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=-12.9 required=3.0 tests=BAYES_00,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 47DC0C55179 for ; Wed, 4 Nov 2020 16:45:11 +0000 (UTC) Received: from vger.kernel.org (vger.kernel.org [23.128.96.18]) by mail.kernel.org (Postfix) with ESMTP id DD015206DB for ; Wed, 4 Nov 2020 16:45:10 +0000 (UTC) Authentication-Results: mail.kernel.org; dkim=pass (1024-bit key) header.d=chromium.org header.i=@chromium.org header.b="UJ7ujSh6" Received: (majordomo@vger.kernel.org) by vger.kernel.org via listexpand id S1731299AbgKDQpJ (ORCPT ); Wed, 4 Nov 2020 11:45:09 -0500 Received: from lindbergh.monkeyblade.net ([23.128.96.19]:38984 "EHLO lindbergh.monkeyblade.net" rhost-flags-OK-OK-OK-OK) by vger.kernel.org with ESMTP id S1731296AbgKDQpJ (ORCPT ); Wed, 4 Nov 2020 11:45:09 -0500 Received: from mail-ej1-x641.google.com (mail-ej1-x641.google.com [IPv6:2a00:1450:4864:20::641]) by lindbergh.monkeyblade.net (Postfix) with ESMTPS id AF248C0613D3 for ; Wed, 4 Nov 2020 08:45:07 -0800 (PST) Received: by mail-ej1-x641.google.com with SMTP id 7so30720212ejm.0 for ; Wed, 04 Nov 2020 08:45:07 -0800 (PST) DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=chromium.org; s=google; h=from:to:cc:subject:date:message-id:in-reply-to:references :mime-version:content-transfer-encoding; bh=bfcTcfxBUXUo/rquci5lcju84a9v0kx0TTqHn8xfiB8=; b=UJ7ujSh69Wh5aI7ynE5Ko1d1SYD+hCUXBTm6PY2XZC0EhamlA/WTizCwz4HLrDZ3J1 wLySxn0C7btP2sZFn0nf8MxA6MfnapjeZWKZ5Yltsy+AfSWp1LlNArnF59KN4V9LXMjB nNGlAo7Kgc/7uysi5AGVaBsBd9lLpBB/cLWDI= X-Google-DKIM-Signature: v=1; a=rsa-sha256; c=relaxed/relaxed; d=1e100.net; s=20161025; h=x-gm-message-state:from:to:cc:subject:date:message-id:in-reply-to :references:mime-version:content-transfer-encoding; bh=bfcTcfxBUXUo/rquci5lcju84a9v0kx0TTqHn8xfiB8=; b=bx+wfvDtKYA+oDj4JdSqP4pQhFoYDAKsKmF+rK9UkvDhXdG4dt1NIV/+pJXG4cfzs7 rzPy3QwDv696g99+WMWbNfCBpcTQD24eX80lHMtdKTGEXuhvcQiwLLV932rpEBxGflPs KYdPUCpKrv+/GLIieRuELlMfOVv7Sgp7dPTxxr1qrJYdboHqFRvGpeViYYTPt5sflSIA g19lMsx6MHjjwdQ9pSpl/ncpxwSVo1+NAnt24eijh1Vt+Gr+/8tEUgcskupxgDAOJgw6 wSxzV8B8oLWEjmEJSrUAekoFf6QysgyCLXEdRi021P75QVnKpYeUKL9JLzpnuIbe63Yo PTCw== X-Gm-Message-State: AOAM530SR6JvPMe+Zns02asbqGt09bX7LCYISkYno0t/tjp3Zzr1l6Za n5eRRHtvCRhK5pDnea/54PSBbw== X-Google-Smtp-Source: ABdhPJz53x2Zvuwsp9K7HsSE1V+e+abzlGxbGGGOJVEgukWWcqaDDKwNcxeIlX6WoRCrc36q/rAz1Q== X-Received: by 2002:a17:906:1955:: with SMTP id b21mr26779304eje.42.1604508306201; Wed, 04 Nov 2020 08:45:06 -0800 (PST) Received: from kpsingh.zrh.corp.google.com ([81.6.44.51]) by smtp.gmail.com with ESMTPSA id g20sm1283551ejz.88.2020.11.04.08.45.05 (version=TLS1_3 cipher=TLS_AES_256_GCM_SHA384 bits=256/256); Wed, 04 Nov 2020 08:45:05 -0800 (PST) From: KP Singh To: linux-kernel@vger.kernel.org, bpf@vger.kernel.org Cc: Alexei Starovoitov , Daniel Borkmann , Martin KaFai Lau , Song Liu , Paul Turner , Jann Horn , Hao Luo Subject: [PATCH bpf-next v3 8/9] bpf: Add tests for task_local_storage Date: Wed, 4 Nov 2020 17:44:52 +0100 Message-Id: <20201104164453.74390-9-kpsingh@chromium.org> X-Mailer: git-send-email 2.29.1.341.ge80a0c044ae-goog In-Reply-To: <20201104164453.74390-1-kpsingh@chromium.org> References: <20201104164453.74390-1-kpsingh@chromium.org> MIME-Version: 1.0 Content-Transfer-Encoding: 8bit Precedence: bulk List-ID: X-Mailing-List: bpf@vger.kernel.org From: KP Singh The test exercises the syscall based map operations by creating a pidfd for the current process. For verifying kernel / LSM functionality, the test implements a simple