wireguard fixes for 5.6-rc7
Subsystems (4)
Jason A. Donenfeld
maintainercover wireguard fixes for 5.6-rc7
Hi Dave,
I originally intended to spend this cycle working on fun optimizations and architecture for WireGuard for 5.7, but I've been a bit neurotic about having 5.6 ship without any show stopper bugs. WireGuard has been stable for a long time now, but that doesn't make me any less nervous about the real deal in 5.6. To that end, I've been doing code reviews and having discussions, and we also had a security firm audit the code. That audit didn't turn up any vulnerabilities, but they did make a good defense-in-depth suggestion. This series contains:
1) Removal of a duplicated header, from YueHaibing.
2) Testing with 64-bit time in our test suite.
3) Account for skb->protocol==0 due to AF_PACKET sockets, suggested by Florian Fainelli.
4) Clean up some code in an unreachable switch/case branch, suggested by Florian Fainelli.
5) Better handling of low-order points, discussed with Mathias Hall-Andersen.
Thanks,
Jason
Jason A. Donenfeld (4):
wireguard: selftests: test using new 64-bit time_t
wireguard: queueing: account for skb->protocol==0
wireguard: receive: remove dead code from default packet type case
wireguard: noise: error out precomputed DH during handshake rather
than config
YueHaibing (1):
wireguard: selftests: remove duplicated include <sys/types.h>
drivers/net/wireguard/device.c | 2 +- drivers/net/wireguard/netlink.c | 8 +-- drivers/net/wireguard/noise.c | 55 ++++++++++--------- drivers/net/wireguard/noise.h | 12 ++-- drivers/net/wireguard/peer.c | 7 +-- drivers/net/wireguard/queueing.h | 8 ++- drivers/net/wireguard/receive.c | 7 +-- tools/testing/selftests/wireguard/netns.sh | 6 -- .../testing/selftests/wireguard/qemu/Makefile | 2 +- tools/testing/selftests/wireguard/qemu/init.c | 1 - .../selftests/wireguard/qemu/kernel.config | 1 - 11 files changed, 51 insertions(+), 58 deletions(-)
Signature · 1 line
2.25.1
Jason A. Donenfeld
maintainerPATCH 1/5 wireguard: selftests: remove duplicated include <sys/types.h>
From: YueHaibing <[email protected]>
This commit removes a duplicated include.
--- tools/testing/selftests/wireguard/qemu/init.c | 1 - 1 file changed, 1 deletion(-)
tools/testing/selftests/wireguard/qemu/init.c +0 -1
@@ -13,7 +13,6 @@#include <fcntl.h> #include <sys/wait.h> #include <sys/mount.h> -#include <sys/types.h> #include <sys/stat.h> #include <sys/types.h> #include <sys/io.h>
Signature · 1 line
2.25.1
Jason A. Donenfeld
maintainerPATCH 2/5 wireguard: selftests: test using new 64-bit time_t
In reply to Jason A. DonenfeldIn case this helps expose bugs with the newer 64-bit time_t types, we do our testing with the newer musl that supports this as well as CONFIG_COMPAT_32BIT_TIME=n. This matters to us, since wireguard does in fact deal with timestamps.
--- tools/testing/selftests/wireguard/qemu/Makefile | 2 +- tools/testing/selftests/wireguard/qemu/kernel.config | 1 - 2 files changed, 1 insertion(+), 2 deletions(-)
tools/testing/selftests/wireguard/qemu/Makefile +1 -1
@@ -41,7 +41,7 @@ $(DISTFILES_PATH)/$(1): flock -x [email protected] -c '[ -f $$@ ] && exit 0; wget -O [email protected] $(MIRROR)$(1) || wget -O [email protected] $(2)$(1) || rm -f [email protected]; [ -f [email protected] ] || exit 1; if echo "$(3) [email protected]" | sha256sum -c -; then mv [email protected] $$@; else rm -f [email protected]; exit 71; fi' endef -$(eval $(call tar_download,MUSL,musl,1.1.24,.tar.gz,https://www.musl-libc.org/releases/,1370c9a812b2cf2a7d92802510cca0058cc37e66a7bedd70051f0a34015022a3)) +$(eval $(call tar_download,MUSL,musl,1.2.0,.tar.gz,https://musl.libc.org/releases/,c6de7b191139142d3f9a7b5b702c9cae1b5ee6e7f57e582da9328629408fd4e8)) $(eval $(call tar_download,IPERF,iperf,3.7,.tar.gz,https://downloads.es.net/pub/iperf/,d846040224317caf2f75c843d309a950a7db23f9b44b94688ccbe557d6d1710c)) $(eval $(call tar_download,BASH,bash,5.0,.tar.gz,https://ftp.gnu.org/gnu/bash/,b4a80f2ac66170b2913efbfb9f2594f1f76c7b1afd11f799e22035d63077fb4d)) $(eval $(call tar_download,IPROUTE2,iproute2,5.4.0,.tar.xz,https://www.kernel.org/pub/linux/utils/net/iproute2/,fe97aa60a0d4c5ac830be18937e18dc3400ca713a33a89ad896ff1e3d46086ae))
tools/testing/selftests/wireguard/qemu/kernel.config +0 -1
