This is a note to let you know that I've just added the patch titled bpf: Fix potential race in tail call compatibility check to the 5.14-stable tree which can be found at: http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary The filename of the patch is: bpf-fix-potential-race-in-tail-call-compatibility-check.patch and it can be found in the queue-5.14 subdirectory. If you, or anyone else, feels it should not be added to the stable tree, please let <stable@xxxxxxxxxxxxxxx> know about it. >From 54713c85f536048e685258f880bf298a74c3620d Mon Sep 17 00:00:00 2001 From: =?UTF-8?q?Toke=20H=C3=B8iland-J=C3=B8rgensen?= <toke@xxxxxxxxxx> Date: Tue, 26 Oct 2021 13:00:19 +0200 Subject: bpf: Fix potential race in tail call compatibility check MIME-Version: 1.0 Content-Type: text/plain; charset=UTF-8 Content-Transfer-Encoding: 8bit From: Toke Høiland-Jørgensen <toke@xxxxxxxxxx> commit 54713c85f536048e685258f880bf298a74c3620d upstream. Lorenzo noticed that the code testing for program type compatibility of tail call maps is potentially racy in that two threads could encounter a map with an unset type simultaneously and both return true even though they are inserting incompatible programs. The race window is quite small, but artificially enlarging it by adding a usleep_range() inside the check in bpf_prog_array_compatible() makes it trivial to trigger from userspace with a program that does, essentially: map_fd = bpf_create_map(BPF_MAP_TYPE_PROG_ARRAY, 4, 4, 2, 0); pid = fork(); if (pid) { key = 0; value = xdp_fd; } else { key = 1; value = tc_fd; } err = bpf_map_update_elem(map_fd, &key, &value, 0); While the race window is small, it has potentially serious ramifications in that triggering it would allow a BPF program to tail call to a program of a different type. So let's get rid of it by protecting the update with a spinlock. The commit in the Fixes tag is the last commit that touches the code in question. v2: - Use a spinlock instead of an atomic variable and cmpxchg() (Alexei) v3: - Put lock and the members it protects into an embedded 'owner' struct (Daniel) Fixes: 3324b584b6f6 ("ebpf: misc core cleanup") Reported-by: Lorenzo Bianconi <lorenzo.bianconi@xxxxxxxxxx> Signed-off-by: Toke Høiland-Jørgensen <toke@xxxxxxxxxx> Signed-off-by: Alexei Starovoitov <ast@xxxxxxxxxx> Link: https://lore.kernel.org/bpf/20211026110019.363464-1-toke@xxxxxxxxxx Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx> --- include/linux/bpf.h | 7 +++++-- kernel/bpf/arraymap.c | 1 + kernel/bpf/core.c | 20 +++++++++++++------- kernel/bpf/syscall.c | 6 ++++-- 4 files changed, 23 insertions(+), 11 deletions(-) --- a/include/linux/bpf.h +++ b/include/linux/bpf.h @@ -900,8 +900,11 @@ struct bpf_array_aux { * stored in the map to make sure that all callers and callees have * the same prog type and JITed flag. */ - enum bpf_prog_type type; - bool jited; + struct { + spinlock_t lock; + enum bpf_prog_type type; + bool jited; + } owner; /* Programs with direct jumps into programs part of this array. */ struct list_head poke_progs; struct bpf_map *map; --- a/kernel/bpf/arraymap.c +++ b/kernel/bpf/arraymap.c @@ -1051,6 +1051,7 @@ static struct bpf_map *prog_array_map_al INIT_WORK(&aux->work, prog_array_map_clear_deferred); INIT_LIST_HEAD(&aux->poke_progs); mutex_init(&aux->poke_mutex); + spin_lock_init(&aux->owner.lock); map = array_map_alloc(attr); if (IS_ERR(map)) { --- a/kernel/bpf/core.c +++ b/kernel/bpf/core.c @@ -1821,20 +1821,26 @@ static unsigned int __bpf_prog_ret0_warn bool bpf_prog_array_compatible(struct bpf_array *array, const struct bpf_prog *fp) { + bool ret; + if (fp->kprobe_override) return false; - if (!array->aux->type) { + spin_lock(&array->aux->owner.lock); + + if (!array->aux->owner.type) { /* There's no owner yet where we could check for * compatibility. */ - array->aux->type = fp->type; - array->aux->jited = fp->jited; - return true; + array->aux->owner.type = fp->type; + array->aux->owner.jited = fp->jited; + ret = true; + } else { + ret = array->aux->owner.type == fp->type && + array->aux->owner.jited == fp->jited; } - - return array->aux->type == fp->type && - array->aux->jited == fp->jited; + spin_unlock(&array->aux->owner.lock); + return ret; } static int bpf_check_tail_call(const struct bpf_prog *fp) --- a/kernel/bpf/syscall.c +++ b/kernel/bpf/syscall.c @@ -543,8 +543,10 @@ static void bpf_map_show_fdinfo(struct s if (map->map_type == BPF_MAP_TYPE_PROG_ARRAY) { array = container_of(map, struct bpf_array, map); - type = array->aux->type; - jited = array->aux->jited; + spin_lock(&array->aux->owner.lock); + type = array->aux->owner.type; + jited = array->aux->owner.jited; + spin_unlock(&array->aux->owner.lock); } seq_printf(m, Patches currently in stable-queue which might be from toke@xxxxxxxxxx are queue-5.14/bpf-fix-potential-race-in-tail-call-compatibility-check.patch