Merge branch 'master' of git://git.kernel.org/pub/scm/linux/kernel/git/klassert/ipsec

Steffen Klassert says:

====================
pull request (net): ipsec 2021-08-04

1) Fix a sysbot reported memory leak in xfrm_user_rcv_msg.
From Pavel Skripkin.

2) Revert "xfrm: policy: Read seqcount outside of rcu-read side
in xfrm_policy_lookup_bytype". This commit tried to fix a
lockin bug, but only cured some of the symptoms. A proper
fix is applied on top of this revert.

3) Fix a locking bug on xfrm state hash resize. A recent change
on sequence counters accidentally repaced a spinlock by a mutex.
Fix from Frederic Weisbecker.

4) Fix possible user-memory-access in xfrm_user_rcv_msg_compat().
From Dmitry Safonov.

5) Add initialiation sefltest fot xfrm_spdattr_type_t.
From Dmitry Safonov.
====================

Signed-off-by: David S. Miller <[e-mail]>

include/net/netns/xfrm.h

···
 #endif
 	spinlock_t		xfrm_state_lock;
 	seqcount_spinlock_t	xfrm_state_hash_generation;
+	seqcount_spinlock_t	xfrm_policy_hash_generation;
 
 	spinlock_t xfrm_policy_lock;
 	struct mutex xfrm_cfg_mutex;

net/xfrm/xfrm_compat.c

···
 	len = nlmsg_attrlen(nlh_src, xfrm_msg_min[type]);
 
 	nla_for_each_attr(nla, attrs, len, remaining) {
-		int err = xfrm_xlate64_attr(dst, nla);
+		int err;
 
+		switch (type) {
+		case XFRM_MSG_NEWSPDINFO:
+			err = xfrm_nla_cpy(dst, nla, nla_len(nla));
+			break;
+		default:
+			err = xfrm_xlate64_attr(dst, nla);
+			break;
+		}
 		if (err)
 			return err;
 	}
···
 
 /* Calculates len of translated 64-bit message. */
 static size_t xfrm_user_rcv_calculate_len64(const struct nlmsghdr *src,
-					    struct nlattr *attrs[XFRMA_MAX+1])
+					    struct nlattr *attrs[XFRMA_MAX + 1],
+					    int maxtype)
 {
 	size_t len = nlmsg_len(src);
 
···
 	case XFRM_MSG_POLEXPIRE:
 		len += 8;
 		break;
+	case XFRM_MSG_NEWSPDINFO:
+		/* attirbutes are xfrm_spdattr_type_t, not xfrm_attr_type_t */
+		return len;
 	default:
 		break;
 	}
+
+	/* Unexpected for anything, but XFRM_MSG_NEWSPDINFO, please
+	 * correct both 64=>32-bit and 32=>64-bit translators to copy
+	 * new attributes.
+	 */
+	if (WARN_ON_ONCE(maxtype))
+		return len;
 
 	if (attrs[XFRMA_SA])
 		len += 4;
···
 
 static int xfrm_xlate32(struct nlmsghdr *dst, const struct nlmsghdr *src,
 			struct nlattr *attrs[XFRMA_MAX+1],
-			size_t size, u8 type, struct netlink_ext_ack *extack)
+			size_t size, u8 type, int maxtype,
+			struct netlink_ext_ack *extack)
 {
 	size_t pos;
 	int i;
···
 	}
 	pos = dst->nlmsg_len;
 
+	if (maxtype) {
+		/* attirbutes are xfrm_spdattr_type_t, not xfrm_attr_type_t */
+		WARN_ON_ONCE(src->nlmsg_type != XFRM_MSG_NEWSPDINFO);
+
+		for (i = 1; i <= maxtype; i++) {
+			int err;
+
+			if (!attrs[i])
+				continue;
+
+			/* just copy - no need for translation */
+			err = xfrm_attr_cpy32(dst, &pos, attrs[i], size,
+					nla_len(attrs[i]), nla_len(attrs[i]));
+			if (err)
+				return err;
+		}
+		return 0;
+	}
+
 	for (i = 1; i < XFRMA_MAX + 1; i++) {
 		int err;
 
···
 	if (err < 0)
 		return ERR_PTR(err);
 
-	len = xfrm_user_rcv_calculate_len64(h32, attrs);
+	len = xfrm_user_rcv_calculate_len64(h32, attrs, maxtype);
 	/* The message doesn't need translation */
 	if (len == nlmsg_len(h32))
 		return NULL;
···
 	if (!h64)
 		return ERR_PTR(-ENOMEM);
 
