Skip to contentSkip to conversation

Netfilter fixes for net

PATCH In review Pablo Neira Ayuso 2 participants 6 messages Jun 14, 2020 – Jun 15, 2020

Subsystems (3)
lore
Share & tools
lore

Pablo Neira Ayuso

author maintainer
#0 lore raw

cover Netfilter fixes for net

Hi,

The following patchset contains Netfilter fixes for net:

1) Fix bogus EEXIST on element insertions to the rbtree with timeouts, from Stefano Brivio.

2) Preempt BUG splat in the pipapo element insertion path, also from Stefano.

3) Release filter from the ctnetlink error path.

4) Release flowtable hooks from the deletion path.

Please, pull these changes from:

git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git

Thank you.

----------------------------------------------------------------

The following changes since commit af7b4801030c07637840191c69eb666917e4135d:

Merge git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net (2020-06-07 17:27:45 -0700)

are available in the Git repository at:

git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git HEAD

for you to fetch changes up to 3003055f50663095472144994dac0339076031a8:

netfilter: nf_tables: hook list memleak in flowtable deletion (2020-06-12 17:48:21 +0200)

----------------------------------------------------------------
Pablo Neira Ayuso (2):
netfilter: ctnetlink: memleak in filter initialization error path
netfilter: nf_tables: hook list memleak in flowtable deletion

Stefano Brivio (2):
netfilter: nft_set_rbtree: Don't account for expired elements on insertion
netfilter: nft_set_pipapo: Disable preemption before getting per-CPU pointer

 net/netfilter/nf_conntrack_netlink.c | 32 ++++++++++++++++++++++----------
 net/netfilter/nf_tables_api.c        | 31 ++++++++++++++++++++++++-------
 net/netfilter/nft_set_pipapo.c       |  6 +++++-
 net/netfilter/nft_set_rbtree.c       | 21 ++++++++++++++-------
 4 files changed, 65 insertions(+), 25 deletions(-)
author maintainer
#1 lore raw

PATCH 1/4 netfilter: nft_set_rbtree: Don't account for expired elements on insertion

From: Stefano Brivio <[email protected]>

While checking the validity of insertion in __nft_rbtree_insert(), we currently ignore conflicting elements and intervals only if they are not active within the next generation.

However, if we consider expired elements and intervals as potentially conflicting and overlapping, we'll return error for entries that should be added instead. This is particularly visible with garbage collection intervals that are comparable with the element timeout itself, as reported by Mike Dillinger.

Other than the simple issue of denying insertion of valid entries, this might also result in insertion of a single element (opening or closing) out of a given interval. With single entries (that are inserted as intervals of size 1), this leads in turn to the creation of new intervals. For example:

  # nft add element t s { 192.0.2.1 }
  # nft list ruleset
  [...]
     elements = { 192.0.2.1-255.255.255.255 }

Always ignore expired elements active in the next generation, while checking for conflicts.

It might be more convenient to introduce a new macro that covers both inactive and expired items, as this type of check also appears quite frequently in other set back-ends. This is however beyond the scope of this fix and can be deferred to a separate patch.

Other than the overlap detection cases introduced by commit 7c84d41416d8 ("netfilter: nft_set_rbtree: Detect partial overlaps on insertion"), we also have to cover the original conflict check dealing with conflicts between two intervals of size 1, which was introduced before support for timeout was introduced. This won't return an error to the user as -EEXIST is masked by nft if NLM_F_EXCL is not given, but would result in a silent failure adding the entry.

