From 7723628101aaeb1d723786747529b4ea65c5b5c5 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sun, 12 Aug 2018 10:49:27 -0700 Subject: [PATCH 1/4] bpf: Introduce bpf_skb_ancestor_cgroup_id helper == Problem description == It's useful to be able to identify cgroup associated with skb in TC so that a policy can be applied to this skb, and existing bpf_skb_cgroup_id helper can help with this. Though in real life cgroup hierarchy and hierarchy to apply a policy to don't map 1:1. It's often the case that there is a container and corresponding cgroup, but there are many more sub-cgroups inside container, e.g. because it's delegated to containerized application to control resources for its subsystems, or to separate application inside container from infra that belongs to containerization system (e.g. sshd). At the same time it may be useful to apply a policy to container as a whole. If multiple containers like this are run on a host (what is often the case) and many of them have sub-cgroups, it may not be possible to apply per-container policy in TC with existing helpers such as bpf_skb_under_cgroup or bpf_skb_cgroup_id: * bpf_skb_cgroup_id will return id of immediate cgroup associated with skb, i.e. if it's a sub-cgroup inside container, it can't be used to identify container's cgroup; * bpf_skb_under_cgroup can work only with one cgroup and doesn't scale, i.e. if there are N containers on a host and a policy has to be applied to M of them (0 <= M <= N), it'd require M calls to bpf_skb_under_cgroup, and, if M changes, it'd require to rebuild & load new BPF program. == Solution == The patch introduces new helper bpf_skb_ancestor_cgroup_id that can be used to get id of cgroup v2 that is an ancestor of cgroup associated with skb at specified level of cgroup hierarchy. That way admin can place all containers on one level of cgroup hierarchy (what is a good practice in general and already used in many configurations) and identify specific cgroup on this level no matter what sub-cgroup skb is associated with. E.g. if there is a cgroup hierarchy: root/ root/container1/ root/container1/app11/ root/container1/app11/sub-app-a/ root/container1/app12/ root/container2/ root/container2/app21/ root/container2/app22/ root/container2/app22/sub-app-b/ , then having skb associated with root/container1/app11/sub-app-a/ it's possible to get ancestor at level 1, what is container1 and apply policy for this container, or apply another policy if it's container2. Policies can be kept e.g. in a hash map where key is a container cgroup id and value is an action. Levels where container cgroups are created are usually known in advance whether cgroup hierarchy inside container may be hard to predict especially in case when its creation is delegated to containerized application. == Implementation details == The helper gets ancestor by walking parents up to specified level. Another option would be to get different kind of "id" from cgroup->ancestor_ids[level] and use it with idr_find() to get struct cgroup for ancestor. But that would require radix lookup what doesn't seem to be better (at least it's not obviously better). Format of return value of the new helper is same as that of bpf_skb_cgroup_id. Signed-off-by: Andrey Ignatov Signed-off-by: Daniel Borkmann --- include/linux/cgroup.h | 30 ++++++++++++++++++++++++++++++ include/uapi/linux/bpf.h | 21 ++++++++++++++++++++- net/core/filter.c | 28 ++++++++++++++++++++++++++++ 