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 for Android: free password hash cracker in your pocket
[<prev] [next>] [<thread-prev] [thread-next>] [day] [month] [year] [list]
Message-ID: <53D1D353-B8F6-4ADC-8F29-8C48A7C9C6F1@kernel.org>
Date: Tue, 07 Jan 2025 15:28:54 -0800
From: Kees Cook <kees@...nel.org>
To: Gal Pressman <gal@...dia.com>, "David S. Miller" <davem@...emloft.net>,
 Jakub Kicinski <kuba@...nel.org>
CC: netdev@...r.kernel.org, Eric Dumazet <edumazet@...gle.com>,
 Paolo Abeni <pabeni@...hat.com>, David Ahern <dsahern@...nel.org>,
 Simon Horman <horms@...nel.org>, linux-hardening@...r.kernel.org,
 Cosmin Ratiu <cratiu@...dia.com>
Subject: Re: [PATCH net-next] net: Silence false field-spanning write warning in ip_tunnel_info_opts_set() memcpy



On January 7, 2025 8:55:09 AM PST, Gal Pressman <gal@...dia.com> wrote:
>When metadata_dst struct is allocated (using metadata_dst_alloc()), it
>reserves room for options at the end of the struct.
>
>Similar to [1], change the memcpy() to unsafe_memcpy() as it is
>guaranteed that enough room (md_size bytes) was allocated and the
>field-spanning write is intentional.

Why not just add an "options" flex array to struct ip_tunnel_info?

E.g.:

struct ip_tunnel_info {
	struct ip_tunnel_key	key;
	struct ip_tunnel_encap	encap;
#ifdef CONFIG_DST_CACHE
	struct dst_cache	dst_cache;
#endif
	u8			options_len;
	u8			mode;
  u8   options[] __counted_by(options_len);
};

>
>This resolves the following warning:
>  memcpy: detected field-spanning write (size 8) of single field "_Generic(info, const struct ip_tunnel_info * : ((const void *)((info) + 1)), struct ip_tunnel_info * : ((void *)((info) + 1)) )" at include/net/ip_tunnels.h:662 (size 0)

Then you can drop this macro and just use: info->options

Looks like you'd need to do it for all the types in struct metadata_dst, but at least you could stop hiding it from the compiler. :)

-Kees


-- 
Kees Cook

Powered by blists - more mailing lists

Powered by Openwall GNU/*/Linux Powered by OpenVZ