@@ -56,7 +56,6 @@ CONFIG_NO_HZ_IDLE=y CONFIG_NO_HZ_FULL=n CONFIG_HZ_PERIODIC=n CONFIG_HIGH_RES_TIMERS=y -CONFIG_COMPAT_32BIT_TIME=y CONFIG_ARCH_RANDOM=y CONFIG_FILE_LOCKING=y CONFIG_POSIX_TIMERS=y
Signature · 1 line
2.25.1
Jason A. Donenfeld
maintainerPATCH 3/5 wireguard: queueing: account for skb->protocol==0
In reply to Jason A. DonenfeldWe carry out checks to the effect of:
if (skb->protocol != wg_examine_packet_protocol(skb))
goto err;
By having wg_skb_examine_untrusted_ip_hdr return 0 on failure, this means that the check above still passes in the case where skb->protocol is zero, which is possible to hit with AF_PACKET:
struct sockaddr_pkt saddr = { .spkt_device = "wg0" };
unsigned char buffer[5] = { 0 };
sendto(socket(AF_PACKET, SOCK_PACKET, /* skb->protocol = */ 0),
buffer, sizeof(buffer), 0, (const struct sockaddr *)&saddr, sizeof(saddr));Additional checks mean that this isn't actually a problem in the code base, but I could imagine it becoming a problem later if the function is used more liberally.
I would prefer to fix this by having wg_examine_packet_protocol return a 32-bit ~0 value on failure, which will never match any value of skb->protocol, which would simply change the generated code from a mov to a movzx. However, sparse complains, and adding __force casts doesn't seem like a good idea, so instead we just add a simple helper function to check for the zero return value. Since wg_examine_packet_protocol itself gets inlined, this winds up not adding an additional branch to the generated code, since the 0 return value already happens in a mergable branch.
--- drivers/net/wireguard/device.c | 2 +- drivers/net/wireguard/queueing.h | 8 +++++++- drivers/net/wireguard/receive.c | 4 ++-- 3 files changed, 10 insertions(+), 4 deletions(-)
drivers/net/wireguard/device.c +1 -1
@@ -122,7 +122,7 @@ static netdev_tx_t wg_xmit(struct sk_buff *skb, struct net_device *dev)u32 mtu;int ret; -if (unlikely (wg_skb_examine_untrusted_ip_hdr (skb) != skb->protocol)) { +if (unlikely (!wg_check_packet_protocol (skb))) { ret = -EPROTONOSUPPORT ;net_dbg_ratelimited ("%s: Invalid IP packet\n" , dev->name); goto err;
drivers/net/wireguard/queueing.h +7 -1
@@ -66,7 +66,7 @@ struct packet_cb {#define PACKET_PEER (skb) (PACKET_CB(skb)->keypair->entry.peer)/* Returns either the correct skb->protocol value, or 0 if invalid. */ -static inline __be16 wg_skb_examine_untrusted_ip_hdr(struct sk_buff *skb) +static inline __be16 wg_examine_packet_protocol(struct sk_buff *skb) { if (skb_network_header(skb) >= skb->head && (skb_network_header(skb) + sizeof(struct iphdr)) <=
@@ -81,6 +81,12 @@ static inline __be16 wg_skb_examine_untrusted_ip_hdr(struct sk_buff *skb)return 0 ; } +static inline bool wg_check_packet_protocol (struct sk_buff *skb) +{ +__be16 real_protocol =wg_examine_packet_protocol (skb); +return real_protocol && skb->protocol == real_protocol; +} +static inline void wg_reset_packet (struct sk_buff *skb) {skb_scrub_packet (skb, true);
drivers/net/wireguard/receive.c +2 -2
@@ -56,7 +56,7 @@ static int prepare_skb_header(struct sk_buff *skb, struct wg_device *wg)size_t data_offset, data_len, header_len;struct udphdr *udp; -if (unlikely (wg_skb_examine_untrusted_ip_hdr (skb) != skb->protocol || +if (unlikely (!wg_check_packet_protocol (skb) ||skb_transport_header (skb) < skb->head || (skb_transport_header (skb) +sizeof (struct udphdr )) >skb_tail_pointer (skb)))
@@ -388,7 +388,7 @@ static void wg_packet_consume_data_done(struct wg_peer *peer,*/ skb->ip_summed =CHECKSUM_UNNECESSARY ; skb->csum_level = ~0 ;/* All levels */ - skb->protocol =wg_skb_examine_untrusted_ip_hdr (skb); + skb->protocol =wg_examine_packet_protocol (skb);if (skb->protocol ==htons (ETH_P_IP )) { len =ntohs (ip_hdr (skb)->tot_len);if (unlikely (len <sizeof (struct iphdr )))
Signature · 1 line
2.25.1
Jason A. Donenfeld
maintainerPATCH 4/5 wireguard: receive: remove dead code from default packet type case
In reply to Jason A. DonenfeldThe situation in which we wind up hitting the default case here indicates a major bug in earlier parsing code. It is not a usual thing that should ever happen, which means a "friendly" message for it doesn't make sense. Rather, replace this with a WARN_ON, just like we do earlier in the file for a similar situation, so that somebody sends us a bug report and we can fix it.
