[PATCH] Audit: fix handling of 'strings' with NULL characters

[PATCH] Audit: fix handling of 'strings' with NULL characters

[Eric Paris]

currently audit_log_n_untrustedstring() uses
audit_string_contains_control() to check if the 'string' has any control
characters.  If the 'string' has an embedded NULL
audit_string_contains_control() will return that the data has no control
characters and will then pass the string to audit_log_n_string with the
total length, not the length up to the first NULL.  audit_log_n_string
does a memcpy of the entire length and so the actual audit record
emitted may then contain a NULL and then whatever random memory is after
the NULL.

Since we want to log the entire octet stream (if we can't trust the data
to be a string we can't trust that a NULL isn't actually a part of it)
we should just consider NULL as a control character.  If the caller is
certain they want to stop at the first NULL they should be using
audit_log_untrustedstring.

Signed-off-by: Eric Paris <eparis@redhat.com>

---

Miloslav, this is also going to take care of nulls in the TTY_AUDIT_USER
message from userspace.  Is it going to be common to have control
characters on that code path as well?  Do you want to change
audit_receive_msg() to also use the hex encoding directly instead of the
_n_untrustedstring interface?

 kernel/audit.c |    2 +-
 1 files changed, 1 insertions(+), 1 deletions(-)

diff --git a/kernel/audit.c b/kernel/audit.c
index 4414e93..ccb8d68 100644
--- a/kernel/audit.c
+++ b/kernel/audit.c
@@ -1370,7 +1370,7 @@ void audit_log_n_string(struct audit_buffer *ab, const char *string,
 int audit_string_contains_control(const char *string, size_t len)
 {
 	const unsigned char *p;
-	for (p = string; p < (const unsigned char *)string + len && *p; p++) {
+	for (p = string; p < (const unsigned char *)string + len; p++) {
 		if (*p == '"' || *p < 0x21 || *p > 0x7e)
 			return 1;
 	}

Re: [PATCH] Audit: fix handling of 'strings' with NULL characters

[Andrew Morton]

On Thu, 11 Sep 2008 17:48:39 -0400
Eric Paris <eparis@redhat.com> wrote:

> currently audit_log_n_untrustedstring() uses
> audit_string_contains_control() to check if the 'string' has any control
> characters.  If the 'string' has an embedded NULL
> audit_string_contains_control() will return that the data has no control
> characters and will then pass the string to audit_log_n_string with the
> total length, not the length up to the first NULL.  audit_log_n_string
> does a memcpy of the entire length and so the actual audit record
> emitted may then contain a NULL and then whatever random memory is after
> the NULL.
> 
> Since we want to log the entire octet stream (if we can't trust the data
> to be a string we can't trust that a NULL isn't actually a part of it)
> we should just consider NULL as a control character.  If the caller is
> certain they want to stop at the first NULL they should be using
> audit_log_untrustedstring.
> 
> Signed-off-by: Eric Paris <eparis@redhat.com>
> 
> ---
> 
> Miloslav, this is also going to take care of nulls in the TTY_AUDIT_USER
> message from userspace.  Is it going to be common to have control
> characters on that code path as well?  Do you want to change
> audit_receive_msg() to also use the hex encoding directly instead of the
> _n_untrustedstring interface?

OK, I am now officially confused about the relationship between this
patch, Miloslav's two patches and 2.6.27/2.6.26/2.6.25.

Think I'll go into hiding for a while - please wake us up when it's all
sorted out.

[PATCH] Audit: Ignore terminating NUL in AUDIT_USER_TTY messages

[Miloslav Trmač]

From: Miloslav Trmac <mitr@redhat.com>

AUDIT_USER_TTY, like all other messages sent from user-space, is sent
NUL-terminated.  Unlike other user-space audit messages, which come only
from trusted sources, AUDIT_USER_TTY messages are processed using
audit_log_n_untrustedstring().

This patch modifies AUDIT_USER_TTY handling to ignore the trailing NUL
and use the "quoted_string" representation of the message if possible.

Signed-Off-By: Miloslav Trmac <mitr@redhat.com>
---
> Miloslav, this is also going to take care of nulls in the TTY_AUDIT_USER
> message from userspace.  Is it going to be common to have control
> characters on that code path as well?
AUDIT_USER_TTY will commonly contain spaces, but not always.  This patch
cleans the AUDIT_USER_TTY messages up a bit.

 kernel/audit.c |    3 +++
 1 file changed, 3 insertions(+)
diff --git a/kernel/audit.c b/kernel/audit.c
index ccb8d68..f3d88c4 100644
--- a/kernel/audit.c
+++ b/kernel/audit.c
@@ -763,6 +763,9 @@ static int audit_receive_msg(struct sk_buff *skb, struct nlmsghdr *nlh)
 
 				audit_log_format(ab, " msg=");
 				size = nlmsg_len(nlh);
+				if (size > 0 &&
+				    ((unsigned char *)data)[size - 1] == '\0')
+					size--;
 				audit_log_n_untrustedstring(ab, data, size);
 			}
 			audit_set_pid(ab, pid);

Re: [PATCH] Audit: fix handling of 'strings' with NULL characters

[Miloslav Trmač]

Andrew Morton píše v Čt 11. 09. 2008 v 17:09 -0700:
> OK, I am now officially confused about the relationship between this
> patch, Miloslav's two patches and 2.6.27/2.6.26/2.6.25.
> 
> Think I'll go into hiding for a while - please wake us up when it's all
> sorted out.

Eric's patch ["Audit: fix handling of 'strings' with NULL characters"]
replaces my 1/2 ["audit: fix NUL handling in untrusted strings"], and
makes my 2/2 ["audit: Handle embedded NUL in TTY input auditing"]
technically unnecessary - but from the semantics POV 2/2 should still be
applied.

Either Eric's patch or my 2/2 should go to the stable releases.  The
other patches do not have to.
	Mirek