netdev.vger.kernel.org archive mirror
 help / color / mirror / Atom feed
From: "Creeley, Brett" <bcreeley@amd.com>
To: Frode Nordahl <fnordahl@ubuntu.com>, netdev@vger.kernel.org
Cc: stable@vger.kernel.org, "David S. Miller" <davem@davemloft.net>,
	David Ahern <dsahern@kernel.org>,
	Eric Dumazet <edumazet@google.com>,
	Jakub Kicinski <kuba@kernel.org>, Paolo Abeni <pabeni@redhat.com>,
	Simon Horman <horms@kernel.org>, Gal Pressman <gal@nvidia.com>,
	Kees Cook <kees@kernel.org>, Cosmin Ratiu <cratiu@nvidia.com>,
	Tariq Toukan <tariqt@nvidia.com>,
	linux-kernel@vger.kernel.org
Subject: Re: [PATCH] erspan: Initialize options_len before referencing options.
Date: Fri, 12 Dec 2025 09:23:06 -0800	[thread overview]
Message-ID: <1735c1c0-e731-4fec-83b1-818012194fc8@amd.com> (raw)
In-Reply-To: <20251212073202.13153-1-fnordahl@ubuntu.com>


On 12/11/2025 11:32 PM, Frode Nordahl wrote:
> Caution: This message originated from an External Source. Use proper caution when opening attachments, clicking links, or responding.
>
>
> The struct ip_tunnel_info has a flexible array member named
> options that is protected by a counted_by(options_len)
> attribute.
>
> The compiler will use this information to enforce runtime bounds
> checking deployed by FORTIFY_SOURCE string helpers.
>
> As laid out in the GCC documentation, the counter must be
> initialized before the first reference to the flexible array
> member.
>
> In the normal case the ip_tunnel_info_opts_set() helper is used
> which would initialize options_len properly, however in the GRE
> ERSPAN code a partial update is done, preventing the use of the
> helper function.
>
> Before this change the handling of ERSPAN traffic in GRE tunnels
> would cause a kernel panic when the kernel is compiled with
> GCC 15+ and having FORTIFY_SOURCE configured:
>
> memcpy: detected buffer overflow: 4 byte write of buffer size 0
>
> Call Trace:
>   <IRQ>
>   __fortify_panic+0xd/0xf
>   erspan_rcv.cold+0x68/0x83
>   ? ip_route_input_slow+0x816/0x9d0
>   gre_rcv+0x1b2/0x1c0
>   gre_rcv+0x8e/0x100
>   ? raw_v4_input+0x2a0/0x2b0
>   ip_protocol_deliver_rcu+0x1ea/0x210
>   ip_local_deliver_finish+0x86/0x110
>   ip_local_deliver+0x65/0x110
>   ? ip_rcv_finish_core+0xd6/0x360
>   ip_rcv+0x186/0x1a0
>
> Link: https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-counted_005fby-variable-attribute
> Reported-at: https://launchpad.net/bugs/2129580
> Fixes: bb5e62f2d547 ("net: Add options as a flexible array to struct ip_tunnel_info")

Should this be [PATCH net]?

It seems like this should be intended for the net tree.

> Signed-off-by: Frode Nordahl <fnordahl@ubuntu.com>
> ---
>   net/ipv4/ip_gre.c  | 18 ++++++++++++++++--
>   net/ipv6/ip6_gre.c | 18 ++++++++++++++++--
>   2 files changed, 32 insertions(+), 4 deletions(-)
>
> diff --git a/net/ipv4/ip_gre.c b/net/ipv4/ip_gre.c
> index 761a53c6a89a..285a656c9e41 100644
> --- a/net/ipv4/ip_gre.c
> +++ b/net/ipv4/ip_gre.c
> @@ -330,6 +330,22 @@ static int erspan_rcv(struct sk_buff *skb, struct tnl_ptk_info *tpi,
>                          if (!tun_dst)
>                                  return PACKET_REJECT;
>
> +                       /* The struct ip_tunnel_info has a flexible array member named
> +                        * options that is protected by a counted_by(options_len)
> +                        * attribute.
> +                        *
> +                        * The compiler will use this information to enforce runtime bounds
> +                        * checking deployed by FORTIFY_SOURCE string helpers.
> +                        *
> +                        * As laid out in the GCC documentation, the counter must be
> +                        * initialized before the first reference to the flexible array
> +                        * member.
> +                        *
> +                        * Link: https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-counted_005fby-variable-attribute

Nit, but I wonder if the Link in the commit message is good enough? Same 
comment below.

