[PATCH 4.4 72/86] netfilter: nf_tables: fix oops when inserting an element into a verdict map

From: Greg Kroah-Hartman
Date: Thu Nov 29 2018 - 09:57:47 EST

4.4-stable review patch. If anyone has any objections, please let me know.


From: Liping Zhang <zlpnobody@xxxxxxxxx>

commit 58c78e104d937c1f560fb10ed9bb2dcde0db4fcf upstream.

Dalegaard says:
The following ruleset, when loaded with 'nft -f bad.txt'
flush ruleset
table ip inlinenat {
map sourcemap {
type ipv4_addr : verdict;

chain postrouting {
ip saddr vmap @sourcemap accept
add chain inlinenat test
add element inlinenat sourcemap { : jump test }

results in a kernel oops:
BUG: unable to handle kernel paging request at 0000000000001344
IP: [<ffffffffa07bf704>] nf_tables_check_loops+0x114/0x1f0 [nf_tables]
Call Trace:
[<ffffffffa07c2aae>] ? nft_data_init+0x13e/0x1a0 [nf_tables]
[<ffffffffa07c1950>] nft_validate_register_store+0x60/0xb0 [nf_tables]
[<ffffffffa07c74b5>] nft_add_set_elem+0x545/0x5e0 [nf_tables]
[<ffffffffa07bfdd0>] ? nft_table_lookup+0x30/0x60 [nf_tables]
[<ffffffff8132c630>] ? nla_strcmp+0x40/0x50
[<ffffffffa07c766e>] nf_tables_newsetelem+0x11e/0x210 [nf_tables]
[<ffffffff8132c400>] ? nla_validate+0x60/0x80
[<ffffffffa030d9b4>] nfnetlink_rcv+0x354/0x5a7 [nfnetlink]

Because we forget to fill the net pointer in bind_ctx, so dereferencing
it may cause kernel crash.

Reported-by: Dalegaard <dalegaard@xxxxxxxxx>
Signed-off-by: Liping Zhang <zlpnobody@xxxxxxxxx>
Signed-off-by: Pablo Neira Ayuso <pablo@xxxxxxxxxxxxx>
Signed-off-by: Amit Pundir <amit.pundir@xxxxxxxxxx>
Signed-off-by: Greg Kroah-Hartman <gregkh@xxxxxxxxxxxxxxxxxxx>

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

--- a/net/netfilter/nf_tables_api.c
+++ b/net/netfilter/nf_tables_api.c
@@ -3452,6 +3452,7 @@ static int nft_add_set_elem(struct nft_c
dreg = nft_type_to_reg(set->dtype);
list_for_each_entry(binding, &set->bindings, list) {
struct nft_ctx bind_ctx = {
+ .net = ctx->net,
.afi = ctx->afi,
.table = ctx->table,
.chain = (struct nft_chain *)binding->chain,