[PATCH AUTOSEL 6.19-5.10] netfilter: nf_conntrack: Add allow_clash to generic protocol handler

[Sasha Levin <sashal@kernel.org>]

From: Yuto Hamaguchi <Hamaguchi.Yuto@da.MitsubishiElectric.co.jp>

[ Upstream commit 8a49fc8d8a3e83dc51ec05bcd4007bdea3c56eec ]

The upstream commit, 71d8c47fc653711c41bc3282e5b0e605b3727956
 ("netfilter: conntrack: introduce clash resolution on insertion race"),
sets allow_clash=true in the UDP/UDPLITE protocol handler
but does not set it in the generic protocol handler.

As a result, packets composed of connectionless protocols at each layer,
such as UDP over IP-in-IP, still drop packets due to conflicts during conntrack insertion.

To resolve this, this patch sets allow_clash in the nf_conntrack_l4proto_generic.

Signed-off-by: Yuto Hamaguchi <Hamaguchi.Yuto@da.MitsubishiElectric.co.jp>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---

LLM Generated explanations, may be completely bogus:

The current file does **not** yet have `allow_clash = true` — this
confirms the patch hasn't been applied yet and is needed.

### 7. STABILITY AND TRUST INDICATORS

- **Signed-off by Florian Westphal** — the netfilter maintainer,
  indicating proper review
- The fix is trivially verifiable: it mirrors what was already done for
  UDP/UDPLITE/GRE
- The author (Yuto Hamaguchi from Mitsubishi Electric) clearly
  encountered this in a real-world deployment scenario

### Summary

| Criterion | Assessment |
|-----------|-----------|
| **Fixes a real bug** | YES — packets dropped for connectionless
protocol stacks |
| **Obviously correct** | YES — mirrors existing pattern for
UDP/UDPLITE/GRE |
| **Small and contained** | YES — single line addition |
| **No new features** | Correct — just extends existing clash resolution
to generic handler |
| **Risk of regression** | Very low — same mechanism already proven for
UDP |
| **User impact** | Moderate-high — affects any setup with
connectionless tunnels + conntrack |
| **Dependencies** | Requires 71d8c47fc653 which is already in mainline
|

This is a textbook stable backport candidate: a one-line fix that
addresses a real user-visible bug (packet drops) with essentially zero
regression risk, fixing an oversight in an earlier commit that only
partially applied clash resolution to all relevant protocol handlers.

**YES**

 net/netfilter/nf_conntrack_proto_generic.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/netfilter/nf_conntrack_proto_generic.c b/net/netfilter/nf_conntrack_proto_generic.c
index e831637bc8ca8..cb260eb3d012c 100644
--- a/net/netfilter/nf_conntrack_proto_generic.c
+++ b/net/netfilter/nf_conntrack_proto_generic.c
@@ -67,6 +67,7 @@ void nf_conntrack_generic_init_net(struct net *net)
 const struct nf_conntrack_l4proto nf_conntrack_l4proto_generic =
 {
 	.l4proto		= 255,
+	.allow_clash            = true,
 #ifdef CONFIG_NF_CONNTRACK_TIMEOUT
 	.ctnl_timeout		= {
 		.nlattr_to_obj	= generic_timeout_nlattr_to_obj,
-- 
2.51.0