bpf: Add per-program recursion prevention mechanism

Since both sleepable and non-sleepable programs execute under migrate_disable
add recursion prevention mechanism to both types of programs when they're
executed via bpf trampoline.

Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: Andrii Nakryiko <andrii@kernel.org>
Link: https://lore.kernel.org/bpf/20210210033634.62081-5-alexei.starovoitov@gmail.com

2 files changed, 27 insertions, 4 deletions

diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index 2cf71fd39c22..334070c4b8a1 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -91,6 +91,12 @@ struct bpf_prog *bpf_prog_alloc_no_stats(unsigned int size, gfp_t gfp_extra_flag
 		vfree(fp);
 		return NULL;
 	}
+	fp->active = alloc_percpu_gfp(int, GFP_KERNEL_ACCOUNT | gfp_extra_flags);
+	if (!fp->active) {
+		vfree(fp);
+		kfree(aux);
+		return NULL;
+	}
 
 	fp->pages = size / PAGE_SIZE;
 	fp->aux = aux;
@@ -116,6 +122,7 @@ struct bpf_prog *bpf_prog_alloc(unsigned int size, gfp_t gfp_extra_flags)
 
 	prog->stats = alloc_percpu_gfp(struct bpf_prog_stats, gfp_flags);
 	if (!prog->stats) {
+		free_percpu(prog->active);
 		kfree(prog->aux);
 		vfree(prog);
 		return NULL;
@@ -253,6 +260,7 @@ void __bpf_prog_free(struct bpf_prog *fp)
 		kfree(fp->aux);
 	}
 	free_percpu(fp->stats);
+	free_percpu(fp->active);
 	vfree(fp);
 }
 
diff --git a/kernel/bpf/trampoline.c b/kernel/bpf/trampoline.c
index 48eb021e1421..89ef6320d19b 100644
--- a/kernel/bpf/trampoline.c
+++ b/kernel/bpf/trampoline.c
@@ -381,13 +381,16 @@ out:
 	mutex_unlock(&trampoline_mutex);
 }
 
-#define NO_START_TIME 0
+#define NO_START_TIME 1
 static u64 notrace bpf_prog_start_time(void)
 {
 	u64 start = NO_START_TIME;
 
-	if (static_branch_unlikely(&bpf_stats_enabled_key))
+	if (static_branch_unlikely(&bpf_stats_enabled_key)) {
 		start = sched_clock();
+		if (unlikely(!start))
+			start = NO_START_TIME;
+	}
 	return start;
 }
 
@@ -397,12 +400,20 @@ static u64 notrace bpf_prog_start_time(void)
  * call __bpf_prog_enter
  * call prog->bpf_func
  * call __bpf_prog_exit
+ *
+ * __bpf_prog_enter returns:
+ * 0 - skip execution of the bpf prog
+ * 1 - execute bpf prog
+ * [2..MAX_U64] - excute bpf prog and record execution time.
+ *     This is start time.
  */
-u64 notrace __bpf_prog_enter(void)
+u64 notrace __bpf_prog_enter(struct bpf_prog *prog)
 	__acquires(RCU)
 {
 	rcu_read_lock();
 	migrate_disable();
+	if (unlikely(__this_cpu_inc_return(*(prog->active)) != 1))
+		return 0;
 	return bpf_prog_start_time();
 }
 
@@ -430,21 +441,25 @@ void notrace __bpf_prog_exit(struct bpf_prog *prog, u64 start)
 	__releases(RCU)
 {
 	update_prog_stats(prog, start);
+	__this_cpu_dec(*(prog->active));
 	migrate_enable();
 	rcu_read_unlock();
 }
 
-u64 notrace __bpf_prog_enter_sleepable(void)
+u64 notrace __bpf_prog_enter_sleepable(struct bpf_prog *prog)
 {
 	rcu_read_lock_trace();
 	migrate_disable();
 	might_fault();
+	if (unlikely(__this_cpu_inc_return(*(prog->active)) != 1))
+		return 0;
 	return bpf_prog_start_time();
 }
 
 void notrace __bpf_prog_exit_sleepable(struct bpf_prog *prog, u64 start)
 {
 	update_prog_stats(prog, start);
+	__this_cpu_dec(*(prog->active));
 	migrate_enable();
 	rcu_read_unlock_trace();
 }