev_setup(null_dev, &port, pool); > + if (ret !=3D 0) > + goto fail; > + > + /* Create temp file from embedded BPF object */ > + tmpfile =3D create_temp_bpf_file(app_test_bpf_filter_o, app_test_bpf_fi= lter_o_len, "rx"); > + if (tmpfile =3D=3D NULL) > + goto fail; > + > + /* Do test with VM */ > + ret =3D bpf_rx_test(port, tmpfile, pool, "drop", 0, 0); > + if (ret !=3D 0) > + goto fail; > + > + /* Repeat with JIT */ > + ret =3D bpf_rx_test(port, tmpfile, pool, "drop", RTE_BPF_ETH_F_JIT, 0); > + if (ret !=3D 0) > + goto fail; > + > + /* Repeat with allow all */ > + ret =3D bpf_rx_test(port, tmpfile, pool, "allow", 0, BPF_TEST_BURST); > + if (ret !=3D 0) > + goto fail; > + > + /* Repeat with JIT */ > + ret =3D bpf_rx_test(port, tmpfile, pool, "allow", RTE_BPF_ETH_F_JIT, BP= F_TEST_BURST); > + if (ret !=3D 0) > + goto fail; > + > + printf("%s: RX ELF load test passed\n", __func__); > + > + /* The filter should free the mbufs */ > + unsigned int avail =3D rte_mempool_avail_count(pool); > + TEST_ASSERT_EQUAL(avail, BPF_TEST_POOLSIZE, > + "Mempool available %u !=3D %u leaks?", avail, BPF_TEST_POOLSIZE); > + > +fail: > + if (tmpfile) { > + unlink(tmpfile); > + free(tmpfile); > + } > + > + if (port !=3D UINT16_MAX) > + rte_vdev_uninit(null_dev); > + > + rte_mempool_free(pool); > + > + return ret =3D=3D 0 ? TEST_SUCCESS : TEST_FAILED; > +} > #else >=20 > static int > @@ -3430,9 +3721,25 @@ test_bpf_elf_load(void) > return TEST_SKIPPED; > } >=20 > +static int > +test_bpf_elf_tx_load(void) > +{ > + printf("BPF compile not supported, skipping Tx test\n"); > + return TEST_SKIPPED; > +} > + > +static int > +test_bpf_elf_rx_load(void) > +{ > + printf("BPF compile not supported, skipping Tx test\n"); > + return TEST_SKIPPED; > +} > + > #endif /* !TEST_BPF_ELF_LOAD */ >=20 > REGISTER_FAST_TEST(bpf_elf_load_autotest, true, true, test_bpf_elf_load)= ; > +REGISTER_FAST_TEST(bpf_eth_tx_elf_load_autotest, true, true, test_bpf_el= f_tx_load); > +REGISTER_FAST_TEST(bpf_eth_rx_elf_load_autotest, true, true, test_bpf_el= f_rx_load); >=20 > #ifndef RTE_HAS_LIBPCAP >=20 > -- > 2.51.0