| Message ID | 20201105144755.214341-2-kpsingh@chromium.org (mailing list archive) |
|---|---|
| State | Superseded |
| Delegated to: | BPF |
| Headers | show |
| Series | Implement task_local_storage | expand |
> On Nov 5, 2020, at 6:47 AM, KP Singh <kpsingh@chromium.org> wrote: > > From: KP Singh <kpsingh@google.com> > > Usage of spin locks was not allowed for tracing programs due to > insufficient preemption checks. The verifier does not currently prevent > LSM programs from using spin locks, but the helpers are not exposed > via bpf_lsm_func_proto. > > Based on the discussion in [1], non-sleepable LSM programs should be > able to use bpf_spin_{lock, unlock}. > > Sleepable LSM programs can be preempted which means that allowng spin > locks will need more work (disabling preemption and the verifier > ensuring that no sleepable helpers are called when a spin lock is held). > > [1]: https://lore.kernel.org/bpf/20201103153132.2717326-1-kpsingh@chromium.org/T/#md601a053229287659071600d3483523f752cd2fb > > Signed-off-by: KP Singh <kpsingh@google.com> Acked-by: Song Liu <songliubraving@fb.com> [...]
On Thu, Nov 05, 2020 at 03:47:47PM +0100, KP Singh wrote: > From: KP Singh <kpsingh@google.com> > > Usage of spin locks was not allowed for tracing programs due to > insufficient preemption checks. The verifier does not currently prevent > LSM programs from using spin locks, but the helpers are not exposed > via bpf_lsm_func_proto. > > Based on the discussion in [1], non-sleepable LSM programs should be > able to use bpf_spin_{lock, unlock}. > > Sleepable LSM programs can be preempted which means that allowng spin > locks will need more work (disabling preemption and the verifier > ensuring that no sleepable helpers are called when a spin lock is held). > > [1]: https://lore.kernel.org/bpf/20201103153132.2717326-1-kpsingh@chromium.org/T/#md601a053229287659071600d3483523f752cd2fb > > Signed-off-by: KP Singh <kpsingh@google.com> Acked-by: Martin KaFai Lau <kafai@fb.com>
diff --git a/kernel/bpf/bpf_lsm.c b/kernel/bpf/bpf_lsm.c index 78ea8a7bd27f..cd8a617f2109 100644 --- a/kernel/bpf/bpf_lsm.c +++ b/kernel/bpf/bpf_lsm.c @@ -59,6 +59,10 @@ bpf_lsm_func_proto(enum bpf_func_id func_id, const struct bpf_prog *prog) return &bpf_sk_storage_get_proto; case BPF_FUNC_sk_storage_delete: return &bpf_sk_storage_delete_proto; + case BPF_FUNC_spin_lock: + return &bpf_spin_lock_proto; + case BPF_FUNC_spin_unlock: + return &bpf_spin_unlock_proto; default: return tracing_prog_func_proto(func_id, prog); } diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c index 6200519582a6..f863aa84d0a2 100644 --- a/kernel/bpf/verifier.c +++ b/kernel/bpf/verifier.c @@ -9719,11 +9719,21 @@ static int check_map_prog_compatibility(struct bpf_verifier_env *env, verbose(env, "trace type programs with run-time allocated hash maps are unsafe. Switch to preallocated hash maps.\n"); } - if ((is_tracing_prog_type(prog_type) || - prog_type == BPF_PROG_TYPE_SOCKET_FILTER) && - map_value_has_spin_lock(map)) { - verbose(env, "tracing progs cannot use bpf_spin_lock yet\n"); - return -EINVAL; + if (map_value_has_spin_lock(map)) { + if (prog_type == BPF_PROG_TYPE_SOCKET_FILTER) { + verbose(env, "socket filter progs cannot use bpf_spin_lock yet\n"); + return -EINVAL; + } + + if (is_tracing_prog_type(prog_type)) { + verbose(env, "tracing progs cannot use bpf_spin_lock yet\n"); + return -EINVAL; + } + + if (prog->aux->sleepable) { + verbose(env, "sleepable progs cannot use bpf_spin_lock yet\n"); + return -EINVAL; + } } if ((bpf_prog_is_dev_bound(prog->aux) || bpf_map_is_dev_bound(map)) &&