Cc <[email protected]> # 5.6.x
Fixes 8d8540c4f5e0 ("netfilter: nft_set_rbtree: add timeout support")
Fixes 7c84d41416d8 ("netfilter: nft_set_rbtree: Detect partial overlaps on insertion")
---
 net/netfilter/nft_set_rbtree.c | 21 ++++++++++++++-------
 1 file changed, 14 insertions(+), 7 deletions(-)
net/netfilter/nft_set_rbtree.c +14 -7
index 62f416bc0579..b6aad3fc46c3 100644
@@ -271,12 +271,14 @@ static int __nft_rbtree_insert(const struct net *net, const struct nft_set *set,
 
 			if (nft_rbtree_interval_start(new)) {
 				if (nft_rbtree_interval_end(rbe) &&
-				    nft_set_elem_active(&rbe->ext, genmask))
+				    nft_set_elem_active(&rbe->ext, genmask) &&
+				    !nft_set_elem_expired(&rbe->ext))
 					overlap = false;
 			} else {
 				overlap = nft_rbtree_interval_end(rbe) &&
 					  nft_set_elem_active(&rbe->ext,
-							      genmask);
+							      genmask) &&
+					  !nft_set_elem_expired(&rbe->ext);
 			}
 		} else if (d > 0) {
 			p = &parent->rb_right;
@@ -284,9 +286,11 @@ static int __nft_rbtree_insert(const struct net *net, const struct nft_set *set,
 			if (nft_rbtree_interval_end(new)) {
 				overlap = nft_rbtree_interval_end(rbe) &&
 					  nft_set_elem_active(&rbe->ext,
-							      genmask);
+							      genmask) &&
+					  !nft_set_elem_expired(&rbe->ext);
 			} else if (nft_rbtree_interval_end(rbe) &&
-				   nft_set_elem_active(&rbe->ext, genmask)) {
+				   nft_set_elem_active(&rbe->ext, genmask) &&
+				   !nft_set_elem_expired(&rbe->ext)) {
 				overlap = true;
 			}
 		} else {
@@ -294,15 +298,18 @@ static int __nft_rbtree_insert(const struct net *net, const struct nft_set *set,
 			    nft_rbtree_interval_start(new)) {
 				p = &parent->rb_left;
 
-				if (nft_set_elem_active(&rbe->ext, genmask))
+				if (nft_set_elem_active(&rbe->ext, genmask) &&
+				    !nft_set_elem_expired(&rbe->ext))
 					overlap = false;
 			} else if (nft_rbtree_interval_start(rbe) &&
 				   nft_rbtree_interval_end(new)) {
 				p = &parent->rb_right;
 
-				if (nft_set_elem_active(&rbe->ext, genmask))
+				if (nft_set_elem_active(&rbe->ext, genmask) &&
+				    !nft_set_elem_expired(&rbe->ext))
 					overlap = false;
-			} else if (nft_set_elem_active(&rbe->ext, genmask)) {
+			} else if (nft_set_elem_active(&rbe->ext, genmask) &&
+				   !nft_set_elem_expired(&rbe->ext)) {
 				*ext = &rbe->ext;
 				return -EEXIST;
 			} else {
Signature · 1 line
2.20.1
author maintainer
#2 lore raw

PATCH 2/4 netfilter: nft_set_pipapo: Disable preemption before getting per-CPU pointer

In reply to Pablo Neira Ayuso

From: Stefano Brivio <[email protected]>

The lkp kernel test robot reports, with CONFIG_DEBUG_PREEMPT enabled:

  [  165.316525] BUG: using smp_processor_id() in preemptible [00000000] code: nft/6247
  [  165.319547] caller is nft_pipapo_insert+0x464/0x610 [nf_tables]
  [  165.321846] CPU: 1 PID: 6247 Comm: nft Not tainted 5.6.0-rc5-01595-ge32a4dc6512ce3 #1
  [  165.332128] Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS 1.12.0-1 04/01/2014
  [  165.334892] Call Trace:
  [  165.336435]  dump_stack+0x8f/0xcb
  [  165.338128]  debug_smp_processor_id+0xb2/0xc0
  [  165.340117]  nft_pipapo_insert+0x464/0x610 [nf_tables]
  [  165.342290]  ? nft_trans_alloc_gfp+0x1c/0x60 [nf_tables]
  [  165.344420]  ? rcu_read_lock_sched_held+0x52/0x80
  [  165.346460]  ? nft_trans_alloc_gfp+0x1c/0x60 [nf_tables]
  [  165.348543]  ? __mmu_interval_notifier_insert+0xa0/0xf0
  [  165.350629]  nft_add_set_elem+0x5ff/0xa90 [nf_tables]
  [  165.352699]  ? __lock_acquire+0x241/0x1400
  [  165.354573]  ? __lock_acquire+0x241/0x1400
  [  165.356399]  ? reacquire_held_locks+0x12f/0x200
  [  165.358384]  ? nf_tables_valid_genid+0x1f/0x40 [nf_tables]
  [  165.360502]  ? nla_strcmp+0x10/0x50
  [  165.362199]  ? nft_table_lookup+0x4f/0xa0 [nf_tables]
  [  165.364217]  ? nla_strcmp+0x10/0x50
  [  165.365891]  ? nf_tables_newsetelem+0xd5/0x150 [nf_tables]
  [  165.367997]  nf_tables_newsetelem+0xd5/0x150 [nf_tables]
  [  165.370083]  nfnetlink_rcv_batch+0x4fd/0x790 [nfnetlink]
  [  165.372205]  ? __lock_acquire+0x241/0x1400
  [  165.374058]  ? __nla_validate_parse+0x57/0x8a0
  [  165.375989]  ? cap_inode_getsecurity+0x230/0x230
  [  165.377954]  ? security_capable+0x38/0x50
  [  165.379795]  nfnetlink_rcv+0x11d/0x140 [nfnetlink]
  [  165.381779]  netlink_unicast+0x1b2/0x280
  [  165.383612]  netlink_sendmsg+0x351/0x470
  [  165.385439]  sock_sendmsg+0x5b/0x60
  [  165.387133]  ____sys_sendmsg+0x200/0x280
  [  165.388871]  ? copy_msghdr_from_user+0xd9/0x160
  [  165.390805]  ___sys_sendmsg+0x88/0xd0
  [  165.392524]  ? __might_fault+0x3e/0x90
  [  165.394273]  ? sock_getsockopt+0x3d5/0xbb0
  [  165.396021]  ? __handle_mm_fault+0x545/0x6a0
  [  165.397822]  ? find_held_lock+0x2d/0x90
  [  165.399593]  ? __sys_sendmsg+0x5e/0xa0
  [  165.401338]  __sys_sendmsg+0x5e/0xa0
  [  165.402979]  do_syscall_64+0x60/0x280
  [  165.404680]  entry_SYSCALL_64_after_hwframe+0x49/0xbe
  [  165.406621] RIP: 0033:0x7ff1fa46e783
  [  165.408299] Code: c7 c0 ff ff ff ff eb bb 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 44 00 00 64 8b 04 25 18 00 00 00 85 c0 75 14 b8 2e 00 00 00 0f 05 <48> 3d 00 f0 ff ff 77 55 c3 0f 1f 40 00 48 83 ec 28 89 54 24 1c 48
  [  165.414163] RSP: 002b:00007ffedf59ea78 EFLAGS: 00000246 ORIG_RAX: 000000000000002e
  [  165.416804] RAX: ffffffffffffffda RBX: 00007ffedf59fc60 RCX: 00007ff1fa46e783
  [  165.419419] RDX: 0000000000000000 RSI: 00007ffedf59fb10 RDI: 0000000000000005
  [  165.421886] RBP: 00007ffedf59fc10 R08: 00007ffedf59ea54 R09: 0000000000000001
  [  165.424445] R10: 00007ff1fa630c6c R11: 0000000000000246 R12: 0000000000020000
  [  165.426954] R13: 0000000000000280 R14: 0000000000000005 R15: 00007ffedf59ea90

Disable preemption before accessing the lookup scratch area in nft_pipapo_insert().

