2022-12-08 14:04:19

by Qingfang Deng

Subject: [PATCH net] netfilter: flowtable: really fix NAT IPv6 offload

The for-loop was broken from the start. It translates to:

for (i = 0; i < 4; i += 4)

which means the loop statement is run only once, so only the highest
32-bit of the IPv6 address gets mangled.

Fix the loop increment.

Fixes: 0e07e25b481a ("netfilter: flowtable: fix NAT IPv6 offload mangling")
Fixes: 5c27d8d76ce8 ("netfilter: nf_flow_table_offload: add IPv6 support")
Signed-off-by: Qingfang DENG <[email protected]>
---
net/netfilter/nf_flow_table_offload.c | 6 +++---
1 file changed, 3 insertions(+), 3 deletions(-)

diff --git a/net/netfilter/nf_flow_table_offload.c b/net/netfilter/nf_flow_table_offload.c
index 00b522890d77..921eddf96dc4 100644
--- a/net/netfilter/nf_flow_table_offload.c
+++ b/net/netfilter/nf_flow_table_offload.c
@@ -383,12 +383,12 @@ static void flow_offload_ipv6_mangle(struct nf_flow_rule *flow_rule,
const __be32 *addr, const __be32 *mask)
{
struct flow_action_entry *entry;
- int i, j;
+ int i;

- for (i = 0, j = 0; i < sizeof(struct in6_addr) / sizeof(u32); i += sizeof(u32), j++) {
+ for (i = 0; i < sizeof(struct in6_addr) / sizeof(u32); i++) {
entry = flow_action_entry_next(flow_rule);
flow_offload_mangle(entry, FLOW_ACT_MANGLE_HDR_TYPE_IP6,
- offset + i, &addr[j], mask);
+ offset + i * sizeof(u32), &addr[i], mask);
}
}

--
2.34.1

2022-12-08 21:29:28

by Pablo Neira Ayuso

Subject: Re: [PATCH net] netfilter: flowtable: really fix NAT IPv6 offload

On Thu, Dec 08, 2022 at 08:35:29PM +0800, Qingfang DENG wrote:
> The for-loop was broken from the start. It translates to:
>
> for (i = 0; i < 4; i += 4)
>
> which means the loop statement is run only once, so only the highest
> 32-bit of the IPv6 address gets mangled.
>
> Fix the loop increment.

Applied, thanks