-	err = xfrm_xlate32(h64, h32, attrs, len, type, extack);
+	err = xfrm_xlate32(h64, h32, attrs, len, type, maxtype, extack);
 	if (err < 0) {
 		kvfree(h64);
 		return ERR_PTR(err);

net/xfrm/xfrm_ipcomp.c

···
 			break;
 	}
 
-	WARN_ON(!pos);
+	WARN_ON(list_entry_is_head(pos, &ipcomp_tfms_list, list));
 
 	if (--pos->users)
 		return;

net/xfrm/xfrm_policy.c

···
 						__read_mostly;
 
 static struct kmem_cache *xfrm_dst_cache __ro_after_init;
-static __read_mostly seqcount_mutex_t xfrm_policy_hash_generation;
 
 static struct rhashtable xfrm_policy_inexact_table;
 static const struct rhashtable_params xfrm_pol_inexact_params;
···
 		return;
 
 	spin_lock_bh(&net->xfrm.xfrm_policy_lock);
-	write_seqcount_begin(&xfrm_policy_hash_generation);
+	write_seqcount_begin(&net->xfrm.xfrm_policy_hash_generation);
 
 	odst = rcu_dereference_protected(net->xfrm.policy_bydst[dir].table,
 				lockdep_is_held(&net->xfrm.xfrm_policy_lock));
···
 	rcu_assign_pointer(net->xfrm.policy_bydst[dir].table, ndst);
 	net->xfrm.policy_bydst[dir].hmask = nhashmask;
 
-	write_seqcount_end(&xfrm_policy_hash_generation);
+	write_seqcount_end(&net->xfrm.xfrm_policy_hash_generation);
 	spin_unlock_bh(&net->xfrm.xfrm_policy_lock);
 
 	synchronize_rcu();
···
 	} while (read_seqretry(&net->xfrm.policy_hthresh.lock, seq));
 
 	spin_lock_bh(&net->xfrm.xfrm_policy_lock);
-	write_seqcount_begin(&xfrm_policy_hash_generation);
+	write_seqcount_begin(&net->xfrm.xfrm_policy_hash_generation);
 
 	/* make sure that we can insert the indirect policies again before
 	 * we start with destructive action.
···
 
 out_unlock:
 	__xfrm_policy_inexact_flush(net);
-	write_seqcount_end(&xfrm_policy_hash_generation);
+	write_seqcount_end(&net->xfrm.xfrm_policy_hash_generation);
 	spin_unlock_bh(&net->xfrm.xfrm_policy_lock);
 
 	mutex_unlock(&hash_resize_mutex);
···
 	if (unlikely(!daddr || !saddr))
 		return NULL;
 
- retry:
-	sequence = read_seqcount_begin(&xfrm_policy_hash_generation);
 	rcu_read_lock();
-
-	chain = policy_hash_direct(net, daddr, saddr, family, dir);
-	if (read_seqcount_retry(&xfrm_policy_hash_generation, sequence)) {
-		rcu_read_unlock();
-		goto retry;
-	}
+ retry:
+	do {
+		sequence = read_seqcount_begin(&net->xfrm.xfrm_policy_hash_generation);
+		chain = policy_hash_direct(net, daddr, saddr, family, dir);
+	} while (read_seqcount_retry(&net->xfrm.xfrm_policy_hash_generation, sequence));
 
 	ret = NULL;
 	hlist_for_each_entry_rcu(pol, chain, bydst) {
···
 	}
 
 skip_inexact:
-	if (read_seqcount_retry(&xfrm_policy_hash_generation, sequence)) {
-		rcu_read_unlock();
+	if (read_seqcount_retry(&net->xfrm.xfrm_policy_hash_generation, sequence))
 		goto retry;
-	}
 
-	if (ret && !xfrm_pol_hold_rcu(ret)) {
-		rcu_read_unlock();
+	if (ret && !xfrm_pol_hold_rcu(ret))
 		goto retry;
-	}
 fail:
 	rcu_read_unlock();
 
···
 	/* Initialize the per-net locks here */
 	spin_lock_init(&net->xfrm.xfrm_state_lock);
 	spin_lock_init(&net->xfrm.xfrm_policy_lock);
+	seqcount_spinlock_init(&net->xfrm.xfrm_policy_hash_generation, &net->xfrm.xfrm_policy_lock);
 	mutex_init(&net->xfrm.xfrm_cfg_mutex);
 
 	rv = xfrm_statistics_init(net);
···
 {
 	register_pernet_subsys(&xfrm_net_ops);
 	xfrm_dev_init();
-	seqcount_mutex_init(&xfrm_policy_hash_generation, &hash_resize_mutex);
 	xfrm_input_init();
 
