linux-kernel - Re: [PATCH] lsm_audit,selinux: add exception handling for possible NULL audit buffers

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]

Date:   Mon, 12 Jul 2021 22:49:48 -0400
From:   Paul Moore <paul@...l-moore.com>
To:     Austin Kim <austindh.kim@...il.com>
Cc:     Stephen Smalley <stephen.smalley.work@...il.com>,
        Eric Paris <eparis@...isplace.org>,
        Ondrej Mosnacek <omosnace@...hat.com>, selinux@...r.kernel.org,
        linux-kernel@...r.kernel.org, austin.kim@....com,
        kernel-team@....com
Subject: Re: [PATCH] lsm_audit,selinux: add exception handling for possible
 NULL audit buffers

On Mon, Jul 12, 2021 at 7:48 PM Austin Kim <austindh.kim@...il.com> wrote:
> From: Austin Kim <austin.kim@....com>
>
> It is possible for audit_log_start() to return NULL on error.
> So add exception handling for possible NULL audit buffers where
> return value can be handled from callers.

Hi.

The patch looks fine to me, but I think the description doesn't tell
the full story and I'm worried that people might misunderstand why
this patch is worthwhile.  I would suggest you revise the commit
description to explain that in these cases it is safe to return early
when audit_log_start() returns NULL because the only non-cleanup
processing left in these functions is to generate an audit record.
Returning early in these cases is a performance optimization and not a
correctness issue; the audit_log_*() functions can support being
passed a NULL audit_buffer argument.

> Signed-off-by: Austin Kim <austin.kim@....com>
> ---
>  security/selinux/hooks.c       | 4 ++++
>  security/selinux/ss/services.c | 2 ++
>  2 files changed, 6 insertions(+)
>
> diff --git a/security/selinux/hooks.c b/security/selinux/hooks.c
> index b0032c42333e..9e84e6635f2f 100644
> --- a/security/selinux/hooks.c
> +++ b/security/selinux/hooks.c
> @@ -3325,6 +3325,8 @@ static int selinux_inode_setxattr(struct user_namespace *mnt_userns,
>                         }
>                         ab = audit_log_start(audit_context(),
>                                              GFP_ATOMIC, AUDIT_SELINUX_ERR);
> +                       if (!ab)
> +                               return rc;
>                         audit_log_format(ab, "op=setxattr invalid_context=");
>                         audit_log_n_untrustedstring(ab, value, audit_size);
>                         audit_log_end(ab);
> @@ -6552,6 +6554,8 @@ static int selinux_setprocattr(const char *name, void *value, size_t size)
>                                 ab = audit_log_start(audit_context(),
>                                                      GFP_ATOMIC,
>                                                      AUDIT_SELINUX_ERR);
> +                               if (!ab)
> +                                       return error;
>                                 audit_log_format(ab, "op=fscreate invalid_context=");
>                                 audit_log_n_untrustedstring(ab, value, audit_size);
>                                 audit_log_end(ab);
> diff --git a/security/selinux/ss/services.c b/security/selinux/ss/services.c
> index d84c77f370dc..e5f1b2757a83 100644
> --- a/security/selinux/ss/services.c
> +++ b/security/selinux/ss/services.c
> @@ -1673,6 +1673,8 @@ static int compute_sid_handle_invalid_context(
>         if (context_struct_to_string(policydb, newcontext, &n, &nlen))
>                 goto out;
>         ab = audit_log_start(audit_context(), GFP_ATOMIC, AUDIT_SELINUX_ERR);
> +       if (!ab)
> +               goto out;
>         audit_log_format(ab,
>                          "op=security_compute_sid invalid_context=");
>         /* no need to record the NUL with untrusted strings */
> --
> 2.20.1

-- 
paul moore
www.paul-moore.com