3 files changed, 78 insertions(+), 1 deletion(-) diff --git a/include/linux/cgroup.h b/include/linux/cgroup.h index c9fdf6f57913..32c553556bbd 100644 --- a/include/linux/cgroup.h +++ b/include/linux/cgroup.h @@ -553,6 +553,36 @@ static inline bool cgroup_is_descendant(struct cgroup *cgrp, return cgrp->ancestor_ids[ancestor->level] == ancestor->id; } +/** + * cgroup_ancestor - find ancestor of cgroup + * @cgrp: cgroup to find ancestor of + * @ancestor_level: level of ancestor to find starting from root + * + * Find ancestor of cgroup at specified level starting from root if it exists + * and return pointer to it. Return NULL if @cgrp doesn't have ancestor at + * @ancestor_level. + * + * This function is safe to call as long as @cgrp is accessible. + */ +static inline struct cgroup *cgroup_ancestor(struct cgroup *cgrp, + int ancestor_level) +{ + struct cgroup *ptr; + + if (cgrp->level < ancestor_level) + return NULL; + + for (ptr = cgrp; + ptr && ptr->level > ancestor_level; + ptr = cgroup_parent(ptr)) + ; + + if (ptr && ptr->level == ancestor_level) + return ptr; + + return NULL; +} + /** * task_under_cgroup_hierarchy - test task's membership of cgroup ancestry * @task: the task to be tested diff --git a/include/uapi/linux/bpf.h b/include/uapi/linux/bpf.h index 3102a2a23c31..66917a4eba27 100644 --- a/include/uapi/linux/bpf.h +++ b/include/uapi/linux/bpf.h @@ -2093,6 +2093,24 @@ union bpf_attr { * Return * The id is returned or 0 in case the id could not be retrieved. * + * u64 bpf_skb_ancestor_cgroup_id(struct sk_buff *skb, int ancestor_level) + * Description + * Return id of cgroup v2 that is ancestor of cgroup associated + * with the *skb* at the *ancestor_level*. The root cgroup is at + * *ancestor_level* zero and each step down the hierarchy + * increments the level. If *ancestor_level* == level of cgroup + * associated with *skb*, then return value will be same as that + * of **bpf_skb_cgroup_id**\ (). + * + * The helper is useful to implement policies based on cgroups + * that are upper in hierarchy than immediate cgroup associated + * with *skb*. + * + * The format of returned id and helper limitations are same as in + * **bpf_skb_cgroup_id**\ (). + * Return + * The id is returned or 0 in case the id could not be retrieved. + * * u64 bpf_get_current_cgroup_id(void) * Return * A 64-bit integer containing the current cgroup id based @@ -2207,7 +2225,8 @@ union bpf_attr { FN(skb_cgroup_id), \ FN(get_current_cgroup_id), \ FN(get_local_storage), \ - FN(sk_select_reuseport), + FN(sk_select_reuseport), \ + FN(skb_ancestor_cgroup_id), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call diff --git a/net/core/filter.c b/net/core/filter.c index 22906b31d43f..15b9d2df92ca 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -3778,6 +3778,32 @@ static const struct bpf_func_proto bpf_skb_cgroup_id_proto = { .ret_type = RET_INTEGER, .arg1_type = ARG_PTR_TO_CTX, }; + +BPF_CALL_2(bpf_skb_ancestor_cgroup_id, const struct sk_buff *, skb, int, + ancestor_level) +{ + struct sock *sk = skb_to_full_sk(skb); + struct cgroup *ancestor; + struct cgroup *cgrp; + + if (!sk || !sk_fullsock(sk)) + return 0; + + cgrp = sock_cgroup_ptr(&sk->sk_cgrp_data); + ancestor = cgroup_ancestor(cgrp, ancestor_level); + if (!ancestor) + return 0; + + return ancestor->kn->id.id; +} + +static const struct bpf_func_proto bpf_skb_ancestor_cgroup_id_proto = { + .func = bpf_skb_ancestor_cgroup_id, + .gpl_only = false, + .ret_type = RET_INTEGER, + .arg1_type = ARG_PTR_TO_CTX, + .arg2_type = ARG_ANYTHING, +}; #endif static unsigned long bpf_xdp_copy(void *dst_buff, const void *src_buff, @@ -4966,6 +4992,8 @@ tc_cls_act_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) #ifdef CONFIG_SOCK_CGROUP_DATA case BPF_FUNC_skb_cgroup_id: return &bpf_skb_cgroup_id_proto; + case BPF_FUNC_skb_ancestor_cgroup_id: + return &bpf_skb_ancestor_cgroup_id_proto; #endif default: return bpf_base_func_proto(func_id); From 539764d07b49c5b322cc065d275f65df275e4991 Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sun, 12 Aug 2018 10:49:28 -0700 Subject: [PATCH 2/4] bpf: Sync bpf.h to tools/ Sync skb_ancestor_cgroup_id() related bpf UAPI changes to tools/. Signed-off-by: Andrey Ignatov Signed-off-by: Daniel Borkmann --- tools/include/uapi/linux/bpf.h | 21 ++++++++++++++++++++- 1 file changed, 20 insertions(+), 1 deletion(-) diff --git a/tools/include/uapi/linux/bpf.h b/tools/include/uapi/linux/bpf.h index 3102a2a23c31..66917a4eba27 100644 --- a/tools/include/uapi/linux/bpf.h +++ b/tools/include/uapi/linux/bpf.h @@ -2093,6 +2093,24 @@ union bpf_attr { * Return * The id is returned or 0 in case the id could not be retrieved. * + * u64 bpf_skb_ancestor_cgroup_id(struct sk_buff *skb, int ancestor_level) + * Description + * Return id of cgroup v2 that is ancestor of cgroup associated + * with the *skb* at the *ancestor_level*. The root cgroup is at + * *ancestor_level* zero and each step down the hierarchy + * increments the level. If *ancestor_level* == level of cgroup + * associated with *skb*, then return value will be same as that + * of **bpf_skb_cgroup_id**\ (). + * + * The helper is useful to implement policies based on cgroups + * that are upper in hierarchy than immediate cgroup associated + * with *skb*. + * + * The format of returned id and helper limitations are same as in + * **bpf_skb_cgroup_id**\ (). + * Return + * The id is returned or 0 in case the id could not be retrieved. + * * u64 bpf_get_current_cgroup_id(void) * Return * A 64-bit integer containing the current cgroup id based @@ -2207,7 +2225,8 @@ union bpf_attr { FN(skb_cgroup_id), \ FN(get_current_cgroup_id), \ FN(get_local_storage), \ - FN(sk_select_reuseport), + FN(sk_select_reuseport), \ + FN(skb_ancestor_cgroup_id), /* integer value in 'imm' field of BPF_CALL instruction selects which helper * function eBPF program intends to call From 02f6ac7456a0d77360cb47b8d2ed4366b883096d Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sun, 12 Aug 2018 10:49:29 -0700 Subject: [PATCH 3/4] selftests/bpf: Add cgroup id helpers to bpf_helpers.h Add bpf_skb_cgroup_id and bpf_skb_ancestor_cgroup_id helpers to bpf_helpers.h to use them in tests and samples. Signed-off-by: Andrey Ignatov Signed-off-by: Daniel Borkmann --- tools/testing/selftests/bpf/bpf_helpers.h | 4 ++++ 1 file changed, 4 insertions(+) diff --git a/tools/testing/selftests/bpf/bpf_helpers.h b/tools/testing/selftests/bpf/bpf_helpers.h index 5c32266c2c38..e4be7730222d 100644 --- a/tools/testing/selftests/bpf/bpf_helpers.h +++ b/tools/testing/selftests/bpf/bpf_helpers.h @@ -139,6 +139,10 @@ static unsigned long long (*bpf_get_current_cgroup_id)(void) = (void *) BPF_FUNC_get_current_cgroup_id; static void *(*bpf_get_local_storage)(void *map, unsigned long long flags) = (void *) BPF_FUNC_get_local_storage; +static unsigned long long (*bpf_skb_cgroup_id)(void *ctx) = + (void *) BPF_FUNC_skb_cgroup_id; +static