--- drivers/net/wireguard/receive.c | 3 +-- 1 file changed, 1 insertion(+), 2 deletions(-)
drivers/net/wireguard/receive.c +1 -2
@@ -587,8 +587,7 @@ void wg_packet_receive(struct wg_device *wg, struct sk_buff *skb)wg_packet_consume_data (wg, skb);break ;default : -net_dbg_skb_ratelimited ("%s: Invalid packet from %pISpfsc\n" , - wg->dev->name, skb); +WARN (1 ,"Non-exhaustive parsing of packet header lead to unknown packet type!\n" ); goto err; }return ;
Signature · 1 line
2.25.1
Jason A. Donenfeld
maintainerPATCH 5/5 wireguard: noise: error out precomputed DH during handshake rather than config
In reply to Jason A. DonenfeldWe precompute the static-static ECDH during configuration time, in order to save an expensive computation later when receiving network packets. However, not all ECDH computations yield a contributory result. Prior, we were just not letting those peers be added to the interface. However, this creates a strange inconsistency, since it was still possible to add other weird points, like a valid public key plus a low-order point, and, like points that result in zeros, a handshake would not complete. In order to make the behavior more uniform and less surprising, simply allow all peers to be added. Then, we'll error out later when doing the crypto if there's an issue. This also adds more separation between the crypto layer and the configuration layer.
Discussed-with: Mathias Hall-Andersen <[email protected]>
--- drivers/net/wireguard/netlink.c | 8 +--- drivers/net/wireguard/noise.c | 55 ++++++++++++---------- drivers/net/wireguard/noise.h | 12 ++--- drivers/net/wireguard/peer.c | 7 +-- tools/testing/selftests/wireguard/netns.sh | 15 ++++-- 5 files changed, 49 insertions(+), 48 deletions(-)
drivers/net/wireguard/netlink.c +2 -6
@@ -411,11 +411,7 @@ static int set_peer(struct wg_device *wg, struct nlattr **attrs) peer =wg_peer_create (wg, public_key, preshared_key);if (IS_ERR (peer)) { -/* Similar to the above, if the key is invalid, we skip -* it without fanfare, so that services don't need to -* worry about doing key validation themselves. -*/ - ret =PTR_ERR (peer) == -EKEYREJECTED ?0 :PTR_ERR (peer); + ret =PTR_ERR (peer); peer =NULL ; goto out; }
@@ -569,7 +565,7 @@ static int wg_set_device(struct sk_buff *skb, struct genl_info *info) private_key);list_for_each_entry_safe (peer, temp, &wg->peer_list, peer_list) { -BUG_ON (!wg_noise_precompute_static_static (peer)); +wg_noise_precompute_static_static (peer);wg_noise_expire_current_peer_keypairs (peer); }wg_cookie_checker_precompute_device_keys (&wg->cookie_checker);
drivers/net/wireguard/noise.c +29 -26
93 lines not on this page · raw message
drivers/net/wireguard/noise.h +6 -6
@@ -94,11 +94,11 @@ struct noise_handshake {struct wg_device ;void wg_noise_init (void ); -bool wg_noise_handshake_init (struct noise_handshake *handshake, -struct noise_static_identity *static_identity, -const u8 peer_public_key[NOISE_PUBLIC_KEY_LEN ], -const u8 peer_preshared_key[NOISE_SYMMETRIC_KEY_LEN ], -struct wg_peer *peer); +void wg_noise_handshake_init (struct noise_handshake *handshake, +struct noise_static_identity *static_identity, +const u8 peer_public_key[NOISE_PUBLIC_KEY_LEN ], +const u8 peer_preshared_key[NOISE_SYMMETRIC_KEY_LEN ], +struct wg_peer *peer);void wg_noise_handshake_clear (struct noise_handshake *handshake);static inline void wg_noise_reset_last_sent_handshake (atomic64_t *handshake_ns) {
