From patchwork Fri Jan 13 14:56:21 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jakub Sitnicki X-Patchwork-Id: 13100912 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 43881C54EBD for ; Fri, 13 Jan 2023 15:05:28 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229659AbjAMPF0 (ORCPT ); Fri, 13 Jan 2023 10:05:26 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56030 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230080AbjAMPE4 (ORCPT ); Fri, 13 Jan 2023 10:04:56 -0500 Received: from mail-ej1-x631.google.com (mail-ej1-x631.google.com [IPv6:2a00:1450:4864:20::631]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 249C1820EA for ; Fri, 13 Jan 2023 06:56:29 -0800 (PST) Received: by mail-ej1-x631.google.com with SMTP id fy8so52786452ejc.13 for ; Fri, 13 Jan 2023 06:56:29 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=cloudflare.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=tRg0fZfB0/F86fleBj0vJva1BOW0i0hgLuZy9INDlfg=; b=tkkrMh3cG4RUa1ibeKgrgxz8pa11m50Hm4gkN9LA/6MW45rjqgNLvHu75ug3xxV/Sq D3VCgVNolpquW4Mf7fbVuTjT4LgS36rQei9cWPb2NG5BINUqjfBq4StPEL+EawIaHUUN VbbU9Dschyomy2I0E6beP6UOHR1WODg3gYpAc= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=tRg0fZfB0/F86fleBj0vJva1BOW0i0hgLuZy9INDlfg=; b=i+/ahMNn0nJ0/HK/pWpa5IAGChvBXHgReVIh++pfDFps20i0fruhjpt+JV6XoqY38p S1xgANgCbHGa59Eb7zXmEfD0cB+7tg5JK+IytNJMQJXDsiOIG4jd7OwaZ5a/N/CB3kU5 R2sU5kwxDYcNr7lnJ0bhd/ARNfUSOPPhBwn5gozzeqYgF+blcSildmY59BVDMhNqsP65 prQfMEnRFpFty39++BYJnKZ5Ym5g/iSbFB5JiFuf+8ZVq4O/rzld4eY3s5xd1SjkL/1g D+vfl8fm7pU6kYNV3voINsak1Gz3a3XfNsFeSR7P1YaH+Tm4FEncivXze/S2n2cszEnB S3GQ== X-Gm-Message-State: AFqh2krnrSZojSsuwFFpQU0hG1Ugk/l+7AWbGAx7X+Mei/yCJaCq0Fx3 M5YGfEBObPsvPUr/4EMy2611woKdG4ftn27O X-Google-Smtp-Source: AMrXdXs8nUXnBEY/5CTr0aMxhZirAu0Zwvx+D8i0tlqIe3k6BW++Iiue2r5k5hQFe1G9b3W06IjT3g== X-Received: by 2002:a17:906:6dcb:b0:7c0:d60b:2887 with SMTP id j11-20020a1709066dcb00b007c0d60b2887mr69713458ejt.69.1673621787334; Fri, 13 Jan 2023 06:56:27 -0800 (PST) Received: from cloudflare.com (79.184.151.107.ipv4.supernova.orange.pl. [79.184.151.107]) by smtp.gmail.com with ESMTPSA id mb9-20020a170906eb0900b0084d34eec68esm7622474ejb.213.2023.01.13.06.56.26 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Jan 2023 06:56:26 -0800 (PST) From: Jakub Sitnicki To: netdev@vger.kernel.org Cc: kernel-team@cloudflare.com, John Fastabend , Eric Dumazet , Daniel Borkmann , Alexei Starovoitov , Andrii Nakryiko , syzbot+04c21ed96d861dccc5cd@syzkaller.appspotmail.com Subject: [PATCH bpf 1/3] bpf, sockmap: Check for any of tcp_bpf_prots when cloning a listener Date: Fri, 13 Jan 2023 15:56:21 +0100 Message-Id: <20230113-sockmap-fix-v1-1-d3cad092ee10@cloudflare.com> X-Mailer: git-send-email 2.39.0 In-Reply-To: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> References: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net A listening socket linked to a sockmap has its sk_prot overridden. It points to one of the struct proto variants in tcp_bpf_prots. The variant depends on the socket's family and which sockmap programs are attached. A child socket cloned from a TCP listener initially inherits their sk_prot. But before cloning is finished, we restore the child's proto to the listener's original non-tcp_bpf_prots one. This happens in tcp_create_openreq_child -> tcp_bpf_clone. Today, in tcp_bpf_clone we detect if the child's proto should be restored by checking only for