him411/lede-Heleguo
1
0
Fork 0
mirror of https://github.com/Heleguo/lede.git synced 2025-12-18 03:41:30 +00:00
Code Issues Actions 1 Packages Projects Releases Wiki Activity

kernel: fix fraglist GRO on linux 6.12

Browse Source 
Merged segments had a broken skb network header offset.

Signed-off-by: Felix Fietkau <nbd@nbd.name>
This commit is contained in:
Felix Fietkau 2025-07-05 13:51:14 +02:00
parent f2c9ca9aaa
commit f25fd50ac1
2 changed files with 36 additions and 1 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

35
target/linux/generic/pending-6.12/680-net-fix-TCP-UDP-fraglist-GRO.patch Normal file
View File

@ -0,0 +1,35 @@
From: Felix Fietkau <nbd@nbd.name>
Date: Sat, 5 Jul 2025 13:44:10 +0200
Subject: [PATCH] net: fix TCP/UDP fraglist GRO
Since "net: gro: use cb instead of skb->network_header", the skb network
header is no longer set in the GRO path.
This breaks fraglist segmentation, which relies on ip_hdr()/tcp_hdr()
to check for address/port changes.
Fix this regression by selectively setting the network header for merged
segment skbs.
Fixes: 186b1ea73ad8 ("net: gro: use cb instead of skb->network_header")
Signed-off-by: Felix Fietkau <nbd@nbd.name>
---
--- a/net/ipv4/tcp_offload.c
+++ b/net/ipv4/tcp_offload.c
@@ -355,6 +355,7 @@ struct sk_buff *tcp_gro_receive(struct l
flush |= skb->ip_summed != p->ip_summed;
flush |= skb->csum_level != p->csum_level;
flush |= NAPI_GRO_CB(p)->count >= 64;
+ skb_set_network_header(skb, skb_gro_receive_network_offset(skb));
if (flush || skb_gro_receive_list(p, skb))
mss = 1;
--- a/net/ipv4/udp_offload.c
+++ b/net/ipv4/udp_offload.c
@@ -604,6 +604,7 @@ static struct sk_buff *udp_gro_receive_s
NAPI_GRO_CB(skb)->flush = 1;
return NULL;
}
+ skb_set_network_header(skb, skb_gro_receive_network_offset(skb));
ret = skb_gro_receive_list(p, skb);
} else {
skb_gro_postpull_rcsum(skb, uh,

2
target/linux/generic/pending-6.12/690-net-add-missing-check-for-TCP-fraglist-GRO.patch
View File

@ -22,5 +22,5 @@ Signed-off-by: Felix Fietkau <nbd@nbd.name>
flush |= skb->csum_level != p->csum_level;
+ flush |= !pskb_may_pull(skb, skb_gro_offset(skb));
flush |= NAPI_GRO_CB(p)->count >= 64;
skb_set_network_header(skb, skb_gro_receive_network_offset(skb));
if (flush || skb_gro_receive_list(p, skb))