 #ifdef CONFIG_XFRM_ESPINTCP

net/xfrm/xfrm_user.c

···
 
 	err = link->doit(skb, nlh, attrs);
 
+	/* We need to free skb allocated in xfrm_alloc_compat() before
+	 * returning from this function, because consume_skb() won't take
+	 * care of frag_list since netlink destructor sets
+	 * sbk->head to NULL. (see netlink_skb_destructor())
+	 */
+	if (skb_has_frag_list(skb)) {
+		kfree_skb(skb_shinfo(skb)->frag_list);
+		skb_shinfo(skb)->frag_list = NULL;
+	}
+
 err:
 	kvfree(nlh64);
 	return err;

tools/testing/selftests/net/ipsec.c

···
 	MONITOR_ACQUIRE,
 	EXPIRE_STATE,
 	EXPIRE_POLICY,
+	SPDINFO_ATTRS,
 };
 const char *desc_name[] = {
 	"create tunnel",
 	"alloc spi",
 	"monitor acquire",
 	"expire state",
-	"expire policy"
+	"expire policy",
+	"spdinfo attributes",
+	""
 };
 struct xfrm_desc {
 	enum desc_type	type;
···
 	return ret;
 }
 
+static int xfrm_spdinfo_set_thresh(int xfrm_sock, uint32_t *seq,
+		unsigned thresh4_l, unsigned thresh4_r,
+		unsigned thresh6_l, unsigned thresh6_r,
+		bool add_bad_attr)
+
+{
+	struct {
+		struct nlmsghdr		nh;
+		union {
+			uint32_t	unused;
+			int		error;
+		};
+		char			attrbuf[MAX_PAYLOAD];
+	} req;
+	struct xfrmu_spdhthresh thresh;
+
+	memset(&req, 0, sizeof(req));
+	req.nh.nlmsg_len	= NLMSG_LENGTH(sizeof(req.unused));
+	req.nh.nlmsg_type	= XFRM_MSG_NEWSPDINFO;
+	req.nh.nlmsg_flags	= NLM_F_REQUEST | NLM_F_ACK;
+	req.nh.nlmsg_seq	= (*seq)++;
+
+	thresh.lbits = thresh4_l;
+	thresh.rbits = thresh4_r;
+	if (rtattr_pack(&req.nh, sizeof(req), XFRMA_SPD_IPV4_HTHRESH, &thresh, sizeof(thresh)))
+		return -1;
+
+	thresh.lbits = thresh6_l;
+	thresh.rbits = thresh6_r;
+	if (rtattr_pack(&req.nh, sizeof(req), XFRMA_SPD_IPV6_HTHRESH, &thresh, sizeof(thresh)))
+		return -1;
+
+	if (add_bad_attr) {
+		BUILD_BUG_ON(XFRMA_IF_ID <= XFRMA_SPD_MAX + 1);
+		if (rtattr_pack(&req.nh, sizeof(req), XFRMA_IF_ID, NULL, 0)) {
+			pr_err("adding attribute failed: no space");
+			return -1;
+		}
+	}
+
+	if (send(xfrm_sock, &req, req.nh.nlmsg_len, 0) < 0) {
+		pr_err("send()");
+		return -1;
+	}
+
+	if (recv(xfrm_sock, &req, sizeof(req), 0) < 0) {
+		pr_err("recv()");
+		return -1;
+	} else if (req.nh.nlmsg_type != NLMSG_ERROR) {
+		printk("expected NLMSG_ERROR, got %d", (int)req.nh.nlmsg_type);
+		return -1;
+	}
+
+	if (req.error) {
+		printk("NLMSG_ERROR: %d: %s", req.error, strerror(-req.error));
+		return -1;
+	}
+
+	return 0;
+}
+
+static int xfrm_spdinfo_attrs(int xfrm_sock, uint32_t *seq)
+{
+	struct {
+		struct nlmsghdr			nh;
+		union {
+			uint32_t	unused;
+			int		error;
+		};
+		char			attrbuf[MAX_PAYLOAD];
+	} req;
+
+	if (xfrm_spdinfo_set_thresh(xfrm_sock, seq, 32, 31, 120, 16, false)) {
+		pr_err("Can't set SPD HTHRESH");
+		return KSFT_FAIL;
+	}
+
+	memset(&req, 0, sizeof(req));
+
+	req.nh.nlmsg_len	= NLMSG_LENGTH(sizeof(req.unused));
+	req.nh.nlmsg_type	= XFRM_MSG_GETSPDINFO;
+	req.nh.nlmsg_flags	= NLM_F_REQUEST;
+	req.nh.nlmsg_seq	= (*seq)++;