unsigned long long (*bpf_skb_ancestor_cgroup_id)(void *ctx, int level) = + (void *) BPF_FUNC_skb_ancestor_cgroup_id; /* llvm builtin functions that eBPF C program may use to * emit BPF_LD_ABS and BPF_LD_IND instructions From 5ecd8c22739b9a5f6d6431234decd912aa3f48ad Mon Sep 17 00:00:00 2001 From: Andrey Ignatov Date: Sun, 12 Aug 2018 10:49:30 -0700 Subject: [PATCH 4/4] selftests/bpf: Selftest for bpf_skb_ancestor_cgroup_id Add selftests for bpf_skb_ancestor_cgroup_id helper. test_skb_cgroup_id.sh prepares testing interface and adds tc qdisc and filter for it using BPF object compiled from test_skb_cgroup_id_kern.c program. BPF program in test_skb_cgroup_id_kern.c gets ancestor cgroup id using the new helper at different levels of cgroup hierarchy that skb belongs to, including root level and non-existing level, and saves it to the map where the key is the level of corresponding cgroup and the value is its id. To trigger BPF program, user space program test_skb_cgroup_id_user is run. It adds itself into testing cgroup and sends UDP datagram to link-local multicast address of testing interface. Then it reads cgroup ids saved in kernel for different levels from the BPF map and compares them with those in user space. They must be equal for every level of ancestry. Example of run: # ./test_skb_cgroup_id.sh Wait for testing link-local IP to become available ... OK Note: 8 bytes struct bpf_elf_map fixup performed due to size mismatch! [PASS] Signed-off-by: Andrey Ignatov Signed-off-by: Daniel Borkmann --- tools/testing/selftests/bpf/Makefile | 9 +- .../selftests/bpf/test_skb_cgroup_id.sh | 62 ++++++ .../selftests/bpf/test_skb_cgroup_id_kern.c | 47 +++++ .../selftests/bpf/test_skb_cgroup_id_user.c | 187 ++++++++++++++++++ 4 files changed, 302 insertions(+), 3 deletions(-) create mode 100755 tools/testing/selftests/bpf/test_skb_cgroup_id.sh create mode 100644 tools/testing/selftests/bpf/test_skb_cgroup_id_kern.c create mode 100644 tools/testing/selftests/bpf/test_skb_cgroup_id_user.c diff --git a/tools/testing/selftests/bpf/Makefile b/tools/testing/selftests/bpf/Makefile index daed162043c2..fff7fb1285fc 100644 --- a/tools/testing/selftests/bpf/Makefile +++ b/tools/testing/selftests/bpf/Makefile @@ -34,7 +34,8 @@ TEST_GEN_FILES = test_pkt_access.o test_xdp.o test_l4lb.o test_tcp_estats.o test test_btf_haskv.o test_btf_nokv.o test_sockmap_kern.o test_tunnel_kern.o \ test_get_stack_rawtp.o test_sockmap_kern.o test_sockhash_kern.o \ test_lwt_seg6local.o sendmsg4_prog.o sendmsg6_prog.o test_lirc_mode2_kern.o \ - get_cgroup_id_kern.o socket_cookie_prog.o test_select_reuseport_kern.o + get_cgroup_id_kern.o socket_cookie_prog.o test_select_reuseport_kern.o \ + test_skb_cgroup_id_kern.o # Order correspond to 'make run_tests' order TEST_PROGS := test_kmod.sh \ @@ -45,10 +46,11 @@ TEST_PROGS := test_kmod.sh \ test_sock_addr.sh \ test_tunnel.sh \ test_lwt_seg6local.sh \ - test_lirc_mode2.sh + test_lirc_mode2.sh \ + test_skb_cgroup_id.sh # Compile but not part of 'make run_tests' -TEST_GEN_PROGS_EXTENDED = test_libbpf_open test_sock_addr +TEST_GEN_PROGS_EXTENDED = test_libbpf_open test_sock_addr test_skb_cgroup_id_user include ../lib.mk @@ -59,6 +61,7 @@ $(TEST_GEN_PROGS): $(BPFOBJ) $(TEST_GEN_PROGS_EXTENDED): $(OUTPUT)/libbpf.a $(OUTPUT)/test_dev_cgroup: cgroup_helpers.c +$(OUTPUT)/test_skb_cgroup_id_user: cgroup_helpers.c $(OUTPUT)/test_sock: cgroup_helpers.c $(OUTPUT)/test_sock_addr: cgroup_helpers.c $(OUTPUT)/test_socket_cookie: cgroup_helpers.c diff --git a/tools/testing/selftests/bpf/test_skb_cgroup_id.sh b/tools/testing/selftests/bpf/test_skb_cgroup_id.sh new file mode 100755 index 000000000000..42544a969abc --- /dev/null +++ b/tools/testing/selftests/bpf/test_skb_cgroup_id.sh @@ -0,0 +1,62 @@ +#!