From: Hou Tao <houtao1@xxxxxxxxxx>
When removing the inner map from the outer map, the inner map will be
freed after one RCU grace period and one RCU tasks trace grace
period, so it is certain that the bpf program, which may access the
inner map, has exited before the inner map is freed.
However there is no need to wait for one RCU tasks trace grace period if
the outer map is only accessed by non-sleepable program. So adding
sleepable_refcnt in bpf_map and increasing sleepable_refcnt when adding
the outer map into env->used_maps for sleepable program. Although the
max number of bpf program is INT_MAX - 1, the number of bpf programs
which are being loaded may be greater than INT_MAX, so using atomic64_t
instead of atomic_t for sleepable_refcnt. When removing the inner map
from the outer map, using sleepable_refcnt to decide whether or not a
RCU tasks trace grace period is needed before freeing the inner map.
Signed-off-by: Hou Tao <houtao1@xxxxxxxxxx>
---
include/linux/bpf.h | 2 ++
kernel/bpf/core.c | 4 ++++
kernel/bpf/map_in_map.c | 14 +++++++++-----
kernel/bpf/syscall.c | 8 ++++++++
kernel/bpf/verifier.c | 4 +++-
5 files changed, 26 insertions(+), 6 deletions(-)
diff --git a/include/linux/bpf.h b/include/linux/bpf.h
index 15a6bb951b70..4a78cc0c81c8 100644
--- a/include/linux/bpf.h
+++ b/include/linux/bpf.h
@@ -297,6 +297,8 @@ struct bpf_map {
bool bypass_spec_v1;
bool frozen; /* write-once; write-protected by freeze_mutex */
bool free_after_mult_rcu_gp;
+ bool free_after_rcu_gp;
+ atomic64_t sleepable_refcnt;
s64 __percpu *elem_count;
};
diff --git a/kernel/bpf/core.c b/kernel/bpf/core.c
index cd3afe57ece3..4b813da8d6c0 100644
--- a/kernel/bpf/core.c
+++ b/kernel/bpf/core.c
@@ -2664,12 +2664,16 @@ void __bpf_free_used_maps(struct bpf_prog_aux *aux,
struct bpf_map **used_maps, u32 len)
{
struct bpf_map *map;
+ bool sleepable;
u32 i;
+ sleepable = aux->sleepable;
for (i = 0; i < len; i++) {
map = used_maps[i];
if (map->ops->map_poke_untrack)
map->ops->map_poke_untrack(map, aux);
+ if (sleepable)
+ atomic64_dec(&map->sleepable_refcnt);
bpf_map_put(map);
}
}
diff --git a/kernel/bpf/map_in_map.c b/kernel/bpf/map_in_map.c
index 3248ff5d8161..8ef269e66ba5 100644
--- a/kernel/bpf/map_in_map.c
+++ b/kernel/bpf/map_in_map.c
@@ -131,12 +131,16 @@ void bpf_map_fd_put_ptr(struct bpf_map *map, void *ptr, bool need_defer)
{
struct bpf_map *inner_map = ptr;
- /* The inner map may still be used by both non-sleepable and sleepable
- * bpf program, so free it after one RCU grace period and one tasks
- * trace RCU grace period.
+ /* Defer the freeing of inner map according to the sleepable attribute
+ * of bpf program which owns the outer map, so unnecessary waiting for
+ * RCU tasks trace grace period can be avoided.
*/
- if (need_defer)
- WRITE_ONCE(inner_map->free_after_mult_rcu_gp, true);
+ if (need_defer) {
+ if (atomic64_read(&map->sleepable_refcnt))
+ WRITE_ONCE(inner_map->free_after_mult_rcu_gp, true);
+ else
+ WRITE_ONCE(inner_map->free_after_rcu_gp, true);
+ }
bpf_map_put(inner_map);
}
diff --git a/kernel/bpf/syscall.c b/kernel/bpf/syscall.c
index 88882cb58121..7ba10570ca97 100644
--- a/kernel/bpf/syscall.c
+++ b/kernel/bpf/syscall.c
@@ -750,8 +750,11 @@ void bpf_map_put(struct bpf_map *map)
bpf_map_free_id(map);
btf_put(map->btf);
+ WARN_ON_ONCE(atomic64_read(&map->sleepable_refcnt));
if (READ_ONCE(map->free_after_mult_rcu_gp))
call_rcu_tasks_trace(&map->rcu, bpf_map_free_mult_rcu_gp);
+ else if (READ_ONCE(map->free_after_rcu_gp))
+ call_rcu(&map->rcu, bpf_map_free_rcu_gp);
else
bpf_map_free_in_work(map);
}
@@ -5343,6 +5346,11 @@ static int bpf_prog_bind_map(union bpf_attr *attr)
goto out_unlock;
}
+ /* The bpf program will not access the bpf map, but for the sake of
+ * simplicity, increase sleepable_refcnt for sleepable program as well.
+ */
+ if (prog->aux->sleepable)
+ atomic64_inc(&map->sleepable_refcnt);
memcpy(used_maps_new, used_maps_old,
sizeof(used_maps_old[0]) * prog->aux->used_map_cnt);
used_maps_new[prog->aux->used_map_cnt] = map;
diff --git a/kernel/bpf/verifier.c b/kernel/bpf/verifier.c
index 6da370a047fe..0c018cedf47e 100644
--- a/kernel/bpf/verifier.c
+++ b/kernel/bpf/verifier.c
@@ -18051,10 +18051,12 @@ static int resolve_pseudo_ldimm64(struct bpf_verifier_env *env)
return -E2BIG;
}
+ if (env->prog->aux->sleepable)
+ atomic64_inc(&map->sleepable_refcnt);
/* hold the map. If the program is rejected by verifier,
* the map will be released by release_maps() or it
* will be used by the valid program until it's unloaded
- * and all maps are released in free_used_maps()
+ * and all maps are released in bpf_free_used_maps()
*/
bpf_map_inc(map);
--
2.29.2
