netdev - Re: [PATCH ipsec] xfrm: avoid using skb->mac_len to decide if mac header is shown

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: <20240912060307.GA18251@breakpoint.cc>
Date: Thu, 12 Sep 2024 08:03:07 +0200
From: Florian Westphal <fw@...len.de>
To: En-Wei Wu <en-wei.wu@...onical.com>
Cc: steffen.klassert@...unet.com, herbert@...dor.apana.org.au,
	davem@...emloft.net, edumazet@...gle.com, kuba@...nel.org,
	pabeni@...hat.com, netdev@...r.kernel.org,
	linux-kernel@...r.kernel.org, kai.heng.feng@...onical.com,
	chia-lin.kao@...onical.com, anthony.wong@...onical.com,
	kuan-ying.lee@...onical.com, chris.chiu@...onical.com
Subject: Re: [PATCH ipsec] xfrm: avoid using skb->mac_len to decide if mac
 header is shown

En-Wei Wu <en-wei.wu@...onical.com> wrote:
> When we use Intel WWAN with xfrm, our system always hangs after
> browsing websites for a few seconds. The error message shows that
> it is a slab-out-of-bounds error:
> 
> The reason is that the eth_hdr(skb) inside the if statement evaluated
> to an unexpected address with skb->mac_header = ~0U (indicating there
> is no MAC header). The unreliability of skb->mac_len causes the if
> statement to become true even if there is no MAC header inside the
> skb data buffer.
> 
> Replace the skb->mac_len in the if statement with the more reliable macro
> skb_mac_header_was_set(skb) fixes this issue.
> 
> Fixes: b3284df1c86f ("xfrm: remove input2 indirection from xfrm_mode")

No, that just moved code around.

> Signed-off-by: En-Wei Wu <en-wei.wu@...onical.com>
> ---
>  net/xfrm/xfrm_input.c | 2 +-
>  1 file changed, 1 insertion(+), 1 deletion(-)
> 
> diff --git a/net/xfrm/xfrm_input.c b/net/xfrm/xfrm_input.c
> index 749e7eea99e4..93b261340105 100644
> --- a/net/xfrm/xfrm_input.c
> +++ b/net/xfrm/xfrm_input.c
> @@ -251,7 +251,7 @@ static int xfrm4_remove_tunnel_encap(struct xfrm_state *x, struct sk_buff *skb)
>  
>  	skb_reset_network_header(skb);
>  	skb_mac_header_rebuild(skb);
> -	if (skb->mac_len)
> +	if (skb_mac_header_was_set(skb))
>  		eth_hdr(skb)->h_proto = skb->protocol;

I think you will need to check both, else you restore the bug fixed in
87cdf3148b11 ("xfrm: Verify MAC header exists before overwriting
eth_hdr(skb)->h_proto").