MAC policy which denies an executable from unlinking itself. The LSM program bprm_committed_creds sets a task_local_storage with a pointer to the inode. This is then used to detect if the task is trying to unlink itself in the inode_unlink LSM hook. The test copies /bin/rm to /tmp and executes it in a child thread with the intention of deleting itself. A successful test should prevent the the running executable from deleting itself. The bpf programs are also updated to call bpf_spin_{lock, unlock} to trigger the verfier checks for spin locks. The temporary file is cleaned up later in the test. Signed-off-by: KP Singh --- .../bpf/prog_tests/test_local_storage.c | 167 ++++++++++++++++-- .../selftests/bpf/progs/local_storage.c | 61 ++++++- 2 files changed, 210 insertions(+), 18 deletions(-) diff --git a/tools/testing/selftests/bpf/prog_tests/test_local_storage.c b/tools/testing/selftests/bpf/prog_tests/test_local_storage.c index 91cd6f357246..feba23f8848b 100644 --- a/tools/testing/selftests/bpf/prog_tests/test_local_storage.c +++ b/tools/testing/selftests/bpf/prog_tests/test_local_storage.c @@ -4,30 +4,149 @@ * Copyright (C) 2020 Google LLC. */ +#define _GNU_SOURCE + +#include +#include +#include #include #include #include "local_storage.skel.h" #include "network_helpers.h" -int create_and_unlink_file(void) +static inline int sys_pidfd_open(pid_t pid, unsigned int flags) +{ + return syscall(__NR_pidfd_open, pid, flags); +} + +unsigned int duration; + +#define TEST_STORAGE_VALUE 0xbeefdead + +struct storage { + void *inode; + unsigned int value; + /* Lock ensures that spin locked versions of local stoage operations + * also work, most operations in this tests are still single threaded + */ + struct bpf_spin_lock lock; +}; + +/* Copies an rm binary to a temp file. dest is a mkstemp template */ +int copy_rm(char *dest) { - char fname[PATH_MAX] = "/tmp/fileXXXXXX"; - int fd; + int ret, fd_in, fd_out; + struct stat stat; - fd = mkstemp(fname); - if (fd < 0) - return fd; + fd_in = open("/bin/rm", O_RDONLY); + if (fd_in < 0) + return fd_in; - close(fd); - unlink(fname); + fd_out = mkstemp(dest); + if (fd_out < 0) + return fd_out; + + ret = fstat(fd_in, &stat); + if (ret == -1) + return errno; + + ret = copy_file_range(fd_in, NULL, fd_out, NULL, stat.st_size, 0); + if (ret == -1) + return errno; + + /* Set executable permission on the copied file */ + ret = chmod(dest, 0100); + if (ret == -1) + return errno; + + close(fd_in); + close(fd_out); return 0; } +/* Fork and exec the provided rm binary and return the exit code of the + * forked process and its pid. + */ +int run_self_unlink(int *monitored_pid, const char *rm_path) +{ + int child_pid, child_status, ret; + int null_fd; + + child_pid = fork(); + if (child_pid == 0) { + null_fd = open("/dev/null", O_WRONLY); + dup2(null_fd, STDOUT_FILENO); + dup2(null_fd, STDERR_FILENO); + close(null_fd); + + *monitored_pid = getpid(); + /* Use the copied /usr/bin/rm to delete itself + * /tmp/copy_of_rm /tmp/copy_of_rm. + */ + ret = execlp(rm_path, rm_path, rm_path, NULL); + if (ret) + exit(errno); + } else if (child_pid > 0) { + waitpid(child_pid, &child_status, 0); + return WEXITSTATUS(child_status); + } + + return -EINVAL; +} + +bool check_syscall_operations(int map_fd, int obj_fd) +{ + struct storage val = { .value = TEST_STORAGE_VALUE, .lock = { 0 } }, + lookup_val = { .value = 0, .lock = { 0 } }; + int err; + + /* Looking up an existing element should fail initially */ + err = bpf_map_lookup_elem_flags(map_fd, &obj_fd, &lookup_val, + BPF_F_LOCK); + if (CHECK(!err || errno != ENOENT, "bpf_map_lookup_elem", + "err:%d errno:%d\n", err, errno)) + return false; + + /* Create a new element */ + err = bpf_map_update_elem(map_fd, &obj_fd, &val, + BPF_NOEXIST | BPF_F_LOCK); + if (CHECK(err < 0, "bpf_map_update_elem", "err:%d errno:%d\n", err, + errno)) + return false; + + /* Lookup the newly created element */ + err = bpf_map_lookup_elem_flags(map_fd, &obj_fd, &lookup_val, + BPF_F_LOCK); + if (CHECK(err < 0, "bpf_map_lookup_elem", "err:%d errno:%d", err, + errno)) + return false; + + /* Check the value of the newly created element */ + if (CHECK(lookup_val.value != val.value, "bpf_map_lookup_elem", + "value got = %x errno:%d", lookup_val.value, val.value)) + return false; + + err = bpf_map_delete_elem(map_fd, &obj_fd); + if (CHECK(err, "bpf_map_delete_elem()", "err:%d errno:%d\n", err, + errno)) + return false; + + /* The lookup should fail, now that the element has been deleted */ + err = bpf_map_lookup_elem_flags(map_fd, &obj_fd, &lookup_val, + BPF_F_LOCK); + if (CHECK(!err || errno != ENOENT, "bpf_map_lookup_elem", + "err:%d errno:%d\n", err, errno)) + return false; + + return true; +} + void test_test_local_storage(void) { + char tmp_exec_path[PATH_MAX] = "/tmp/copy_of_rmXXXXXX"; + int err, serv_sk = -1, task_fd = -1; struct local_storage *skel = NULL; - int err, duration = 0, serv_sk = -1; skel = local_storage__open_and_load(); if (CHECK(!skel, "skel_load", "lsm skeleton failed\n")) @@ -37,10 +156,35 @@ void test_test_local_storage(void) if (CHECK(err, "attach", "lsm attach failed: %d\n", err)) goto close_prog; + task_fd = sys_pidfd_open(getpid(), 0); + if (CHECK(task_fd < 0, "pidfd_open", + "failed to get pidfd err:%d, errno:%d", task_fd, errno)) + goto close_prog; + + if (!check_syscall_operations(bpf_map__fd(skel->maps.task_storage_map), + task_fd)) + goto close_prog; + + err = copy_rm(tmp_exec_path); + if (CHECK(err < 0, "copy_rm", "err %d errno %d\n", err, errno)) + goto close_prog; + + /* Sets skel->bss->monitored_pid to the pid of the forked child + * forks a child process that executes tmp_exec_path and tries to + * unlink its executable. This operation should be denied by the loaded + * LSM program. + */ + err = run_self_unlink(&skel->bss->monitored_pid, tmp_exec_path); + if (CHECK(err != EPERM, "run_self_unlink", "err %d want EPERM\n", err)) + goto close_prog; + + /* Set the process being monitored to be the current process */ skel->bss->monitored_pid = getpid(); - err = create_and_unlink_file(); - if (CHECK(err < 0, "exec_cmd", "err %d errno %d\n", err, errno)) + /* Remove the temporary created executable */ + err = unlink(tmp_exec_path); + if (CHECK(err != 0, "unlink", "unable to unlink %s: %d", tmp_exec_path, + errno)) goto close_prog; CHECK(skel->data->inode_storage_result != 0, "inode_storage_result", @@ -56,5 +200,6 @@ void test_test_local_storage(void) close(serv_sk); close_prog: + close(task_fd); local_storage__destroy(skel); } diff --git a/tools/testing/selftests/bpf/progs/local_storage.c b/tools/testing/selftests/bpf/progs/local_storage.c index ef3822bc7542..3e3de130f28f 100644 --- a/tools/testing/selftests/bpf/progs/local_storage.c +++ b/tools/testing/selftests/bpf/progs/local_storage.c @@ -17,41 +17,64 @@ int monitored_pid = 0; int inode_storage_result = -1; int sk_storage_result = -1; -struct dummy_storage { +struct local_storage { + struct inode *exec_inode; __u32 value; + struct bpf_spin_lock lock; }; struct { __uint(type, BPF_MAP_TYPE_INODE_STORAGE); __uint(map_flags, BPF_F_NO_PREALLOC); __type(key, int); - __type(value, struct