Re: [PATCH] SELinux: Measure state and hash of policy using IMA

On Tue, Sep 8, 2020 at 2:37 PM Stephen Smalley
[off-list ref] wrote:

On Mon, Sep 7, 2020 at 5:39 PM Lakshmi Ramasubramanian
[off-list ref] wrote:

<snip>

quoted

diff --git a/security/selinux/measure.c b/security/selinux/measure.c
new file mode 100644
index 000000000000..caf9107937d9
--- /dev/null
+++ b/security/selinux/measure.c

<snip>

quoted

+static int read_selinux_state(char **state_str, int *state_str_len,
+                             struct selinux_state *state)
+{
+       char *buf, *str_fmt = "%s=%d;";
+       int i, buf_len, curr;

<snip>

quoted

+       for (i = 0; i < __POLICYDB_CAPABILITY_MAX; i++) {
+               buf_len += snprintf(NULL, 0, str_fmt,
+                                   selinux_policycap_names[i],
+                                   state->policycap[i]);
+       }

This will need to be converted to use
security_policycap_supported(state, i) rather than state->policycap[i]
since the latter is going to be removed by Ondrej's patches I think.

Based on my testing so far, even with just moving the array under
struct selinux_policy, the RCU accessing still brings a significant
overhead (relative to the whole syscalls it is probably negligible,
but relative to the rest of the simpler hooks it is about 30%), so I
don't think it is necessary to adapt other patches to it yet. It will
be my responsibility to adapt to the newly added code when/if I rebase
and respin my patch.

quoted

+       for (i = 0; i < __POLICYDB_CAPABILITY_MAX; i++) {
+               curr += snprintf((buf + curr), (buf_len - curr), str_fmt,
+                                selinux_policycap_names[i],
+                                state->policycap[i]);

Ditto.

-- 
Ondrej Mosnacek
Software Engineer, Platform Security - SELinux kernel
Red Hat, Inc.