linux-kernel - Re: [PATCH net] selinux: fix potential memory leak in selinux_socket

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]

Date:   Wed, 17 Apr 2019 08:24:06 -0400
From:   Paul Moore <paul@...l-moore.com>
To:     Mao Wenan <maowenan@...wei.com>
Cc:     Stephen Smalley <sds@...ho.nsa.gov>, selinux@...r.kernel.org,
        linux-kernel@...r.kernel.org, kernel-janitors@...r.kernel.org
Subject: Re: [PATCH net] selinux: fix potential memory leak in selinux_socket_bind()

On Wed, Apr 17, 2019 at 5:15 AM Mao Wenan <maowenan@...wei.com> wrote:
>
> There might be memory leak if avc_has_perm() is failed after calling
> sel_netport_sid() or sel_netnode_sid(), port and node list must be deleted
> and freed firstly before it goto out.
> call trace:
> __sys_bind
>  security_socket_bind
>   selinux_socket_bind
>    sel_netport_sid
>    sel_netnode_sid
>
> Fixes: 3e11217263("SELinux: Add network port SID cache")
> Fixes: 88b7d370bb("selinux: fix address family in bind() and connect() to match address/port")
> Signed-off-by: Mao Wenan <maowenan@...wei.com>
> ---
>  security/selinux/hooks.c           | 15 +++++++++++----
>  security/selinux/include/netnode.h |  1 +
>  security/selinux/include/netport.h |  1 +
>  security/selinux/netnode.c         | 38 ++++++++++++++++++++++++++++++++++++++
>  security/selinux/netport.c         | 27 +++++++++++++++++++++++++++
>  5 files changed, 78 insertions(+), 4 deletions(-)

These are object label caches and as such it really isn't necessary,
or desirable, to remove entries.  Regardless of if the access is
allowed or not, the system is attempting to access these objects, and
likely to do so again, so having the object labels "hot" in the cache
is a performance win.

-- 
paul moore
www.paul-moore.com