On 02/15/2017 09:24 PM, Andy Lutomirski wrote:
On Mon, Feb 13, 2017 at 7:45 PM, Tyler Hicks
<tyhicks(a)canonical.com> wrote:
> This patch set is the third revision of the following two previously
> submitted patch sets:
>
> v1:
http://lkml.kernel.org/r/1483375990-14948-1-git-send-email-tyhicks@canoni...
> v1:
http://lkml.kernel.org/r/1483377999-15019-2-git-send-email-tyhicks@canoni...
>
> v2:
http://lkml.kernel.org/r/1486100262-32391-1-git-send-email-tyhicks@canoni...
>
> The patch set aims to address some known deficiencies in seccomp's current
> logging capabilities:
>
> 1. Inability to log all filter actions.
> 2. Inability to selectively enable filtering; e.g. devs want noisy logging,
> users want relative quiet.
> 3. Consistent behavior with audit enabled and disabled.
> 4. Inability to easily develop a filter due to the lack of a
> permissive/complain mode.
I think I dislike this, but I think my dislikes may be fixable with
minor changes.
What I dislike is that this mixes app-specific built-in configuration
(seccomp) with global privileged stuff (audit). The result is a
potentially difficult to use situation in which you need to modify an
app to make it loggable (using RET_LOG) and then fiddle with
privileged config (auditctl, etc) to actually see the logs.
There's no need to fiddle with audictl. You would only need to write
"log" to /proc/sys/kernel/seccomp/log_max_action. I wouldn't think that
would be an issue considering that this will typically be be done on a
developer's system. Additionally, I plan to make "log" the default for
log_max_action in Ubuntu so there'd be nothing to change to get RET_LOG
working.
If auditd is running, the messages will go to the audit.log. If auditd
is not running, the messages will go to the syslog.
What if we make "log" the default for log_max_action in this patch set?
Would that address your concerns?
What if, instead of logging straight to the audit log,
SECCOMP_RET_LOG
[1] merely meant "tell our parent about this syscall"? (Ideally we'd
also figure out a way to express "log this and allow", "log this and
do ERRNO", etc.) Then we could have another mechanism that installs a
layer in the seccomp stack that, instead of catching syscalls, catches
log events and sticks them in a ring buffer (or audit).
Concretely, it might work like this. If a filter returns
SECCOMP_RET_LOG, then we "log" and keep processing. SECCOMP_RET_LOG
is otherwise treated literally like SECCOMP_RET_ALLOW and has no
effect on return value. If you want log-and-kill, you install two
filters.
There's a new seccomp(2) action that returns an fd. That fd
references a new thing in the seccomp stack that is a BPF program that
is called whenever SECCOMP_RET_LOG is returned from lower down. The
output of this filter determines whether the log event is ignored,
stuck in the ring buffer, or passed up the stack for further
processing. You read(2) the fd to access the ring buffer.
Using this mechanism, you could write a simple seccomptrace tool that
needs no privilege and dumps SECCOMP_RET_LOG events from the target
program to stderr.
Thoughts?
IMHO, this sounds like yet another logging daemon. It is also more
complex than what's needed for my use case of SECCOMP_RET_LOG. I only
need a seccomp learning mode, to accompany the learning modes
implemented in the various LSMs, that allows for everything marked with
RET_LOG to hit the syslog (or audit log).
Having to wedge seccomptrace tool between my application launcher, which
handles sandboxing, and the application being launched is less than
ideal. The application launcher could reimplement what seccomptrace is
doing but it'd be nice if that was left up to auditd/rsyslogd/journald/etc.
Tyler
[1] If we went this route, it might want to be renamed.
P.S. We ought to be able to write a BPF verifier pass that makes sure
that filters don't return unsupported return values if we cared to do
so.