Skip to contentSkip to conversation

: Netfilter fixes

PATCH Applied Patrick McHardy 2 participants 5 messages Dec 11, 2007 – Dec 12, 2007

Subsystems (2)
lore
Share & tools
lore
#0 lore raw

cover : Netfilter fixes

Hi Dave,

these two patches fix a missing bit on conntrack entries with master connections created through ctnetlink and some brokeness in the iptables compat code, causing it to use pointers dumped to userspace and copied back again to the kernel without any checks for validity.

Pleasy apply, thanks.

 net/ipv4/netfilter/ip_tables.c       |   57 +++++++--------------------------
 net/netfilter/nf_conntrack_netlink.c |    4 ++-
 net/netfilter/x_tables.c             |    8 +++-
 3 files changed, 21 insertions(+), 48 deletions(-)

Pablo Neira Ayuso (1):
[NETFILTER]: ctnetlink: set expected bit for related conntracks

Patrick McHardy (1):
[NETFILTER]: ip_tables: fix compat copy race

#1 lore raw

PATCH 1/2 : ctnetlink: set expected bit for related conntracks

[NETFILTER]: ctnetlink: set expected bit for related conntracks

This patch is a fix. It sets IPS_EXPECTED for related conntracks.

---
commit 4e5080caf7901c74983fcd47af4b2789bcf9fb1c
tree b240fe271290ba1a6912abc317238385187d548f
parent 82d29bf6dc7317aeb0a3a13c2348ca8591965875
author Pablo Neira Ayuso <[email protected]> Tue, 11 Dec 2007 18:24:41 +0100 committer Patrick McHardy <[email protected]> Tue, 11 Dec 2007 18:24:41 +0100

 net/netfilter/nf_conntrack_netlink.c |    4 +++-
 1 files changed, 3 insertions(+), 1 deletions(-)
net/netfilter/nf_conntrack_netlink.c +3 -1
index 9be1826..7d23124 100644
@@ -1024,8 +1024,10 @@ ctnetlink_create_conntrack(struct nlattr *cda[],
 	}
 
 	/* setup master conntrack: this is a confirmed expectation */
-	if (master_ct)
+	if (master_ct) {
+		__set_bit(IPS_EXPECTED_BIT, &ct->status);
 		ct->master = master_ct;
+	}
 
 	add_timer(&ct->timeout);
 	nf_conntrack_hash_insert(ct);
#3 lore raw

PATCH 2/2 : ip_tables: fix compat copy race

In reply to Patrick McHardy

[NETFILTER]: ip_tables: fix compat copy race

When copying entries to user, the kernel makes two passes through the data, first copying all the entries, then fixing up names and counters. On the second pass it copies the kernel and match data from userspace to the kernel again to find the corresponding structures, expecting that kernel pointers contained in the data are still valid.

This is obviously broken, fix by avoiding the second pass completely and fixing names and counters while dumping the ruleset, using the kernel-internal data structures.

---
commit 060d71fa0ce1d805708890701e55a40b1f15d4a7
tree ca0488fb9ec5b7c3a1f22dc851ab26f989641138
parent 4e5080caf7901c74983fcd47af4b2789bcf9fb1c
author Patrick McHardy <[email protected]> Tue, 11 Dec 2007 18:24:45 +0100 committer Patrick McHardy <[email protected]> Tue, 11 Dec 2007 18:24:45 +0100

 net/ipv4/netfilter/ip_tables.c |   57 ++++++++--------------------------------
 net/netfilter/x_tables.c       |    8 ++++--
 2 files changed, 18 insertions(+), 47 deletions(-)
net/ipv4/netfilter/ip_tables.c +12 -45
index 4b10b98..b9b189c 100644
@@ -1492,8 +1492,10 @@ static inline int compat_copy_match_to_user(struct ipt_entry_match *m,
 	return xt_compat_match_to_user(m, dstptr, size);
 }
 
-static int compat_copy_entry_to_user(struct ipt_entry *e,
-		void __user **dstptr, compat_uint_t *size)
+static int
+compat_copy_entry_to_user(struct ipt_entry *e, void __user **dstptr,
+			  compat_uint_t *size, struct xt_counters *counters,
+			  unsigned int *i)
 {
 	struct ipt_entry_target *t;
 	struct compat_ipt_entry __user *ce;
@@ -1507,6 +1509,9 @@ static int compat_copy_entry_to_user(struct ipt_entry *e,
 	if (copy_to_user(ce, e, sizeof(struct ipt_entry)))
 		goto out;
 