/bin/sh +# SPDX-License-Identifier: GPL-2.0 +# Copyright (c) 2018 Facebook + +set -eu + +wait_for_ip() +{ + local _i + echo -n "Wait for testing link-local IP to become available " + for _i in $(seq ${MAX_PING_TRIES}); do + echo -n "." + if ping -6 -q -c 1 -W 1 ff02::1%${TEST_IF} >/dev/null 2>&1; then + echo " OK" + return + fi + sleep 1 + done + echo 1>&2 "ERROR: Timeout waiting for test IP to become available." + exit 1 +} + +setup() +{ + # Create testing interfaces not to interfere with current environment. + ip link add dev ${TEST_IF} type veth peer name ${TEST_IF_PEER} + ip link set ${TEST_IF} up + ip link set ${TEST_IF_PEER} up + + wait_for_ip + + tc qdisc add dev ${TEST_IF} clsact + tc filter add dev ${TEST_IF} egress bpf obj ${BPF_PROG_OBJ} \ + sec ${BPF_PROG_SECTION} da + + BPF_PROG_ID=$(tc filter show dev ${TEST_IF} egress | \ + awk '/ id / {sub(/.* id /, "", $0); print($1)}') +} + +cleanup() +{ + ip link del ${TEST_IF} 2>/dev/null || : + ip link del ${TEST_IF_PEER} 2>/dev/null || : +} + +main() +{ + trap cleanup EXIT 2 3 6 15 + setup + ${PROG} ${TEST_IF} ${BPF_PROG_ID} +} + +DIR=$(dirname $0) +TEST_IF="test_cgid_1" +TEST_IF_PEER="test_cgid_2" +MAX_PING_TRIES=5 +BPF_PROG_OBJ="${DIR}/test_skb_cgroup_id_kern.o" +BPF_PROG_SECTION="cgroup_id_logger" +BPF_PROG_ID=0 +PROG="${DIR}/test_skb_cgroup_id_user" + +main diff --git a/tools/testing/selftests/bpf/test_skb_cgroup_id_kern.c b/tools/testing/selftests/bpf/test_skb_cgroup_id_kern.c new file mode 100644 index 000000000000..68cf9829f5a7 --- /dev/null +++ b/tools/testing/selftests/bpf/test_skb_cgroup_id_kern.c @@ -0,0 +1,47 @@ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2018 Facebook + +#include +#include + +#include + +#include "bpf_helpers.h" + +#define NUM_CGROUP_LEVELS 4 + +struct bpf_map_def SEC("maps") cgroup_ids = { + .type = BPF_MAP_TYPE_ARRAY, + .key_size = sizeof(__u32), + .value_size = sizeof(__u64), + .max_entries = NUM_CGROUP_LEVELS, +}; + +static __always_inline void log_nth_level(struct __sk_buff *skb, __u32 level) +{ + __u64 id; + + /* [1] &level passed to external function that may change it, it's + * incompatible with loop unroll. + */ + id = bpf_skb_ancestor_cgroup_id(skb, level); + bpf_map_update_elem(&cgroup_ids, &level, &id, 0); +} + +SEC("cgroup_id_logger") +int log_cgroup_id(struct __sk_buff *skb) +{ + /* Loop unroll can't be used here due to [1]. Unrolling manually. + * Number of calls should be in sync with NUM_CGROUP_LEVELS. + */ + log_nth_level(skb, 0); + log_nth_level(skb, 1); + log_nth_level(skb, 2); + log_nth_level(skb, 3); + + return TC_ACT_OK; +} + +int _version SEC("version") = 1; + +char _license[] SEC("license") = "GPL"; diff --git a/tools/testing/selftests/bpf/test_skb_cgroup_id_user.c b/tools/testing/selftests/bpf/test_skb_cgroup_id_user.c new file mode 100644 index 000000000000..c121cc59f314 --- /dev/null +++ b/tools/testing/selftests/bpf/test_skb_cgroup_id_user.c @@ -0,0 +1,187 @@ +// SPDX-License-Identifier: GPL-2.0 +// Copyright (c) 2018 Facebook + +#include +#include +#include + +#include +#include +#include +#include +#include + + +#include +#include + +#include "bpf_rlimit.h" +#include "cgroup_helpers.h" + +#define CGROUP_PATH "/skb_cgroup_test" +#define NUM_CGROUP_LEVELS 4 + +/* RFC 4291, Section 2.7.1 */ +#define LINKLOCAL_MULTICAST "ff02::1" + +static int mk_dst_addr(const char *ip, const char *iface, + struct sockaddr_in6 *dst) +{ + memset(dst, 0, sizeof(*dst)); + + dst->sin6_family = AF_INET6; + dst->sin6_port = htons(1025); + + if (inet_pton(AF_INET6, ip, &dst->sin6_addr) != 1) { + log_err("Invalid IPv6: %s", ip); + return -1; + } + + dst->sin6_scope_id = if_nametoindex(iface); + if (!dst->sin6_scope_id) { + log_err("Failed to get index of iface: %s", iface); + return -1; + } + + return 0; +} + +static int send_packet(const char *iface) +{ + struct sockaddr_in6 dst; + char msg[] = "msg"; + int err = 0; + int fd = -1; + + if (mk_dst_addr(LINKLOCAL_MULTICAST, iface, &dst)) + goto err; + + fd = socket(AF_INET6, SOCK_DGRAM, 0); + if (fd == -1) { + log_err("Failed to create UDP socket"); + goto err; + } + + if (sendto(fd, &msg, sizeof(msg), 0, (const struct sockaddr *)&dst, + sizeof(dst)) == -1) { + log_err("Failed to send datagram"); + goto err; + } + + goto out; +err: + err = -1; +out: + if (fd >= 0) + close(fd); + return err; +} + +int get_map_fd_by_prog_id(int prog_id) +{ + struct bpf_prog_info info = {}; + __u32 info_len = sizeof(info); + __u32 map_ids[1]; + int prog_fd = -1; + int map_fd = -1; + + prog_fd = bpf_prog_get_fd_by_id(prog_id); + if (prog_fd < 0) { + log_err("Failed to get fd by prog id %d", prog_id); + goto err; + } + + info.nr_map_ids = 1; + info.map_ids = (__u64) (unsigned long) map_ids; + + if (bpf_obj_get_info_by_fd(prog_fd, &info, &info_len)) { + log_err("Failed to get info by prog fd %d", prog_fd); + goto err; + } + + if (!info.nr_map_ids) { + log_err("No maps found for prog fd %d", prog_fd); + goto err; + } + + map_fd = bpf_map_get_fd_by_id(map_ids[0]); + if (map_fd < 0) + log_err("Failed to get fd by map id %d", map_ids[0]); +err: + if (prog_fd >= 0) + close(prog_fd); + return map_fd; +} + +int check_ancestor_cgroup_ids(int prog_id) +{ + __u64 actual_ids[NUM_CGROUP_LEVELS], expected_ids[NUM_CGROUP_LEVELS]; + __u32 level; + int err = 0; + int map_fd; + + expected_ids[0] = 0x100000001; /* root cgroup */ + expected_ids[1] = get_cgroup_id(""); + expected_ids[2] = get_cgroup_id(CGROUP_PATH); + expected_ids[3] = 0; /* non-existent cgroup */ + + map_fd = get_map_fd_by_prog_id(prog_id); + if (map_fd < 0) + goto err; + + for (level = 0; level < NUM_CGROUP_LEVELS; ++level) { + if (bpf_map_lookup_elem(map_fd, &level, &actual_ids[level])) { + log_err("Failed to lookup key %d", level); + goto err; + } + if (actual_ids[level] != expected_ids[level]) { + log_err("%llx (actual) != %llx (expected), level: %u\n", + actual_ids[level], expected_ids[level], level); + goto err; + } + } + + goto out; +err: + err = -1; +out: + if (map_fd >= 0) + close(map_fd); + return err; +} + +int main(int argc, char **argv) +{ + int cgfd = -1; + int err = 0; + + if (argc < 3) { + fprintf(stderr, "Usage: %s iface prog_id\n", argv[0]); + exit(EXIT_FAILURE); + } + + if (setup_cgroup_environment()) + goto err; + + cgfd = create_and_get_cgroup(CGROUP_PATH); + if (!cgfd) + goto err; + + if (join_cgroup(CGROUP_PATH)) + goto err; + + if (send_packet(argv[1])) + goto err; + + if (check_ancestor_cgroup_ids(atoi(argv[2]))) + goto err; + + goto out; +err: + err = -1; +out: + close(cgfd); + cleanup_cgroup_environment(); + printf("[%s]\n", err ? "FAIL" : "PASS"); + return err; +}