From: Florian Westphal <fw@strlen•de>
To: <netdev@vger•kernel.org>
Cc: syzkaller-bugs@googlegroups•com, steffen.klassert@secunet•com,
herbert@gondor•apana.org.au, Florian Westphal <fw@strlen•de>
Subject: [PATCH net] xfrm: defer daddr pointer assignment after spi parsing
Date: Wed, 1 Nov 2017 20:30:49 +0100 [thread overview]
Message-ID: <20171101193049.12336-1-fw@strlen.de> (raw)
In-Reply-To: <94eb2c05815016ca9c055cef9236@google.com>
syzbot reports:
BUG: KASAN: use-after-free in __xfrm_state_lookup+0x695/0x6b0
Read of size 4 at addr ffff8801d434e538 by task syzkaller647520/2991
[..]
__xfrm_state_lookup+0x695/0x6b0 net/xfrm/xfrm_state.c:833
xfrm_state_lookup+0x8a/0x160 net/xfrm/xfrm_state.c:1592
xfrm_input+0x8e5/0x22f0 net/xfrm/xfrm_input.c:302
The use-after-free is the ipv4 destination address, which points
to an skb head area that has been reallocated:
pskb_expand_head+0x36b/0x1210 net/core/skbuff.c:1494
__pskb_pull_tail+0x14a/0x17c0 net/core/skbuff.c:1877
pskb_may_pull include/linux/skbuff.h:2102 [inline]
xfrm_parse_spi+0x3d3/0x4d0 net/xfrm/xfrm_input.c:170
xfrm_input+0xce2/0x22f0 net/xfrm/xfrm_input.c:291
so the real bug is that xfrm_parse_spi() uses pskb_may_pull, but
for now do smaller workaround that makes xfrm_input fetch daddr
after spi parsing.
Reported-by: syzbot <syzkaller@googlegroups•com>
Signed-off-by: Florian Westphal <fw@strlen•de>
---
net/xfrm/xfrm_input.c | 4 ++--
1 file changed, 2 insertions(+), 2 deletions(-)
diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
index 8ac9d32fb79d..1c6051cb7733 100644
--- a/net/xfrm/xfrm_input.c
+++ b/net/xfrm/xfrm_input.c
@@ -265,8 +265,6 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
goto lock;
}
- daddr = (xfrm_address_t *)(skb_network_header(skb) +
- XFRM_SPI_SKB_CB(skb)->daddroff);
family = XFRM_SPI_SKB_CB(skb)->family;
/* if tunnel is present override skb->mark value with tunnel i_key */
@@ -293,6 +291,8 @@ int xfrm_input(struct sk_buff *skb, int nexthdr, __be32 spi, int encap_type)
goto drop;
}
+ daddr = (xfrm_address_t *)(skb_network_header(skb) +
+ XFRM_SPI_SKB_CB(skb)->daddroff);
do {
if (skb->sp->len == XFRM_MAX_DEPTH) {
XFRM_INC_STATS(net, LINUX_MIB_XFRMINBUFFERERROR);
--
2.13.6
next prev parent reply other threads:[~2017-11-01 19:30 UTC|newest]
Thread overview: 4+ messages / expand[flat|nested] mbox.gz Atom feed top
2017-11-01 17:55 KASAN: use-after-free Read in __xfrm_state_lookup syzbot
2017-11-01 19:30 ` Florian Westphal [this message]
2017-11-03 7:16 ` [PATCH net] xfrm: defer daddr pointer assignment after spi parsing Steffen Klassert
2018-01-31 2:05 ` KASAN: use-after-free Read in __xfrm_state_lookup Eric Biggers
Reply instructions:
You may reply publicly to this message via plain-text email
using any one of the following methods:
* Save the following mbox file, import it into your mail client,
and reply-to-all from there: mbox
Avoid top-posting and favor interleaved quoting:
https://en.wikipedia.org/wiki/Posting_style#Interleaved_style
* Reply using the --to, --cc, and --in-reply-to
switches of git-send-email(1):
git send-email \
--in-reply-to=20171101193049.12336-1-fw@strlen.de \
--to=fw@strlen$(echo .)de \
--cc=herbert@gondor$(echo .)apana.org.au \
--cc=netdev@vger$(echo .)kernel.org \
--cc=steffen.klassert@secunet$(echo .)com \
--cc=syzkaller-bugs@googlegroups$(echo .)com \
/path/to/YOUR_REPLY
https://kernel.org/pub/software/scm/git/docs/git-send-email.html
Be sure your reply has a Subject: header at the top and a blank line
before the message body.
This is a public inbox, see mirroring instructions
for how to clone and mirror all data and code used for this inbox