6.1-stable review patch. If anyone has any objections, please let me know.
------------------
From: Pedro Tammela pctammela@mojatatu.com
[ Upstream commit 6b78debe1c07e6aa3c91ca0b1384bf3cb8217c50 ]
Proper refcounts will always warn splat when something goes wrong, be it underflow, saturation or object resurrection. As these are always a source of bugs, use it in cls_u32 as a safeguard to prevent/catch issues. Another benefit is that the refcount API self documents the code, making clear when transitions to dead are expected.
For such an update we had to make minor adaptations on u32 to fit the refcount API. First we set explicitly to '1' when objects are created, then the objects are alive until a 1 -> 0 happens, which is then released appropriately.
The above made clear some redundant operations in the u32 code around the root_ht handling that were removed. The root_ht is created with a refcnt set to 1. Then when it's associated with tcf_proto it increments the refcnt to 2. Throughout the entire code the root_ht is an exceptional case and can never be referenced, therefore the refcnt never incremented/decremented. Its lifetime is always bound to tcf_proto, meaning if you delete tcf_proto the root_ht is deleted as well. The code made up for the fact that root_ht refcnt is 2 and did a double decrement to free it, which is not a fit for the refcount API.
Even though refcount_t is implemented using atomics, we should observe a negligible control plane impact.
Signed-off-by: Pedro Tammela pctammela@mojatatu.com Acked-by: Jamal Hadi Salim jhs@mojatatu.com Link: https://lore.kernel.org/r/20231114141856.974326-2-pctammela@mojatatu.com Signed-off-by: Jakub Kicinski kuba@kernel.org Stable-dep-of: 73af53d82076 ("net: sched: cls_u32: Fix u32's systematic failure to free IDR entries for hnodes.") Signed-off-by: Sasha Levin sashal@kernel.org --- net/sched/cls_u32.c | 36 ++++++++++++++++++------------------ 1 file changed, 18 insertions(+), 18 deletions(-)
diff --git a/net/sched/cls_u32.c b/net/sched/cls_u32.c index 04448bfb4d3db..adcc8de1d01be 100644 --- a/net/sched/cls_u32.c +++ b/net/sched/cls_u32.c @@ -70,7 +70,7 @@ struct tc_u_hnode { struct tc_u_hnode __rcu *next; u32 handle; u32 prio; - int refcnt; + refcount_t refcnt; unsigned int divisor; struct idr handle_idr; bool is_root; @@ -85,7 +85,7 @@ struct tc_u_hnode { struct tc_u_common { struct tc_u_hnode __rcu *hlist; void *ptr; - int refcnt; + refcount_t refcnt; struct idr handle_idr; struct hlist_node hnode; long knodes; @@ -357,7 +357,7 @@ static int u32_init(struct tcf_proto *tp) if (root_ht == NULL) return -ENOBUFS;
- root_ht->refcnt++; + refcount_set(&root_ht->refcnt, 1); root_ht->handle = tp_c ? gen_new_htid(tp_c, root_ht) : 0x80000000; root_ht->prio = tp->prio; root_ht->is_root = true; @@ -369,18 +369,20 @@ static int u32_init(struct tcf_proto *tp) kfree(root_ht); return -ENOBUFS; } + refcount_set(&tp_c->refcnt, 1); tp_c->ptr = key; INIT_HLIST_NODE(&tp_c->hnode); idr_init(&tp_c->handle_idr);
hlist_add_head(&tp_c->hnode, tc_u_hash(key)); + } else { + refcount_inc(&tp_c->refcnt); }
- tp_c->refcnt++; RCU_INIT_POINTER(root_ht->next, tp_c->hlist); rcu_assign_pointer(tp_c->hlist, root_ht);
- root_ht->refcnt++; + /* root_ht must be destroyed when tcf_proto is destroyed */ rcu_assign_pointer(tp->root, root_ht); tp->data = tp_c; return 0; @@ -391,7 +393,7 @@ static void __u32_destroy_key(struct tc_u_knode *n) struct tc_u_hnode *ht = rtnl_dereference(n->ht_down);
tcf_exts_destroy(&n->exts); - if (ht && --ht->refcnt == 0) + if (ht && refcount_dec_and_test(&ht->refcnt)) kfree(ht); kfree(n); } @@ -599,8 +601,6 @@ static int u32_destroy_hnode(struct tcf_proto *tp, struct tc_u_hnode *ht, struct tc_u_hnode __rcu **hn; struct tc_u_hnode *phn;
- WARN_ON(--ht->refcnt); - u32_clear_hnode(tp, ht, extack);
hn = &tp_c->hlist; @@ -628,10 +628,10 @@ static void u32_destroy(struct tcf_proto *tp, bool rtnl_held,
WARN_ON(root_ht == NULL);
- if (root_ht && --root_ht->refcnt == 1) + if (root_ht && refcount_dec_and_test(&root_ht->refcnt)) u32_destroy_hnode(tp, root_ht, extack);
- if (--tp_c->refcnt == 0) { + if (refcount_dec_and_test(&tp_c->refcnt)) { struct tc_u_hnode *ht;
hlist_del(&tp_c->hnode); @@ -643,7 +643,7 @@ static void u32_destroy(struct tcf_proto *tp, bool rtnl_held, /* u32_destroy_key() will later free ht for us, if it's * still referenced by some knode */ - if (--ht->refcnt == 0) + if (refcount_dec_and_test(&ht->refcnt)) kfree_rcu(ht, rcu); }
@@ -672,7 +672,7 @@ static int u32_delete(struct tcf_proto *tp, void *arg, bool *last, return -EINVAL; }
- if (ht->refcnt == 1) { + if (refcount_dec_if_one(&ht->refcnt)) { u32_destroy_hnode(tp, ht, extack); } else { NL_SET_ERR_MSG_MOD(extack, "Can not delete in-use filter"); @@ -680,7 +680,7 @@ static int u32_delete(struct tcf_proto *tp, void *arg, bool *last, }
out: - *last = tp_c->refcnt == 1 && tp_c->knodes == 0; + *last = refcount_read(&tp_c->refcnt) == 1 && tp_c->knodes == 0; return ret; }
@@ -764,14 +764,14 @@ static int u32_set_parms(struct net *net, struct tcf_proto *tp, NL_SET_ERR_MSG_MOD(extack, "Not linking to root node"); return -EINVAL; } - ht_down->refcnt++; + refcount_inc(&ht_down->refcnt); }
ht_old = rtnl_dereference(n->ht_down); rcu_assign_pointer(n->ht_down, ht_down);
if (ht_old) - ht_old->refcnt--; + refcount_dec(&ht_old->refcnt); }
if (ifindex >= 0) @@ -850,7 +850,7 @@ static struct tc_u_knode *u32_init_knode(struct net *net, struct tcf_proto *tp,
/* bump reference count as long as we hold pointer to structure */ if (ht) - ht->refcnt++; + refcount_inc(&ht->refcnt);
return new; } @@ -930,7 +930,7 @@ static int u32_change(struct net *net, struct sk_buff *in_skb,
ht_old = rtnl_dereference(n->ht_down); if (ht_old) - ht_old->refcnt++; + refcount_inc(&ht_old->refcnt); } __u32_destroy_key(new); return err; @@ -978,7 +978,7 @@ static int u32_change(struct net *net, struct sk_buff *in_skb, return err; } } - ht->refcnt = 1; + refcount_set(&ht->refcnt, 1); ht->divisor = divisor; ht->handle = handle; ht->prio = tp->prio;