On 6/24/23 12:53 AM, Hou Tao wrote:
Hi,
On 6/24/2023 11:13 AM, Alexei Starovoitov wrote:
From: Alexei Starovoitov <ast@xxxxxxxxxx>
Introduce bpf_mem_[cache_]free_rcu() similar to kfree_rcu().
Unlike bpf_mem_[cache_]free() that links objects for immediate reuse into
per-cpu free list the _rcu() flavor waits for RCU grace period and then moves
objects into free_by_rcu_ttrace list where they are waiting for RCU
task trace grace period to be freed into slab.
SNIP
+static void check_free_by_rcu(struct bpf_mem_cache *c)
+{
+ struct llist_node *llnode, *t;
+
+ if (llist_empty(&c->free_by_rcu) && llist_empty(&c->free_llist_extra_rcu))
+ return;
+
+ /* drain free_llist_extra_rcu */
+ llist_for_each_safe(llnode, t, llist_del_all(&c->free_llist_extra_rcu))
+ if (__llist_add(llnode, &c->free_by_rcu))
+ c->free_by_rcu_tail = llnode;
Just like add_obj_to_free_list(), we should do conditional
local_irq_save(flags) and local_inc_return(&c->active) as well for
free_by_rcu, otherwise free_by_rcu may be corrupted by bpf program
running in a NMI context.
Good catch. Will do.
