bpf: Revert "bpf: drop KF_ACQUIRE flag on BPF kfunc bpf_get_root_mem_cgroup()"

This reverts commit e463b6de9da1 ("bpf: drop KF_ACQUIRE flag on BPF
kfunc bpf_get_root_mem_cgroup()").

The original commit removed the KF_ACQUIRE flag from
bpf_get_root_mem_cgroup() under the assumption that it resulted in
simplified usage. This stemmed from the fact that
bpf_get_root_mem_cgroup() inherently returns a reference to an object
which technically isn't reference counted, therefore there is no
strong requirement to call a matching bpf_put_mem_cgroup() on the
returned reference.

Although technically correct, as per the arguments in the thread [0],
dropping the KF_ACQUIRE flag and losing reference tracking semantics
negatively impacted the usability of bpf_get_root_mem_cgroup() in
practice.

[0] https://lore.kernel.org/bpf/878qdx6yut.fsf@linux.dev/

Link: https://lore.kernel.org/bpf/CAADnVQ+6d1Lj4dteAv8u62d7kj3Ze5io6bqM0xeQd-UPk9ZgJQ@mail.gmail.com/
Signed-off-by: Matt Bobrowski <mattbobrowski@google.com>
Link: https://lore.kernel.org/r/20260121090001.240166-1-mattbobrowski@google.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

diff --git a/mm/bpf_memcontrol.c b/mm/bpf_memcontrol.c
index f95cd5d16f4cd93acd9a53e6d5e62c55b17b86f1..716df49d76477d135a55874a5989968a3b645653 100644 (file)
--- a/mm/bpf_memcontrol.c
+++ b/mm/bpf_memcontrol.c
@@ -13,12 +13,12 @@ __bpf_kfunc_start_defs();
 /**
  * bpf_get_root_mem_cgroup - Returns a pointer to the root memory cgroup
  *
- * Return: A pointer to the root memory cgroup. Notably, the pointer to the
- * returned struct mem_cgroup is trusted by default, so it's perfectably
- * acceptable to also pass this pointer into other BPF kfuncs (e.g.,
- * bpf_mem_cgroup_usage()). Additionally, this BPF kfunc does not make use of
- * KF_ACQUIRE semantics, so there's no requirement for the BPF program to call
- * bpf_put_mem_cgroup() on the pointer returned by this BPF kfunc.
+ * The function has KF_ACQUIRE semantics, even though the root memory
+ * cgroup is never destroyed after being created and doesn't require
+ * reference counting. And it's perfectly safe to pass it to
+ * bpf_put_mem_cgroup()
+ *
+ * Return: A pointer to the root memory cgroup.
  */
 __bpf_kfunc struct mem_cgroup *bpf_get_root_mem_cgroup(void)
 {
@@ -162,7 +162,7 @@ __bpf_kfunc void bpf_mem_cgroup_flush_stats(struct mem_cgroup *memcg)
 __bpf_kfunc_end_defs();
 
 BTF_KFUNCS_START(bpf_memcontrol_kfuncs)
-BTF_ID_FLAGS(func, bpf_get_root_mem_cgroup, KF_RET_NULL)
+BTF_ID_FLAGS(func, bpf_get_root_mem_cgroup, KF_ACQUIRE | KF_RET_NULL)
 BTF_ID_FLAGS(func, bpf_get_mem_cgroup, KF_ACQUIRE | KF_RET_NULL | KF_RCU)
 BTF_ID_FLAGS(func, bpf_put_mem_cgroup, KF_RELEASE)