current/SOURCES/af_key-do-not-call-xfrm_probe_algs-in-parallel.patch

From fd4bc5d9d5c2f39b406cde4ca4dcc2e5c08a06b0 Mon Sep 17 00:00:00 2001
From: Sasha Levin <sashal@kernel.org>
Date: Thu, 4 Aug 2022 18:03:46 +0800
Subject: af_key: Do not call xfrm_probe_algs in parallel

From: Herbert Xu <herbert@gondor.apana.org.au>

[ Upstream commit ba953a9d89a00c078b85f4b190bc1dde66fe16b5 ]

When namespace support was added to xfrm/afkey, it caused the
previously single-threaded call to xfrm_probe_algs to become
multi-threaded.  This is buggy and needs to be fixed with a mutex.

Reported-by: Abhishek Shah <abhishek.shah@columbia.edu>
Fixes: 283bc9f35bbb ("xfrm: Namespacify xfrm state/policy locks")
Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Signed-off-by: Sasha Levin <sashal@kernel.org>
---
 net/key/af_key.c | 3 +++
 1 file changed, 3 insertions(+)

diff --git a/net/key/af_key.c b/net/key/af_key.c
index fb16d7c4e1b8d..20e73643b9c89 100644
--- a/net/key/af_key.c
+++ b/net/key/af_key.c
@@ -1697,9 +1697,12 @@ static int pfkey_register(struct sock *sk, struct sk_buff *skb, const struct sad
                pfk->registered |= (1<<hdr->sadb_msg_satype);
        }
 
+       mutex_lock(&pfkey_mutex);
        xfrm_probe_algs();
 
        supp_skb = compose_sadb_supported(hdr, GFP_KERNEL | __GFP_ZERO);
+       mutex_unlock(&pfkey_mutex);
+
        if (!supp_skb) {
                if (hdr->sadb_msg_satype != SADB_SATYPE_UNSPEC)
                        pfk->registered &= ~(1<<hdr->sadb_msg_satype);
-- 
2.35.1

1	From fd4bc5d9d5c2f39b406cde4ca4dcc2e5c08a06b0 Mon Sep 17 00:00:00 2001
2	From: Sasha Levin <sashal@kernel.org>
3	Date: Thu, 4 Aug 2022 18:03:46 +0800
4	Subject: af_key: Do not call xfrm_probe_algs in parallel
5
6	From: Herbert Xu <herbert@gondor.apana.org.au>
7
8	[ Upstream commit ba953a9d89a00c078b85f4b190bc1dde66fe16b5 ]
9
10	When namespace support was added to xfrm/afkey, it caused the
11	previously single-threaded call to xfrm_probe_algs to become
12	multi-threaded. This is buggy and needs to be fixed with a mutex.
13
14	Reported-by: Abhishek Shah <abhishek.shah@columbia.edu>
15	Fixes: 283bc9f35bbb ("xfrm: Namespacify xfrm state/policy locks")
16	Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au>
17	Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
18	Signed-off-by: Sasha Levin <sashal@kernel.org>
19	---
20	net/key/af_key.c \| 3 +++
21	1 file changed, 3 insertions(+)
22
23	diff --git a/net/key/af_key.c b/net/key/af_key.c
24	index fb16d7c4e1b8d..20e73643b9c89 100644
25	--- a/net/key/af_key.c
26	+++ b/net/key/af_key.c
27	@@ -1697,9 +1697,12 @@ static int pfkey_register(struct sock sk, struct sk_buff skb, const struct sad
28	pfk->registered \|= (1<<hdr->sadb_msg_satype);
29	}
30
31	+ mutex_lock(&pfkey_mutex);
32	xfrm_probe_algs();
33
34	supp_skb = compose_sadb_supported(hdr, GFP_KERNEL \| __GFP_ZERO);
35	+ mutex_unlock(&pfkey_mutex);
36	+
37	if (!supp_skb) {
38	if (hdr->sadb_msg_satype != SADB_SATYPE_UNSPEC)
39	pfk->registered &= ~(1<<hdr->sadb_msg_satype);
40	--
41	2.35.1
42