the TCP_BPF_BASE proto variant. This is not correct. The sk_prot of listening socket linked to a sockmap can point to to any variant in tcp_bpf_prots. If the listeners sk_prot happens to be not the TCP_BPF_BASE variant, then the child socket unintentionally is left if the inherited sk_prot by tcp_bpf_clone. This leads to issues like infinite recursion on close [1], because the child state is otherwise not set up for use with tcp_bpf_prot operations. Adjust the check in tcp_bpf_clone to detect all of tcp_bpf_prots variants. Note that it wouldn't be sufficient to check the socket state when overriding the sk_prot in tcp_bpf_update_proto in order to always use the TCP_BPF_BASE variant for listening sockets. Since commit b8b8315e39ff ("bpf, sockmap: Remove unhash handler for BPF sockmap usage") it is possible for a socket to transition to TCP_LISTEN state while already linked to a sockmap, e.g. connect() -> insert into map -> connect(AF_UNSPEC) -> listen(). [1]: https://lore.kernel.org/all/00000000000073b14905ef2e7401@google.com/ Fixes: e80251555f0b ("tcp_bpf: Don't let child socket inherit parent protocol ops on copy") Reported-by: syzbot+04c21ed96d861dccc5cd@syzkaller.appspotmail.com Signed-off-by: Jakub Sitnicki --- net/ipv4/tcp_bpf.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-) diff --git a/net/ipv4/tcp_bpf.c b/net/ipv4/tcp_bpf.c index 94aad3870c5f..8db00647e0a4 100644 --- a/net/ipv4/tcp_bpf.c +++ b/net/ipv4/tcp_bpf.c @@ -639,10 +639,9 @@ EXPORT_SYMBOL_GPL(tcp_bpf_update_proto); */ void tcp_bpf_clone(const struct sock *sk, struct sock *newsk) { - int family = sk->sk_family == AF_INET6 ? TCP_BPF_IPV6 : TCP_BPF_IPV4; struct proto *prot = newsk->sk_prot; - if (prot == &tcp_bpf_prots[family][TCP_BPF_BASE]) + if (tcp_bpf_prots[0] <= prot && prot < tcp_bpf_prots[ARRAY_SIZE(tcp_bpf_prots)]) newsk->sk_prot = sk->sk_prot_creator; } #endif /* CONFIG_BPF_SYSCALL */ From patchwork Fri Jan 13 14:56:22 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jakub Sitnicki X-Patchwork-Id: 13100911 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 3B41FC61DB3 for ; Fri, 13 Jan 2023 15:05:27 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S230085AbjAMPFZ (ORCPT ); Fri, 13 Jan 2023 10:05:25 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:59208 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230031AbjAMPE5 (ORCPT ); Fri, 13 Jan 2023 10:04:57 -0500 Received: from mail-ej1-x632.google.com (mail-ej1-x632.google.com [IPv6:2a00:1450:4864:20::632]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id E697B840B3 for ; Fri, 13 Jan 2023 06:56:30 -0800 (PST) Received: by mail-ej1-x632.google.com with SMTP id u9so52958288ejo.0 for ; Fri, 13 Jan 2023 06:56:30 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=cloudflare.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=mIDvWvTceP3dMNuGMTkeWoOyskDpFQfCLlKp/mFzwdo=; b=GzTy4J5xXqTXF4Jw+mLcdbqI2yMawGLiUvGCYBM1LyBD57VFTxVY7TDieiskByvqEd eXs3ubeSrFZE29KA+Emjgwr274+J0QFPMm/Js9kJSooQiGDTSZPRF8YtVGazcXcKMKt/ usoXQ1DvgKt0blFGXo3k58LGoBe9SKR5MgjJs= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=mIDvWvTceP3dMNuGMTkeWoOyskDpFQfCLlKp/mFzwdo=; b=aQhqt+SOa33b5kHyDYvNJWRJ4OKcC3oNRAr/e0nFZvr6wD4CQqpNiLFoxEfjlxX0Rc 2XN5h6AC5Yq3OUKSI9qDComcBIOFEd+k+dopswhhaod8z0kMua/lk/SRlaL2FcqfgvLj OVtuqltxLgzO2L6vMav100AM2yeGu+/yxDI2GjtyJNFEMOoGMRj+K13KSNPYuEoGZ00T QuzmZwGfvzzTlDmFpfcPn5DxiD/4iCu8NCzUY39sJd6O4Oc4YLT/blDcoM+rIK3INf1p