@@ -116,7 +116,7 @@ void wg_noise_expire_current_peer_keypairs(struct wg_peer *peer);void wg_noise_set_static_identity_private_key (struct noise_static_identity *static_identity,const u8 private_key[NOISE_PUBLIC_KEY_LEN ]); -bool wg_noise_precompute_static_static (struct wg_peer *peer); +void wg_noise_precompute_static_static (struct wg_peer *peer); bool wg_noise_handshake_create_initiation(struct message_handshake_initiation *dst,
drivers/net/wireguard/peer.c +2 -5
@@ -34,11 +34,8 @@ struct wg_peer *wg_peer_create(struct wg_device *wg,return ERR_PTR (ret); peer->device = wg; -if (!wg_noise_handshake_init (&peer->handshake, &wg->static_identity, - public_key, preshared_key, peer)) { - ret = -EKEYREJECTED ; - goto err_1; - } +wg_noise_handshake_init (&peer->handshake, &wg->static_identity, + public_key, preshared_key, peer);if (dst_cache_init (&peer->endpoint_cache,GFP_KERNEL )) goto err_1; if (wg_packet_queue_init(&peer->tx_queue, wg_packet_tx_worker, false,
tools/testing/selftests/wireguard/netns.sh +10 -5
@@ -527,11 +527,16 @@ n0 wg set wg0 peer "$pub2" allowed-ips 0.0.0.0/0 n0 wg set wg0 peer "$pub2" allowed-ips ::/0,1700::/111,5000::/4,e000::/37,9000::/75 n0 wg set wg0 peer "$pub2" allowed-ips ::/0 n0 wg set wg0 peer "$pub2" remove -low_order_points=( AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA= AQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA= 4Ot6fDtBuK4WVuP68Z/EatoJjeucMrH9hmIFFl9JuAA= X5yVvKNQjCSx0LFVnIPvWwREXMRYHI6G2CJO3dCfEVc= 7P///////////////////////////////////////38= 7f///////////////////////////////////////38= 7v///////////////////////////////////////38= ) -n0 wg set wg0 private-key /dev/null ${low_order_points[@]/#/peer } -[[ -z $(n0 wg show wg0 peers) ]] -n0 wg set wg0 private-key <(echo "$key1") ${low_order_points[@]/#/peer } -[[ -z $(n0 wg show wg0 peers) ]] +for low_order_point in AAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA= AQAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAAA= 4Ot6fDtBuK4WVuP68Z/EatoJjeucMrH9hmIFFl9JuAA= X5yVvKNQjCSx0LFVnIPvWwREXMRYHI6G2CJO3dCfEVc= 7P///////////////////////////////////////38= 7f///////////////////////////////////////38= 7v///////////////////////////////////////38=; do + n0 wg set wg0 peer "$low_order_point" persistent-keepalive 1 endpoint 127.0.0.1:1111 +done +[[ -n $(n0 wg show wg0 peers) ]] +exec 4< <(n0 ncat -l -u -p 1111) +ncat_pid=$! +waitncatudp $netns0 $ncat_pid +ip0 link set wg0 up +! read -r -n 1 -t 2 <&4 || false +kill $ncat_pid ip0 link del wg0 declare -A objects
Signature · 1 line
2.25.1
David Miller
maintainerFrom: "Jason A. Donenfeld" <[email protected]>
Date: Wed, 18 Mar 2020 18:30:42 -0600
Quoted 9 lines
I originally intended to spend this cycle working on fun optimizations and architecture for WireGuard for 5.7, but I've been a bit neurotic about having 5.6 ship without any show stopper bugs. WireGuard has been stable for a long time now, but that doesn't make me any less nervous about the real deal in 5.6. To that end, I've been doing code reviews and having discussions, and we also had a security firm audit the code. That audit didn't turn up any vulnerabilities, but they did make a good defense-in-depth suggestion. This series contains:
Quoted 8 lines
1) Removal of a duplicated header, from YueHaibing.
2) Testing with 64-bit time in our test suite.
3) Account for skb->protocol==0 due toAF_PACKETsockets, suggested by Florian Fainelli.
4) Clean up some code in an unreachable switch/case branch, suggested by Florian Fainelli.
5) Better handling of low-order points, discussed with Mathias Hall-Andersen.
Series applied, please start providing appropriate Fixes: tags in the future.
Thank you.
Jason A. Donenfeld
maintainerQuoted 2 lines · David Miller
Series applied, please start providing appropriate Fixes: tags in the future.
No problem, will do.
Jason