vxlan: keep the last remote linked during FDB flush
Subsystems (3)
PATCH v3 1/1 vxlan: keep the last remote linked during FDB flush
From: Kyle Zeng <[email protected]>
A non-nexthop FDB entry is expected to have at least one remote while it remains reachable through the FDB hash table. A filtered bulk flush violates this invariant when every remote matches: It unlinks the last remote in vxlan_fdb_dst_destroy() and only afterwards tells vxlan_flush() to destroy the parent FDB entry.
An RCU reader can find the parent during this interval. first_remote_rcu() then applies list_entry_rcu() to the empty list head, producing an invalid remote pointer that the receive learning path can read from and write to.
When a matching remote is the sole remaining remote, leave it linked and ask the caller to destroy the entire FDB entry. vxlan_fdb_destroy() keeps the remote attached while sending the deletion notification and removing the parent from the lookup structures.
---
Changes in v3:
- Remove remotes_flushed and the redundant final p_destroy_fdb assignment, as suggested by Ido Schimmel; the caller initializes the flag to false.
Changes in v2:
- Add the net tree prefix to the subject.
- Restore Kyle Zeng as the patch author and correct the sign-off chain.
- Move the research credit below the commit-message separator.
- Add the recipients reported by netdev CI.
v1: https://lore.kernel.org/all/[email protected]/
Bug found and triaged by OpenAI Security Research and validated by Trail of Bits.
Trail of Bits has a reproducer for this bug that triggers a KASAN slab-out-of-bounds read in vxlan_snoop() and can share if needed.
drivers/net/vxlan/vxlan_core.c | 9 +++++---- 1 file changed, 5 insertions(+), 4 deletions(-)
drivers/net/vxlan/vxlan_core.c +5 -4
@@ -3051,18 +3051,19 @@ vxlan_fdb_flush_match_remotes(struct vxlan_fdb *f, struct vxlan_dev *vxlan,const struct vxlan_fdb_flush_desc *desc, bool *p_destroy_fdb) { - bool remotes_flushed = false; struct vxlan_rdst *rd, *tmp;list_for_each_entry_safe (rd, tmp, &f->remotes, list) {if (!vxlan_fdb_flush_remote_matches (desc, rd))continue ; +if (list_is_singular (&f->remotes)) { + *p_destroy_fdb = true; +return ; + } +vxlan_fdb_dst_destroy (vxlan, f, rd, true); - remotes_flushed = true; } - - *p_destroy_fdb = remotes_flushed &&list_empty (&f->remotes); }/* Purge the forwarding table */
Signature · 1 line
2.53.0
Quoted 13 lines · David Lee
From: Kyle Zeng <[email protected]>
A non-nexthop FDB entry is expected to have at least one remote while it remains reachable through the FDB hash table. A filtered bulk flush violates this invariant when every remote matches: It unlinks the last remote in
vxlan_fdb_dst_destroy()and only afterwards tellsvxlan_flush()to destroy the parent FDB entry.An RCU reader can find the parent during this interval.
first_remote_rcu()then applieslist_entry_rcu()to the empty list head, producing an invalid remote pointer that the receive learning path can read from and write to.
Quoted 8 lines
When a matching remote is the sole remaining remote, leave it linked and ask the caller to destroy the entire FDB entry.
vxlan_fdb_destroy()keeps the remote attached while sending the deletion notification and removing the parent from the lookup structures.Fixes c499fccb71cb ("vxlan: vxlan_core: Support FDB flushing by destination VNI")Suggested-by Ido Schimmel <[email protected]>
This tag is not really appropriate here since I didn't suggest the fix, but only a small cleanup. Anyway, not a big deal.
Quoted 4 lines
Assisted-by Codex:gpt-5.6-sol Codex:gpt-5.5-cyberSigned-off-by Kyle Zeng <[email protected]>Co-developed-by David Lee <[email protected]>Signed-off-by David Lee <[email protected]>
A message that is not in the archive (sent off-list or not yet received): CAC_etQG00pK5AT7n+BCxiRFbc_LWBkyBYSDhV-Ngnie6XBq=mQ@mail.gmail.com
Jakub Kicinski
maintainerQuoted 1 line · On Tue, 11 Aug 2026 00:52:12 +0900 David Lee wrote:
Do you want me to send v3? Or would it be fine? Thank you.
No need, I'll strip it. Just keep that in mind going forward.
This patch was applied to netdev/net.git (main) by Jakub Kicinski <[email protected]>:
Hello:
This patch was applied to netdev/net.git (main)
by Jakub Kicinski <[email protected]>:
Quoted 9 lines · On Mon, 10 Aug 2026 14:41:14 +0000 you wrote:
From: Kyle Zeng <[email protected]>
A non-nexthop FDB entry is expected to have at least one remote while it remains reachable through the FDB hash table. A filtered bulk flush violates this invariant when every remote matches: It unlinks the last remote in
vxlan_fdb_dst_destroy()and only afterwards tellsvxlan_flush()to destroy the parent FDB entry.[...]
Here is the summary with links:
- [net,v3] vxlan: keep the last remote linked during FDB flush
https://git.kernel.org/netdev/net/c/d5d4a7b538b5
You are awesome, thank you!
Signature · 2 lines
Deet-doot-dot, I am a bot. https://korg.docs.kernel.org/patchwork/pwbot.html