[PATCH net-next v3 1/2] pppoe: pass bound packets directly to generic PPP

[Qingfang Deng <qingfang.deng@linux.dev>]

Bound PPPoE sockets pass received frames to the generic PPP layer. They
currently do so through __sk_receive_skb(), which takes the socket BH
lock and serializes calls to ppp_input().

The lock originally prevented ppp_input() from racing with
ppp_unregister_channel(). Commit ec4215683e47 ("ppp: defer channel free
to an RCU grace period to fix pppol2tp RX UAF") now keeps the generic
PPP channel alive until in-flight RCU readers have completed, so bound
packets can be passed directly to ppp_input() from pppoe_rcv().

The direct bound receive path also skips the socket filter check in
__sk_receive_skb(). This does not affect normal pppd PPPoE operation:
pppd uses the kernel PPP channel and does not attach a receive filter to
a socket. Instead, it uses the generic PPP pass and active filters,
installed through PPPIOCSPASS and PPPIOCSACTIVE, unaffected by this
change.

Signed-off-by: Qingfang Deng <qingfang.deng@linux.dev>
---
A local receive-path benchmark sends 1492-byte IPv4 packets through
multiqueue TAP -> PPPoE -> PPP in a 16-vCPU x86-64 QEMU KVM guest.
Senders are pinned to distinct guest CPUs. This synchronous TAP receive
path does not use GRO. The generic PPP receive lock (ppp->rlock) is
retained in both variants. Median throughput over 12 samples per cell
across four boots, expressed as IP-layer Gbps:

  Senders       Before       After     Change
        1        19.08       19.63     +2.89%
        2        31.52       33.77     +7.15%
        4        53.67       63.48    +18.28%
        8        39.79       51.50    +29.42%
       16        49.40       61.33    +24.15%

Repeating the test with a separate benchmark-only change that removes
the generic PPP receive lock:

  Senders       Before       After     Change
        1        19.45       19.89     +2.23%
        2        33.29       34.92     +4.90%
        4        61.72       67.01     +8.57%
        8        44.46      123.76   +178.35%
       16        54.62      220.48   +303.69%

Eight and sixteen senders perform worse than four whenever at least
one receive lock remains. With both locks removed, throughput
rises from 67.01 Gbps with four senders to 123.76 Gbps with eight
and 220.48 Gbps with sixteen.

GPT Codex assisted in the benchmarks.
---
v3:
 - Update the benchmark results. Code remains unchanged.
 - Sashiko reported a potential dangling skb->dev pointer in v2:
   https://sashiko.dev/#/patchset/20261028030509.23294-1-qingfang.deng%40linux.dev?part=1
   That requires a very specific and uncommon setup. Both extensions
   require a non NULL skb->dev, so I doubt anyone had hit that path.
   PPPoL2TP also bypasses __sk_receive_skb().
v2: https://lore.kernel.org/netdev/20261028030509.23294-1-qingfang.deng@linux.dev/
 - Split out the synchronize_net() change (already in net tree e7b01bc0202a)
 - Do not change the banner comment
 - Include the benchmark results

v1: https://lore.kernel.org/netdev/20260804074452.75548-1-qingfang.deng@linux.dev/
---

 drivers/net/ppp/pppoe.c | 4 ++++
 1 file changed, 4 insertions(+)

diff --git a/drivers/net/ppp/pppoe.c b/drivers/net/ppp/pppoe.c
index 67c28ac763ef..a1ed756a6220 100644
--- a/drivers/net/ppp/pppoe.c
+++ b/drivers/net/ppp/pppoe.c
@@ -415,6 +415,10 @@ static int pppoe_rcv(struct sk_buff *skb, struct net_device *dev,
 	if (!po)
 		goto drop;
 
+	if (likely(po->sk.sk_state & PPPOX_BOUND)) {
+		ppp_input(&po->chan, skb);
+		return NET_RX_SUCCESS;
+	}
 	return __sk_receive_skb(&po->sk, skb, 0, 1, false);
 
 drop:
-- 
2.43.0