Cc <[email protected]> # 5.6.x
Fixes 3c4287f62044 ("nf_tables: Add set type for arbitrary concatenation of ranges")
---
 net/netfilter/nft_set_pipapo.c | 6 +++++-
 1 file changed, 5 insertions(+), 1 deletion(-)
net/netfilter/nft_set_pipapo.c +5 -1
index 8b5acc6910fd..8c04388296b0 100644
@@ -1242,7 +1242,9 @@ static int nft_pipapo_insert(const struct net *net, const struct nft_set *set,
 		end += NFT_PIPAPO_GROUPS_PADDED_SIZE(f);
 	}
 
-	if (!*this_cpu_ptr(m->scratch) || bsize_max > m->bsize_max) {
+	if (!*get_cpu_ptr(m->scratch) || bsize_max > m->bsize_max) {
+		put_cpu_ptr(m->scratch);
+
 		err = pipapo_realloc_scratch(m, bsize_max);
 		if (err)
 			return err;
@@ -1250,6 +1252,8 @@ static int nft_pipapo_insert(const struct net *net, const struct nft_set *set,
 		this_cpu_write(nft_pipapo_scratch_index, false);
 
 		m->bsize_max = bsize_max;
+	} else {
+		put_cpu_ptr(m->scratch);
 	}
 
 	*ext2 = &e->ext;
Signature · 1 line
2.20.1
author maintainer
#3 lore raw

PATCH 3/4 netfilter: ctnetlink: memleak in filter initialization error path

In reply to Pablo Neira Ayuso

Release the filter object in case of error.

Fixes cb8aa9a3affb ("netfilter: ctnetlink: add kernel side filtering for dump")
Reported-by [email protected]
---
 net/netfilter/nf_conntrack_netlink.c | 32 +++++++++++++++++++---------
 1 file changed, 22 insertions(+), 10 deletions(-)
net/netfilter/nf_conntrack_netlink.c +22 -10
index d7bd8b1f27d5..832eabecfbdd 100644
@@ -939,7 +939,8 @@ ctnetlink_alloc_filter(const struct nlattr * const cda[], u8 family)
 			filter->mark.mask = 0xffffffff;
 		}
 	} else if (cda[CTA_MARK_MASK]) {
-		return ERR_PTR(-EINVAL);
+		err = -EINVAL;
+		goto err_filter;
 	}
 #endif
 	if (!cda[CTA_FILTER])
@@ -947,15 +948,17 @@ ctnetlink_alloc_filter(const struct nlattr * const cda[], u8 family)
 
 	err = ctnetlink_parse_zone(cda[CTA_ZONE], &filter->zone);
 	if (err < 0)
-		return ERR_PTR(err);
+		goto err_filter;
 
 	err = ctnetlink_parse_filter(cda[CTA_FILTER], filter);
 	if (err < 0)
-		return ERR_PTR(err);
+		goto err_filter;
 
 	if (filter->orig_flags) {
-		if (!cda[CTA_TUPLE_ORIG])
-			return ERR_PTR(-EINVAL);
+		if (!cda[CTA_TUPLE_ORIG]) {
+			err = -EINVAL;
+			goto err_filter;
+		}
 
 		err = ctnetlink_parse_tuple_filter(cda, &filter->orig,
 						   CTA_TUPLE_ORIG,
@@ -963,23 +966,32 @@ ctnetlink_alloc_filter(const struct nlattr * const cda[], u8 family)
 						   &filter->zone,
 						   filter->orig_flags);
 		if (err < 0)
-			return ERR_PTR(err);
+			goto err_filter;
 	}
 
 	if (filter->reply_flags) {
-		if (!cda[CTA_TUPLE_REPLY])
-			return ERR_PTR(-EINVAL);
+		if (!cda[CTA_TUPLE_REPLY]) {
+			err = -EINVAL;
+			goto err_filter;
+		}
 
 		err = ctnetlink_parse_tuple_filter(cda, &filter->reply,
 						   CTA_TUPLE_REPLY,
 						   filter->family,
 						   &filter->zone,
 						   filter->orig_flags);
-		if (err < 0)
-			return ERR_PTR(err);
+		if (err < 0) {
+			err = -EINVAL;
+			goto err_filter;
+		}
 	}
 
 	return filter;
+
+err_filter:
+	kfree(filter);
+
+	return ERR_PTR(err);
 }
 
 static bool ctnetlink_needs_filter(u8 family, const struct nlattr * const *cda)
Signature · 1 line
2.20.1
author maintainer
#4 lore raw

PATCH 4/4 netfilter: nf_tables: hook list memleak in flowtable deletion

In reply to Pablo Neira Ayuso

After looking up for the flowtable hooks that need to be removed, release the hook objects in the deletion list. The error path needs to released these hook objects too.