+	if (copy_to_user(&ce->counters, &counters[*i], sizeof(counters[*i])))
+		goto out;
+
 	*dstptr += sizeof(struct compat_ipt_entry);
 	ret = IPT_MATCH_ITERATE(e, compat_copy_match_to_user, dstptr, size);
 	target_offset = e->target_offset - (origsize - *size);
@@ -1522,6 +1527,8 @@ static int compat_copy_entry_to_user(struct ipt_entry *e,
 		goto out;
 	if (put_user(next_offset, &ce->next_offset))
 		goto out;
+
+	(*i)++;
 	return 0;
 out:
 	return ret;
@@ -1937,14 +1944,13 @@ struct compat_ipt_get_entries
 static int compat_copy_entries_to_user(unsigned int total_size,
 		     struct xt_table *table, void __user *userptr)
 {
-	unsigned int off, num;
-	struct compat_ipt_entry e;
 	struct xt_counters *counters;
 	struct xt_table_info *private = table->private;
 	void __user *pos;
 	unsigned int size;
 	int ret = 0;
 	void *loc_cpu_entry;
+	unsigned int i = 0;
 
 	counters = alloc_counters(table);
 	if (IS_ERR(counters))
@@ -1958,48 +1964,9 @@ static int compat_copy_entries_to_user(unsigned int total_size,
 	pos = userptr;
 	size = total_size;
 	ret = IPT_ENTRY_ITERATE(loc_cpu_entry, total_size,
-			compat_copy_entry_to_user, &pos, &size);
-	if (ret)
-		goto free_counters;
-
-	/* ... then go back and fix counters and names */
-	for (off = 0, num = 0; off < size; off += e.next_offset, num++) {
-		unsigned int i;
-		struct ipt_entry_match m;
-		struct ipt_entry_target t;
+				compat_copy_entry_to_user,
+				&pos, &size, counters, &i);
 
-		ret = -EFAULT;
-		if (copy_from_user(&e, userptr + off,
-					sizeof(struct compat_ipt_entry)))
-			goto free_counters;
-		if (copy_to_user(userptr + off +
-			offsetof(struct compat_ipt_entry, counters),
-			 &counters[num], sizeof(counters[num])))
-			goto free_counters;
-
-		for (i = sizeof(struct compat_ipt_entry);
-				i < e.target_offset; i += m.u.match_size) {
-			if (copy_from_user(&m, userptr + off + i,
-					sizeof(struct ipt_entry_match)))
-				goto free_counters;
-			if (copy_to_user(userptr + off + i +
-				offsetof(struct ipt_entry_match, u.user.name),
-				m.u.kernel.match->name,
-				strlen(m.u.kernel.match->name) + 1))
-				goto free_counters;
-		}
-
-		if (copy_from_user(&t, userptr + off + e.target_offset,
-					sizeof(struct ipt_entry_target)))
-			goto free_counters;
-		if (copy_to_user(userptr + off + e.target_offset +
-			offsetof(struct ipt_entry_target, u.user.name),
-			t.u.kernel.target->name,
-			strlen(t.u.kernel.target->name) + 1))
-			goto free_counters;
-	}
-	ret = 0;
-free_counters:
 	vfree(counters);
 	return ret;
 }
net/netfilter/x_tables.c +6 -2
index d9a3bde..b6160e4 100644
@@ -377,7 +377,9 @@ int xt_compat_match_to_user(struct xt_entry_match *m, void __user **dstptr,
 	u_int16_t msize = m->u.user.match_size - off;
 
 	if (copy_to_user(cm, m, sizeof(*cm)) ||
-	    put_user(msize, &cm->u.user.match_size))
+	    put_user(msize, &cm->u.user.match_size) ||
+	    copy_to_user(cm->u.user.name, m->u.kernel.match->name,
+			 strlen(m->u.kernel.match->name) + 1))
 		return -EFAULT;
 
 	if (match->compat_to_user) {
@@ -468,7 +470,9 @@ int xt_compat_target_to_user(struct xt_entry_target *t, void __user **dstptr,
 	u_int16_t tsize = t->u.user.target_size - off;
 
 	if (copy_to_user(ct, t, sizeof(*ct)) ||
-	    put_user(tsize, &ct->u.user.target_size))
+	    put_user(tsize, &ct->u.user.target_size) ||
+	    copy_to_user(ct->u.user.name, t->u.kernel.target->name,
+			 strlen(t->u.kernel.target->name) + 1))
 		return -EFAULT;
 
 	if (target->compat_to_user) {
maintainer
#4 lore raw

Re: [NETFILTER 02/02]: ip_tables: fix compat copy race

From: Patrick McHardy <[email protected]>
Date: Tue, 11 Dec 2007 18:42:11 +0100 (MET)

Quoted 8 lines

[NETFILTER]: ip_tables: fix compat copy race

When copying entries to user, the kernel makes two passes through the data, first copying all the entries, then fixing up names and counters. On the second pass it copies the kernel and match data from userspace to the kernel again to find the corresponding structures, expecting that kernel pointers contained in the data are still valid.

Quoted 5 lines

This is obviously broken, fix by avoiding the second pass completely and fixing names and counters while dumping the ruleset, using the kernel-internal data structures.

Also applied to net-2.6, thanks Patrick!