+	if (send(xfrm_sock, &req, req.nh.nlmsg_len, 0) < 0) {
+		pr_err("send()");
+		return KSFT_FAIL;
+	}
+
+	if (recv(xfrm_sock, &req, sizeof(req), 0) < 0) {
+		pr_err("recv()");
+		return KSFT_FAIL;
+	} else if (req.nh.nlmsg_type == XFRM_MSG_NEWSPDINFO) {
+		size_t len = NLMSG_PAYLOAD(&req.nh, sizeof(req.unused));
+		struct rtattr *attr = (void *)req.attrbuf;
+		int got_thresh = 0;
+
+		for (; RTA_OK(attr, len); attr = RTA_NEXT(attr, len)) {
+			if (attr->rta_type == XFRMA_SPD_IPV4_HTHRESH) {
+				struct xfrmu_spdhthresh *t = RTA_DATA(attr);
+
+				got_thresh++;
+				if (t->lbits != 32 || t->rbits != 31) {
+					pr_err("thresh differ: %u, %u",
+							t->lbits, t->rbits);
+					return KSFT_FAIL;
+				}
+			}
+			if (attr->rta_type == XFRMA_SPD_IPV6_HTHRESH) {
+				struct xfrmu_spdhthresh *t = RTA_DATA(attr);
+
+				got_thresh++;
+				if (t->lbits != 120 || t->rbits != 16) {
+					pr_err("thresh differ: %u, %u",
+							t->lbits, t->rbits);
+					return KSFT_FAIL;
+				}
+			}
+		}
+		if (got_thresh != 2) {
+			pr_err("only %d thresh returned by XFRM_MSG_GETSPDINFO", got_thresh);
+			return KSFT_FAIL;
+		}
+	} else if (req.nh.nlmsg_type != NLMSG_ERROR) {
+		printk("expected NLMSG_ERROR, got %d", (int)req.nh.nlmsg_type);
+		return KSFT_FAIL;
+	} else {
+		printk("NLMSG_ERROR: %d: %s", req.error, strerror(-req.error));
+		return -1;
+	}
+
+	/* Restore the default */
+	if (xfrm_spdinfo_set_thresh(xfrm_sock, seq, 32, 32, 128, 128, false)) {
+		pr_err("Can't restore SPD HTHRESH");
+		return KSFT_FAIL;
+	}
+
+	/*
+	 * At this moment xfrm uses nlmsg_parse_deprecated(), which
+	 * implies NL_VALIDATE_LIBERAL - ignoring attributes with
+	 * (type > maxtype). nla_parse_depricated_strict() would enforce
+	 * it. Or even stricter nla_parse().
+	 * Right now it's not expected to fail, but to be ignored.
+	 */
+	if (xfrm_spdinfo_set_thresh(xfrm_sock, seq, 32, 32, 128, 128, true))
+		return KSFT_PASS;
+
+	return KSFT_PASS;
+}
+
 static int child_serv(int xfrm_sock, uint32_t *seq,
 		unsigned int nr, int cmd_fd, void *buf, struct xfrm_desc *desc)
 {
···
 			break;
 		case EXPIRE_POLICY:
 			ret = xfrm_expire_policy(xfrm_sock, &seq, nr, &desc);
+			break;
+		case SPDINFO_ATTRS:
+			ret = xfrm_spdinfo_attrs(xfrm_sock, &seq);
 			break;
 		default:
 			printk("Unknown desc type %d", desc.type);
···
  *   sizeof(xfrm_user_polexpire)  = 168  |  sizeof(xfrm_user_polexpire)  = 176
  *
  * Check the affected by the UABI difference structures.
+ * Also, check translation for xfrm_set_spdinfo: it has it's own attributes
+ * which needs to be correctly copied, but not translated.
  */
-const unsigned int compat_plan = 4;
+const unsigned int compat_plan = 5;
 static int write_compat_struct_tests(int test_desc_fd)
 {
 	struct xfrm_desc desc = {};
···
 		return -1;
 
 	desc.type = EXPIRE_POLICY;
+	if (__write_desc(test_desc_fd, &desc))
+		return -1;
+
+	desc.type = SPDINFO_ATTRS;
 	if (__write_desc(test_desc_fd, &desc))
 		return -1;