DUs1gYC4ts1AFQcTY7UYs3BLAZVfgB4+APkTpSUxPlGNub+4tgCg/6gBcV3E7Lb8ZWji LTtw== X-Gm-Message-State: AFqh2koL+X5BZMrP/h7vI1G6hvDzgvWNPkyJvpQI/sq8PXHxkhhWioDL UE+k8W8RZwopJZMGDPBYly0LB3OFRREexXev X-Google-Smtp-Source: AMrXdXsx3Bg3ETDgCUDmg2wWRk1RcdxV5L2gpmU5Z42R+vRNrSxtd24irNcFI4HpaSd3/k307RcUjQ== X-Received: by 2002:a17:907:874c:b0:7fc:4242:f9ec with SMTP id qo12-20020a170907874c00b007fc4242f9ecmr69143299ejc.43.1673621789067; Fri, 13 Jan 2023 06:56:29 -0800 (PST) Received: from cloudflare.com (79.184.151.107.ipv4.supernova.orange.pl. [79.184.151.107]) by smtp.gmail.com with ESMTPSA id d9-20020a1709063ec900b007bd9e683639sm8564377ejj.130.2023.01.13.06.56.28 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Jan 2023 06:56:28 -0800 (PST) From: Jakub Sitnicki To: netdev@vger.kernel.org Cc: kernel-team@cloudflare.com, John Fastabend , Eric Dumazet , Daniel Borkmann , Alexei Starovoitov , Andrii Nakryiko Subject: [PATCH bpf 2/3] selftests/bpf: Pass BPF skeleton to sockmap_listen ops tests Date: Fri, 13 Jan 2023 15:56:22 +0100 Message-Id: <20230113-sockmap-fix-v1-2-d3cad092ee10@cloudflare.com> X-Mailer: git-send-email 2.39.0 In-Reply-To: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> References: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net Following patch extends the sockmap ops tests to cover the scenario when a sockmap with attached programs holds listening sockets. Pass the BPF skeleton to sockmap ops test so that the can access and attach the BPF programs. Signed-off-by: Jakub Sitnicki --- .../selftests/bpf/prog_tests/sockmap_listen.c | 55 +++++++++++++++------- 1 file changed, 37 insertions(+), 18 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c index 2cf0c7a3fe23..499fba8f55b9 100644 --- a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c +++ b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c @@ -30,6 +30,8 @@ #define MAX_STRERR_LEN 256 #define MAX_TEST_NAME 80 +#define __always_unused __attribute__((__unused__)) + #define _FAIL(errnum, fmt...) \ ({ \ error_at_line(0, (errnum), __func__, __LINE__, fmt); \ @@ -321,7 +323,8 @@ static int socket_loopback(int family, int sotype) return socket_loopback_reuseport(family, sotype, -1); } -static void test_insert_invalid(int family, int sotype, int mapfd) +static void test_insert_invalid(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u32 key = 0; u64 value; @@ -338,7 +341,8 @@ static void test_insert_invalid(int family, int sotype, int mapfd) FAIL_ERRNO("map_update: expected EBADF"); } -static void test_insert_opened(int family, int sotype, int mapfd) +static void test_insert_opened(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u32 key = 0; u64 value; @@ -359,7 +363,8 @@ static void test_insert_opened(int family, int sotype, int mapfd) xclose(s); } -static void test_insert_bound(int family, int sotype, int mapfd) +static void test_insert_bound(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct sockaddr_storage addr; socklen_t len; @@ -386,7 +391,8 @@ static void test_insert_bound(int family, int sotype, int mapfd) xclose(s); } -static void test_insert(int family, int sotype, int mapfd) +static void test_insert(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u64 value; u32 key; @@ -402,7 +408,8 @@ static void test_insert(int family, int sotype, int mapfd) xclose(s); } -static void test_delete_after_insert(int family, int sotype, int mapfd) +static void