Thanks,

Brett

> +                        */
> +                       info = &tun_dst->u.tun_info;
> +                       info->options_len = sizeof(*md);
> +
>                          /* skb can be uncloned in __iptunnel_pull_header, so
>                           * old pkt_md is no longer valid and we need to reset
>                           * it
> @@ -344,10 +360,8 @@ static int erspan_rcv(struct sk_buff *skb, struct tnl_ptk_info *tpi,
>                          memcpy(md2, pkt_md, ver == 1 ? ERSPAN_V1_MDSIZE :
>                                                         ERSPAN_V2_MDSIZE);
>
> -                       info = &tun_dst->u.tun_info;
>                          __set_bit(IP_TUNNEL_ERSPAN_OPT_BIT,
>                                    info->key.tun_flags);
> -                       info->options_len = sizeof(*md);
>                  }
>
>                  skb_reset_mac_header(skb);
> diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c
> index c82a75510c0e..eb840a11b93b 100644
> --- a/net/ipv6/ip6_gre.c
> +++ b/net/ipv6/ip6_gre.c
> @@ -535,6 +535,22 @@ static int ip6erspan_rcv(struct sk_buff *skb,
>                          if (!tun_dst)
>                                  return PACKET_REJECT;
>
> +                       /* The struct ip_tunnel_info has a flexible array member named
> +                        * options that is protected by a counted_by(options_len)
> +                        * attribute.
> +                        *
> +                        * The compiler will use this information to enforce runtime bounds
> +                        * checking deployed by FORTIFY_SOURCE string helpers.
> +                        *
> +                        * As laid out in the GCC documentation, the counter must be
> +                        * initialized before the first reference to the flexible array
> +                        * member.
> +                        *
> +                        * Link: https://gcc.gnu.org/onlinedocs/gcc/Common-Variable-Attributes.html#index-counted_005fby-variable-attribute
> +                        */
> +                       info = &tun_dst->u.tun_info;
> +                       info->options_len = sizeof(*md);
> +
>                          /* skb can be uncloned in __iptunnel_pull_header, so
>                           * old pkt_md is no longer valid and we need to reset
>                           * it
> @@ -543,7 +559,6 @@ static int ip6erspan_rcv(struct sk_buff *skb,
>                               skb_network_header_len(skb);
>                          pkt_md = (struct erspan_metadata *)(gh + gre_hdr_len +
>                                                              sizeof(*ershdr));
> -                       info = &tun_dst->u.tun_info;
>                          md = ip_tunnel_info_opts(info);
>                          md->version = ver;
>                          md2 = &md->u.md2;
> @@ -551,7 +566,6 @@ static int ip6erspan_rcv(struct sk_buff *skb,
>                                                         ERSPAN_V2_MDSIZE);
>                          __set_bit(IP_TUNNEL_ERSPAN_OPT_BIT,
>                                    info->key.tun_flags);
> -                       info->options_len = sizeof(*md);
>
>                          ip6_tnl_rcv(tunnel, skb, tpi, tun_dst, log_ecn_error);
>
> --
> 2.43.0
>
>

  parent reply	other threads:[~2025-12-12 17:23 UTC|newest]

Thread overview: 6+ messages / expand[flat|nested]  mbox.gz  Atom feed  top
2025-12-12  7:32 [PATCH] erspan: Initialize options_len before referencing options Frode Nordahl
2025-12-12 15:13 ` Simon Horman
2025-12-12 18:21   ` Frode Nordahl
2025-12-13 16:27     ` Simon Horman
2025-12-12 17:23 ` Creeley, Brett [this message]
2025-12-12 18:24   ` Frode Nordahl

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=1735c1c0-e731-4fec-83b1-818012194fc8@amd.com \
    --to=bcreeley@amd.com \
    --cc=cratiu@nvidia.com \
    --cc=davem@davemloft.net \
    --cc=dsahern@kernel.org \
    --cc=edumazet@google.com \
    --cc=fnordahl@ubuntu.com \
    --cc=gal@nvidia.com \
    --cc=horms@kernel.org \
    --cc=kees@kernel.org \
    --cc=kuba@kernel.org \
    --cc=linux-kernel@vger.kernel.org \
    --cc=netdev@vger.kernel.org \
    --cc=pabeni@redhat.com \
    --cc=stable@vger.kernel.org \
    --cc=tariqt@nvidia.com \
    /path/to/YOUR_REPLY

  https://kernel.org/pub/software/scm/git/docs/git-send-email.html

* If your mail client supports setting the In-Reply-To header
  via mailto: links, try the mailto: link
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;
as well as URLs for NNTP newsgroup(s).