dummy_storage); + __type(value, struct local_storage); } inode_storage_map SEC(".maps"); struct { __uint(type, BPF_MAP_TYPE_SK_STORAGE); __uint(map_flags, BPF_F_NO_PREALLOC | BPF_F_CLONE); __type(key, int); - __type(value, struct dummy_storage); + __type(value, struct local_storage); } sk_storage_map SEC(".maps"); +struct { + __uint(type, BPF_MAP_TYPE_TASK_STORAGE); + __uint(map_flags, BPF_F_NO_PREALLOC); + __type(key, int); + __type(value, struct local_storage); +} task_storage_map SEC(".maps"); + SEC("lsm/inode_unlink") int BPF_PROG(unlink_hook, struct inode *dir, struct dentry *victim) { __u32 pid = bpf_get_current_pid_tgid() >> 32; - struct dummy_storage *storage; + struct local_storage *storage; + bool is_self_unlink; int err; if (pid != monitored_pid) return 0; + storage = bpf_task_storage_get(&task_storage_map, + bpf_get_current_task_btf(), 0, 0); + if (storage) { + /* Don't let an executable delete itself */ + bpf_spin_lock(&storage->lock); + is_self_unlink = storage->exec_inode == victim->d_inode; + bpf_spin_unlock(&storage->lock); + if (is_self_unlink) + return -EPERM; + } + storage = bpf_inode_storage_get(&inode_storage_map, victim->d_inode, 0, BPF_LOCAL_STORAGE_GET_F_CREATE); if (!storage) return 0; + bpf_spin_lock(&storage->lock); if (storage->value != DUMMY_STORAGE_VALUE) inode_storage_result = -1; + bpf_spin_unlock(&storage->lock); err = bpf_inode_storage_delete(&inode_storage_map, victim->d_inode); if (!err) @@ -65,7 +88,7 @@ int BPF_PROG(socket_bind, struct socket *sock, struct sockaddr *address, int addrlen) { __u32 pid = bpf_get_current_pid_tgid() >> 32; - struct dummy_storage *storage; + struct local_storage *storage; int err; if (pid != monitored_pid) @@ -76,8 +99,10 @@ int BPF_PROG(socket_bind, struct socket *sock, struct sockaddr *address, if (!storage) return 0; + bpf_spin_lock(&storage->lock); if (storage->value != DUMMY_STORAGE_VALUE) sk_storage_result = -1; + bpf_spin_unlock(&storage->lock); err = bpf_sk_storage_delete(&sk_storage_map, sock->sk); if (!err) @@ -91,7 +116,7 @@ int BPF_PROG(socket_post_create, struct socket *sock, int family, int type, int protocol, int kern) { __u32 pid = bpf_get_current_pid_tgid() >> 32; - struct dummy_storage *storage; + struct local_storage *storage; if (pid != monitored_pid) return 0; @@ -101,7 +126,9 @@ int BPF_PROG(socket_post_create, struct socket *sock, int family, int type, if (!storage) return 0; + bpf_spin_lock(&storage->lock); storage->value = DUMMY_STORAGE_VALUE; + bpf_spin_unlock(&storage->lock); return 0; } @@ -110,7 +137,7 @@ SEC("lsm/file_open") int BPF_PROG(file_open, struct file *file) { __u32 pid = bpf_get_current_pid_tgid() >> 32; - struct dummy_storage *storage; + struct local_storage *storage; if (pid != monitored_pid) return 0; @@ -123,6 +150,26 @@ int BPF_PROG(file_open, struct file *file) if (!storage) return 0; + bpf_spin_lock(&storage->lock); storage->value = DUMMY_STORAGE_VALUE; + bpf_spin_unlock(&storage->lock); return 0; } + +/* This uses the local storage to remember the inode of the binary that a + * process was originally executing. + */ +SEC("lsm/bprm_committed_creds") +void BPF_PROG(exec, struct linux_binprm *bprm) +{ + struct local_storage *storage; + + storage = bpf_task_storage_get(&task_storage_map, + bpf_get_current_task_btf(), 0, + BPF_LOCAL_STORAGE_GET_F_CREATE); + if (storage) { + bpf_spin_lock(&storage->lock); + storage->exec_inode = bprm->file->f_inode; + bpf_spin_unlock(&storage->lock); + } +} -- 2.29.1.341.ge80a0c044ae-goog