2017-09-19 17:19:24

by Meng Xu

Subject: [PATCH] net: compat: assert the size of cmsg copied in is as expected

The actual length of cmsg fetched in during the second loop
(i.e., kcmsg - kcmsg_base) could be different from what we
get from the first loop (i.e., kcmlen).

The main reason is that the two get_user() calls in the two
loops (i.e., get_user(ucmlen, &ucmsg->cmsg_len) and
__get_user(ucmlen, &ucmsg->cmsg_len)) could cause ucmlen
to have different values even they fetch from the same userspace
address, as user can race to change the memory content in
&ucmsg->cmsg_len across fetches.

Although in the second loop, the sanity check
if ((char *)kcmsg_base + kcmlen - (char *)kcmsg < CMSG_ALIGN(tmp))
is inplace, it only ensures that the cmsg fetched in during the
second loop does not exceed the length of kcmlen, but not
necessarily equal to kcmlen. But indicated by the assignment
kmsg->msg_controllen = kcmlen, we should enforce that.

This patch adds this additional sanity check and ensures that
what is recorded in kmsg->msg_controllen is the actual cmsg length.

Signed-off-by: Meng Xu <[email protected]>
---
net/compat.c | 7 +++++++
1 file changed, 7 insertions(+)

diff --git a/net/compat.c b/net/compat.c
index 6ded6c8..2238171 100644
--- a/net/compat.c
+++ b/net/compat.c
@@ -185,6 +185,13 @@ int cmsghdr_from_user_compat_to_kern(struct msghdr *kmsg, struct sock *sk,
ucmsg = cmsg_compat_nxthdr(kmsg, ucmsg, ucmlen);
}

+ /*
+ * check the length of messages copied in is the same as the
+ * what we get from the first loop
+ */
+ if ((char *)kcmsg - (char *)kcmsg_base != kcmlen)
+ goto Einval;
+
/* Ok, looks like we made it. Hook it up and return success. */
kmsg->msg_control = kcmsg_base;
kmsg->msg_controllen = kcmlen;
--
2.7.4

2017-09-20 22:36:52

by David Miller

Subject: Re: [PATCH] net: compat: assert the size of cmsg copied in is as expected

From: Meng Xu <[email protected]>
Date: Tue, 19 Sep 2017 13:19:13 -0400

> The actual length of cmsg fetched in during the second loop
> (i.e., kcmsg - kcmsg_base) could be different from what we
> get from the first loop (i.e., kcmlen).
>
> The main reason is that the two get_user() calls in the two
> loops (i.e., get_user(ucmlen, &ucmsg->cmsg_len) and
> __get_user(ucmlen, &ucmsg->cmsg_len)) could cause ucmlen
> to have different values even they fetch from the same userspace
> address, as user can race to change the memory content in
> &ucmsg->cmsg_len across fetches.
>
> Although in the second loop, the sanity check
> if ((char *)kcmsg_base + kcmlen - (char *)kcmsg < CMSG_ALIGN(tmp))
> is inplace, it only ensures that the cmsg fetched in during the
> second loop does not exceed the length of kcmlen, but not
> necessarily equal to kcmlen. But indicated by the assignment
> kmsg->msg_controllen = kcmlen, we should enforce that.
>
> This patch adds this additional sanity check and ensures that
> what is recorded in kmsg->msg_controllen is the actual cmsg length.
>
> Signed-off-by: Meng Xu <[email protected]>

Applied, thanks.