test_delete_after_insert(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u64 value; u32 key; @@ -419,7 +426,8 @@ static void test_delete_after_insert(int family, int sotype, int mapfd) xclose(s); } -static void test_delete_after_close(int family, int sotype, int mapfd) +static void test_delete_after_close(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { int err, s; u64 value; @@ -442,7 +450,8 @@ static void test_delete_after_close(int family, int sotype, int mapfd) FAIL_ERRNO("map_delete: expected EINVAL/EINVAL"); } -static void test_lookup_after_insert(int family, int sotype, int mapfd) +static void test_lookup_after_insert(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u64 cookie, value; socklen_t len; @@ -470,7 +479,8 @@ static void test_lookup_after_insert(int family, int sotype, int mapfd) xclose(s); } -static void test_lookup_after_delete(int family, int sotype, int mapfd) +static void test_lookup_after_delete(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { int err, s; u64 value; @@ -493,7 +503,8 @@ static void test_lookup_after_delete(int family, int sotype, int mapfd) xclose(s); } -static void test_lookup_32_bit_value(int family, int sotype, int mapfd) +static void test_lookup_32_bit_value(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { u32 key, value32; int err, s; @@ -523,7 +534,8 @@ static void test_lookup_32_bit_value(int family, int sotype, int mapfd) xclose(s); } -static void test_update_existing(int family, int sotype, int mapfd) +static void test_update_existing(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { int s1, s2; u64 value; @@ -551,7 +563,8 @@ static void test_update_existing(int family, int sotype, int mapfd) /* Exercise the code path where we destroy child sockets that never * got accept()'ed, aka orphans, when parent socket gets closed. */ -static void test_destroy_orphan_child(int family, int sotype, int mapfd) +static void test_destroy_orphan_child(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct sockaddr_storage addr; socklen_t len; @@ -585,7 +598,8 @@ static void test_destroy_orphan_child(int family, int sotype, int mapfd) /* Perform a passive open after removing listening socket from SOCKMAP * to ensure that callbacks get restored properly. */ -static void test_clone_after_delete(int family, int sotype, int mapfd) +static void test_clone_after_delete(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct sockaddr_storage addr; socklen_t len; @@ -621,7 +635,8 @@ static void test_clone_after_delete(int family, int sotype, int mapfd) * SOCKMAP, but got accept()'ed only after the parent has been removed * from SOCKMAP, gets cloned without parent psock state or callbacks. */ -static void test_accept_after_delete(int family, int sotype, int mapfd) +static void test_accept_after_delete(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct sockaddr_storage addr; const u32 zero = 0; @@ -675,7 +690,8 @@ static void test_accept_after_delete(int family, int sotype, int mapfd) /* Check that child socket that got created and accepted while parent * was in a SOCKMAP is cloned without parent psock state or callbacks. */ -static void test_accept_before_delete(int family, int sotype, int mapfd) +static void test_accept_before_delete(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct sockaddr_storage addr; const u32 zero = 0, one = 1; @@ -784,7 +800,8 @@ static void *connect_accept_thread(void *arg) return NULL; } -static void test_syn_recv_insert_delete(int family, int sotype, int mapfd) +static void test_syn_recv_insert_delete(struct test_sockmap_listen *skel __always_unused, + int family, int sotype, int mapfd) { struct connect_accept_ctx ctx = { 0 }; struct sockaddr_storage addr; @@ -847,7 +864,8 @@ static void *listen_thread(void *arg) return NULL; } -static void test_race_insert_listen(int family, int socktype, int mapfd) +static void test_race_insert_listen(struct test_sockmap_listen *skel __always_unused, + int family, int socktype, int mapfd) { struct connect_accept_ctx ctx = { 0 }; const u32 zero = 0; @@ -1473,7 +1491,8 @@ static void test_ops(struct test_sockmap_listen *skel, struct bpf_map *map, int family, int sotype) { const struct op_test { - void (*fn)(int family, int sotype, int mapfd); + void (*fn)(struct test_sockmap_listen *skel, + int family, int sotype, int mapfd); const char *name; int sotype; } tests[] = { @@ -1520,7 +1539,7 @@ static void test_ops(struct test_sockmap_listen *skel, struct bpf_map *map, if (!test__start_subtest(s)) continue; - t->fn(family, sotype, map_fd); + t->fn(skel, family, sotype, map_fd); test_ops_cleanup(map); } } From patchwork Fri Jan 13 14:56:23 2023 Content-Type: text/plain; charset="utf-8" MIME-Version: 1.0 Content-Transfer-Encoding: 7bit X-Patchwork-Submitter: Jakub Sitnicki X-Patchwork-Id: 13100913 X-Patchwork-Delegate: bpf@iogearbox.net Return-Path: X-Spam-Checker-Version: SpamAssassin 3.4.0 (2014-02-07) on aws-us-west-2-korg-lkml-1.web.codeaurora.org Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by smtp.lore.kernel.org (Postfix) with ESMTP id 648EEC54EBE for ; Fri, 13 Jan 2023 15:05:30 +0000 (UTC) Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S229774AbjAMPF2 (ORCPT ); Fri, 13 Jan 2023 10:05:28 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:56028 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S230090AbjAMPE6 (ORCPT ); Fri, 13 Jan 2023 10:04:58 -0500 Received: from mail-ej1-x629.google.com (mail-ej1-x629.google.com [IPv6:2a00:1450:4864:20::629]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id 6D3DB840BD for ; Fri, 13 Jan 2023 06:56:32 -0800 (PST) Received: by mail-ej1-x629.google.com with SMTP id v6so9900385ejg.6 for ; Fri, 13 Jan 2023 06:56:32 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=cloudflare.com; s=google; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:from:to:cc:subject:date :message-id:reply-to; bh=LjX4FJ8gRkEtjzAf9WCo0Kq35+A6/erPh54GkDAerRk=; b=RfXMs7/aeN1F0Luowy5Aok/0MUmvWvZ3FaoYpRALRgx0SweCHM1wGPBr8Zc7JeZaHV SiezbtQ39pChfrPSyLvsZzdt1sNLEx4SWY5sZEQaz6QBjruK9VqUnDayRpPO2bTGLXzs /I+WiyOd164IMMuPGQ4lU7tz8TapDa+40ObB0= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20210112; h=content-transfer-encoding:mime-version:references:in-reply-to :message-id:date:subject:cc:to:from:x-gm-message-state:from:to:cc :subject:date:message-id:reply-to; bh=LjX4FJ8gRkEtjzAf9WCo0Kq35+A6/erPh54GkDAerRk=; b=hJmoeY0UUuGYfGmL7FLdxVNXjYXDIz1M22R6nGrSlHc2PrGKXv2lqGKUnWAB33+sJ0 m41n6cr0Q5Zon69ukj7gV7JQuFsFgc6BPe2XU78npqOoRyivC5Z5RCvAFiXjAa1KeJwj 57/j4EEoKikmmxzLPA5LwxETngcrMq/vuIt2g8P+5ukra5eaDqD/Tea15KjekpiwwgF2 +xfCsFPremia2Y3gOOvSpu3vt0gobZ6pdunY84/O+Kp5k7d9v7LIPOotErZEu6K6Sc+d emDOdqxgVIbPrGB8RBxt5ORQs7jzpHitm5b2+i+nhpL3uowqN+bSmDa9mduZ2VHT5Xze ggLA== X-Gm-Message-State: AFqh2kp9GXoTauYNWEa5YRS5Q9ejWHwPdoEiVm9D/G7lX66nMT+Gm3YA