This is a note to let you know that I've just added the patch titled
netfilter: nft_set_rbtree: fix overlap expiration walk
to the 5.4-stable tree which can be found at:
http://www.kernel.org/git/?p=linux/kernel/git/stable/stable-queue.git;a=summary
The filename of the patch is:
netfilter-nft_set_rbtree-fix-overlap-expiration-walk.patch
and it can be found in the queue-5.4 subdirectory.
If you, or anyone else, feels it should not be added to the stable tree,
please let <stable@xxxxxxxxxxxxxxx> know about it.
>From stable-owner@xxxxxxxxxxxxxxx Tue Nov 21 12:13:56 2023
From: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
Date: Tue, 21 Nov 2023 13:13:13 +0100
Subject: netfilter: nft_set_rbtree: fix overlap expiration walk
To: netfilter-devel@xxxxxxxxxxxxxxx
Cc: gregkh@xxxxxxxxxxxxxxxxxxx, sashal@xxxxxxxxxx, stable@xxxxxxxxxxxxxxx
Message-ID: <20231121121333.294238-7-pablo@xxxxxxxxxxxxx>
From: Florian Westphal <fw@xxxxxxxxx>
commit f718863aca469a109895cb855e6b81fff4827d71 upstream.
The lazy gc on insert that should remove timed-out entries fails to release
the other half of the interval, if any.
Can be reproduced with tests/shell/testcases/sets/0044interval_overlap_0
in nftables.git and kmemleak enabled kernel.
Second bug is the use of rbe_prev vs. prev pointer.
If rbe_prev() returns NULL after at least one iteration, rbe_prev points
to element that is not an end interval, hence it should not be removed.
Lastly, check the genmask of the end interval if this is active in the
current generation.
Fixes: c9e6978e2725 ("netfilter: nft_set_rbtree: Switch to node list walk for overlap detection")
Signed-off-by: Florian Westphal <fw@xxxxxxxxx>
Signed-off-by: Sasha Levin <sashal@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>
---
net/netfilter/nft_set_rbtree.c | 20 ++++++++++++++------
1 file changed, 14 insertions(+), 6 deletions(-)
--- a/net/netfilter/nft_set_rbtree.c
+++ b/net/netfilter/nft_set_rbtree.c
@@ -216,29 +216,37 @@ static void *nft_rbtree_get(const struct
static int nft_rbtree_gc_elem(const struct nft_set *__set,
struct nft_rbtree *priv,
- struct nft_rbtree_elem *rbe)
+ struct nft_rbtree_elem *rbe,
+ u8 genmask)
{
struct nft_set *set = (struct nft_set *)__set;
struct rb_node *prev = rb_prev(&rbe->node);
- struct nft_rbtree_elem *rbe_prev = NULL;
+ struct nft_rbtree_elem *rbe_prev;
struct nft_set_gc_batch *gcb;
gcb = nft_set_gc_batch_check(set, NULL, GFP_ATOMIC);
if (!gcb)
return -ENOMEM;
- /* search for expired end interval coming before this element. */
+ /* search for end interval coming before this element.
+ * end intervals don't carry a timeout extension, they
+ * are coupled with the interval start element.
+ */
while (prev) {
rbe_prev = rb_entry(prev, struct nft_rbtree_elem, node);
- if (nft_rbtree_interval_end(rbe_prev))
+ if (nft_rbtree_interval_end(rbe_prev) &&
+ nft_set_elem_active(&rbe_prev->ext, genmask))
break;
prev = rb_prev(prev);
}
- if (rbe_prev) {
+ if (prev) {
+ rbe_prev = rb_entry(prev, struct nft_rbtree_elem, node);
+
rb_erase(&rbe_prev->node, &priv->root);
atomic_dec(&set->nelems);
+ nft_set_gc_batch_add(gcb, rbe_prev);
}
rb_erase(&rbe->node, &priv->root);
@@ -320,7 +328,7 @@ static int __nft_rbtree_insert(const str
/* perform garbage collection to avoid bogus overlap reports. */
if (nft_set_elem_expired(&rbe->ext)) {
- err = nft_rbtree_gc_elem(set, priv, rbe);
+ err = nft_rbtree_gc_elem(set, priv, rbe, genmask);
if (err < 0)
return err;
Patches currently in stable-queue which might be from stable-owner@xxxxxxxxxxxxxxx are
queue-5.4/netfilter-nf_tables-fix-memleak-when-more-than-255-elements-expired.patch
queue-5.4/netfilter-nft_set_rbtree-fix-overlap-expiration-walk.patch
queue-5.4/netfilter-nf_tables-use-correct-lock-to-protect-gc_list.patch
queue-5.4/netfilter-nf_tables-disable-toggling-dormant-table-state-more-than-once.patch
queue-5.4/netfilter-nf_tables-gc-transaction-race-with-netns-dismantle.patch
queue-5.4/netfilter-nf_tables-drop-map-element-references-from-preparation-phase.patch
queue-5.4/netfilter-nf_tables-fix-gc-transaction-races-with-netns-and-netlink-event-exit-path.patch
queue-5.4/netfilter-nf_tables-don-t-skip-expired-elements-during-walk.patch
queue-5.4/netfilter-nf_tables-remove-busy-mark-and-gc-batch-api.patch
queue-5.4/netfilter-nf_tables-gc-transaction-race-with-abort-path.patch
queue-5.4/netfilter-nf_tables-unregister-flowtable-hooks-on-netns-exit.patch
queue-5.4/netfilter-nft_set_rbtree-switch-to-node-list-walk-for-overlap-detection.patch
queue-5.4/netfilter-nf_tables-adapt-set-backend-to-use-gc-transaction-api.patch
queue-5.4/netfilter-nftables-rename-set-element-data-activation-deactivation-functions.patch
queue-5.4/netfilter-nft_set_rbtree-skip-sync-gc-for-new-elements-in-this-transaction.patch
queue-5.4/netfilter-nf_tables-pass-context-to-nft_set_destroy.patch
queue-5.4/netfilter-nf_tables-bogus-ebusy-when-deleting-flowtable-after-flush-for-5.4.patch
queue-5.4/netfilter-nft_set_hash-try-later-when-gc-hits-eagain-on-iteration.patch
queue-5.4/netfilter-nf_tables-defer-gc-run-if-previous-batch-is-still-pending.patch
queue-5.4/netfilter-nft_set_rbtree-use-read-spinlock-to-avoid-datapath-contention.patch
queue-5.4/netfilter-nf_tables-double-hook-unregistration-in-netns-path.patch
queue-5.4/netfilter-nftables-update-table-flags-from-the-commit-phase.patch
queue-5.4/netfilter-nft_set_hash-mark-set-element-as-dead-when-deleting-from-packet-path.patch
queue-5.4/netfilter-nf_tables-gc-transaction-api-to-avoid-race-with-control-plane.patch
queue-5.4/netfilter-nf_tables-fix-table-flag-updates.patch
queue-5.4/netfilter-nft_set_rbtree-fix-null-deref-on-element-insertion.patch