Fixes abadb2f865d7 ("netfilter: nf_tables: delete devices from flowtable")
Reported-by [email protected]
---
 net/netfilter/nf_tables_api.c | 31 ++++++++++++++++++++++++-------
 1 file changed, 24 insertions(+), 7 deletions(-)
net/netfilter/nf_tables_api.c +24 -7
index 073aa1051d43..7647ecfa0d40 100644
@@ -6550,12 +6550,22 @@ static int nf_tables_newflowtable(struct net *net, struct sock *nlsk,
 	return err;
 }
 
+static void nft_flowtable_hook_release(struct nft_flowtable_hook *flowtable_hook)
+{
+	struct nft_hook *this, *next;
+
+	list_for_each_entry_safe(this, next, &flowtable_hook->list, list) {
+		list_del(&this->list);
+		kfree(this);
+	}
+}
+
 static int nft_delflowtable_hook(struct nft_ctx *ctx,
 				 struct nft_flowtable *flowtable)
 {
 	const struct nlattr * const *nla = ctx->nla;
 	struct nft_flowtable_hook flowtable_hook;
-	struct nft_hook *this, *next, *hook;
+	struct nft_hook *this, *hook;
 	struct nft_trans *trans;
 	int err;
 
@@ -6564,33 +6574,40 @@ static int nft_delflowtable_hook(struct nft_ctx *ctx,
 	if (err < 0)
 		return err;
 
-	list_for_each_entry_safe(this, next, &flowtable_hook.list, list) {
+	list_for_each_entry(this, &flowtable_hook.list, list) {
 		hook = nft_hook_list_find(&flowtable->hook_list, this);
 		if (!hook) {
 			err = -ENOENT;
 			goto err_flowtable_del_hook;
 		}
 		hook->inactive = true;
-		list_del(&this->list);
-		kfree(this);
 	}
 
 	trans = nft_trans_alloc(ctx, NFT_MSG_DELFLOWTABLE,
 				sizeof(struct nft_trans_flowtable));
-	if (!trans)
-		return -ENOMEM;
+	if (!trans) {
+		err = -ENOMEM;
+		goto err_flowtable_del_hook;
+	}
 
 	nft_trans_flowtable(trans) = flowtable;
 	nft_trans_flowtable_update(trans) = true;
 	INIT_LIST_HEAD(&nft_trans_flowtable_hooks(trans));
+	nft_flowtable_hook_release(&flowtable_hook);
 
 	list_add_tail(&trans->list, &ctx->net->nft.commit_list);
 
 	return 0;
 
 err_flowtable_del_hook:
-	list_for_each_entry(hook, &flowtable_hook.list, list)
+	list_for_each_entry(this, &flowtable_hook.list, list) {
+		hook = nft_hook_list_find(&flowtable->hook_list, this);
+		if (!hook)
+			break;
+
 		hook->inactive = false;
+	}
+	nft_flowtable_hook_release(&flowtable_hook);
 
 	return err;
 }
Signature · 1 line
2.20.1
maintainer
#5 lore raw
In reply to Pablo Neira Ayuso

From: Pablo Neira Ayuso <[email protected]>
Date: Sun, 14 Jun 2020 23:52:57 +0200

Quoted 8 lines

The following patchset contains Netfilter fixes for net:

1) Fix bogus EEXIST on element insertions to the rbtree with timeouts, from Stefano Brivio.

2) Preempt BUG splat in the pipapo element insertion path, also from Stefano.

Quoted 7 lines

3) Release filter from the ctnetlink error path.

4) Release flowtable hooks from the deletion path.

Please, pull these changes from:

git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf.git

Pulled, thanks Pablo.