j+v5SFJA5RidYqqQJB3mffsomZxqp4KNgveR X-Google-Smtp-Source: AMrXdXuA/B7ppypRtstNR3RalKoTpfUXLkQGkLy3JGT6BWhvudDUOeLbHK75uqrUnlOiFDXqxFDUMg== X-Received: by 2002:a17:907:d10f:b0:83f:cbfd:31a9 with SMTP id uy15-20020a170907d10f00b0083fcbfd31a9mr62946200ejc.47.1673621790638; Fri, 13 Jan 2023 06:56:30 -0800 (PST) Received: from cloudflare.com (79.184.151.107.ipv4.supernova.orange.pl. [79.184.151.107]) by smtp.gmail.com with ESMTPSA id b9-20020a1709063ca900b007c073be0127sm8589112ejh.202.2023.01.13.06.56.29 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Fri, 13 Jan 2023 06:56:30 -0800 (PST) From: Jakub Sitnicki To: netdev@vger.kernel.org Cc: kernel-team@cloudflare.com, John Fastabend , Eric Dumazet , Daniel Borkmann , Alexei Starovoitov , Andrii Nakryiko Subject: [PATCH bpf 3/3] selftests/bpf: Cover listener cloning with progs attached to sockmap Date: Fri, 13 Jan 2023 15:56:23 +0100 Message-Id: <20230113-sockmap-fix-v1-3-d3cad092ee10@cloudflare.com> X-Mailer: git-send-email 2.39.0 In-Reply-To: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> References: <20230113-sockmap-fix-v1-0-d3cad092ee10@cloudflare.com> MIME-Version: 1.0 Precedence: bulk List-ID: X-Mailing-List: netdev@vger.kernel.org X-Patchwork-Delegate: bpf@iogearbox.net Today we test if a child socket is cloned properly from a listening socket inside a sockmap only when there are no BPF programs attached to the map. A bug has been reported [1] for the case when sockmap has a verdict program attached. So cover this case as well to prevent regressions. [1]: https://lore.kernel.org/r/00000000000073b14905ef2e7401@google.com Signed-off-by: Jakub Sitnicki --- .../selftests/bpf/prog_tests/sockmap_listen.c | 30 ++++++++++++++++++++-- 1 file changed, 28 insertions(+), 2 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c index 499fba8f55b9..567e07c19ecc 100644 --- a/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c +++ b/tools/testing/selftests/bpf/prog_tests/sockmap_listen.c @@ -563,8 +563,7 @@ static void test_update_existing(struct test_sockmap_listen *skel __always_unuse /* Exercise the code path where we destroy child sockets that never * got accept()'ed, aka orphans, when parent socket gets closed. */ -static void test_destroy_orphan_child(struct test_sockmap_listen *skel __always_unused, - int family, int sotype, int mapfd) +static void do_destroy_orphan_child(int family, int sotype, int mapfd) { struct sockaddr_storage addr; socklen_t len; @@ -595,6 +594,33 @@ static void test_destroy_orphan_child(struct test_sockmap_listen *skel __always_ xclose(s); } +static void test_destroy_orphan_child(struct test_sockmap_listen *skel, + int family, int sotype, int mapfd) +{ + int msg_verdict = bpf_program__fd(skel->progs.prog_msg_verdict); + int skb_verdict = bpf_program__fd(skel->progs.prog_skb_verdict); + const struct test { + int progfd; + enum bpf_attach_type atype; + } tests[] = { + { -1, -1 }, + { msg_verdict, BPF_SK_MSG_VERDICT }, + { skb_verdict, BPF_SK_SKB_VERDICT }, + }; + const struct test *t; + + for (t = tests; t < tests + ARRAY_SIZE(tests); t++) { + if (t->progfd != -1 && + xbpf_prog_attach(t->progfd, mapfd, t->atype, 0) != 0) + return; + + do_destroy_orphan_child(family, sotype, mapfd); + + if (t->progfd != -1) + xbpf_prog_detach2(t->progfd, mapfd, t->atype); + } +} + /* Perform a passive open after removing listening socket from SOCKMAP * to ensure that callbacks get restored properly. */