netdev - Re: [PATCH] ppp: Prevent out-of-bounds access in ppp_sync

lists.openwall.net		lists / announce owl-users owl-dev john-users john-dev passwdqc-users yescrypt popa3d-users / oss-security kernel-hardening musl sabotage tlsify passwords / crypt-dev xvendor / Bugtraq Full-Disclosure linux-kernel linux-netdev linux-ext4 linux-hardening linux-cve-announce PHC
Open Source and information security mailing list archives

Hash Suite: Windows password security audit tool. GUI, reports in PDF.

[<prev] [next>] [<thread-prev] [day] [month] [year] [list]

Message-ID: <20250218185033.26399-1-kuniyu@amazon.com>
Date: Tue, 18 Feb 2025 10:50:33 -0800
From: Kuniyuki Iwashima <kuniyu@...zon.com>
To: <purvayeshi550@...il.com>
CC: <andrew+netdev@...n.ch>, <davem@...emloft.net>, <edumazet@...gle.com>,
	<kuba@...nel.org>, <kuniyu@...zon.com>, <linux-kernel@...r.kernel.org>,
	<linux-ppp@...r.kernel.org>, <netdev@...r.kernel.org>, <pabeni@...hat.com>,
	<skhan@...uxfoundation.org>,
	<syzbot+29fc8991b0ecb186cf40@...kaller.appspotmail.com>
Subject: Re: [PATCH] ppp: Prevent out-of-bounds access in ppp_sync_txmunge

From: Purva Yeshi <purvayeshi550@...il.com>
Date: Tue, 18 Feb 2025 11:58:17 +0530
> On 18/02/25 02:46, Kuniyuki Iwashima wrote:
> > From: Purva Yeshi <purvayeshi550@...il.com>
> > Date: Sun, 16 Feb 2025 11:34:46 +0530
> >> Fix an issue detected by syzbot with KMSAN:
> >>
> >> BUG: KMSAN: uninit-value in ppp_sync_txmunge
> >> drivers/net/ppp/ppp_synctty.c:516 [inline]
> >> BUG: KMSAN: uninit-value in ppp_sync_send+0x21c/0xb00
> >> drivers/net/ppp/ppp_synctty.c:568
> >>
> >> Ensure sk_buff is valid and has at least 3 bytes before accessing its
> >> data field in ppp_sync_txmunge(). Without this check, the function may
> >> attempt to read uninitialized or invalid memory, leading to undefined
> >> behavior.
> >>
> >> To address this, add a validation check at the beginning of the function
> >> to safely handle cases where skb is NULL or too small. If either condition
> >> is met, free the skb and return NULL to prevent processing an invalid
> >> packet.
> >>
> >> Reported-by: syzbot+29fc8991b0ecb186cf40@...kaller.appspotmail.com
> >> Closes: https://syzkaller.appspot.com/bug?extid=29fc8991b0ecb186cf40
> >> Tested-by: syzbot+29fc8991b0ecb186cf40@...kaller.appspotmail.com
> >> Signed-off-by: Purva Yeshi <purvayeshi550@...il.com>
> >> ---
> >>   drivers/net/ppp/ppp_synctty.c | 6 ++++++
> >>   1 file changed, 6 insertions(+)
> >>
> >> diff --git a/drivers/net/ppp/ppp_synctty.c b/drivers/net/ppp/ppp_synctty.c
> >> index 644e99fc3..e537ea3d9 100644
> >> --- a/drivers/net/ppp/ppp_synctty.c
> >> +++ b/drivers/net/ppp/ppp_synctty.c
> >> @@ -506,6 +506,12 @@ ppp_sync_txmunge(struct syncppp *ap, struct sk_buff *skb)
> >>   	unsigned char *data;
> >>   	int islcp;
> >>   
> >> +	/* Ensure skb is not NULL and has at least 3 bytes */
> >> +	if (!skb || skb->len < 3) {
> > 
> > When is skb NULL ?
> 
> skb pointer can be NULL in cases where memory allocation for the socket 
> buffer fails, or if an upstream function incorrectly passes a NULL 
> reference due to improper error handling.

Which caller passes NULL skb ?

If it's really possible, you'll see null-ptr-deref at

  data = skb->data;

below instead of KMSAN's uninit splat.


> 
> Additionally, skb->len being less than 3 can occur if the received 
> packet is truncated or malformed, leading to out-of-bounds memory access 
> when attempting to read data[2].
> 
> > 
> > 
> >> +		kfree_skb(skb);
> >> +		return NULL;
> >> +	}
> >> +
> >>   	data  = skb->data;
> >>   	proto = get_unaligned_be16(data);
> >>   
> >> -- 
> >> 2.34.1