From 0e78a87306a6f55b1c7bbafad1de62c3975953ca Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Wed, 3 May 2017 08:44:27 +0200 Subject: esp4: Fix udpencap for local TCP packets. Locally generated TCP packets are usually cloned, so we do skb_cow_data() on this packets. After that we need to reload the pointer to the esp header. On udpencap this header has an offset to skb_transport_header, so take this offset into account. Fixes: 67d349ed603 ("net/esp4: Fix invalid esph pointer crash") Fixes: fca11ebde3f0 ("esp4: Reorganize esp_output") Reported-by: Don Bowman Signed-off-by: Steffen Klassert --- net/ipv4/esp4.c | 5 ++++- 1 file changed, 4 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/net/ipv4/esp4.c b/net/ipv4/esp4.c index 65cc02bd82bc..93322f895eab 100644 --- a/net/ipv4/esp4.c +++ b/net/ipv4/esp4.c @@ -248,6 +248,7 @@ int esp_output_head(struct xfrm_state *x, struct sk_buff *skb, struct esp_info * u8 *tail; u8 *vaddr; int nfrags; + int esph_offset; struct page *page; struct sk_buff *trailer; int tailen = esp->tailen; @@ -313,11 +314,13 @@ int esp_output_head(struct xfrm_state *x, struct sk_buff *skb, struct esp_info * } cow: + esph_offset = (unsigned char *)esp->esph - skb_transport_header(skb); + nfrags = skb_cow_data(skb, tailen, &trailer); if (nfrags < 0) goto out; tail = skb_tail_pointer(trailer); - esp->esph = ip_esp_hdr(skb); + esp->esph = (struct ip_esp_hdr *)(skb_transport_header(skb) + esph_offset); skip_cow: esp_output_fill_trailer(tail, esp->tfclen, esp->plen, esp->proto); -- cgit v1.2.3 From 9b3eb54106cf6acd03f07cf0ab01c13676a226c2 Mon Sep 17 00:00:00 2001 From: Sabrina Dubroca Date: Wed, 3 May 2017 16:43:19 +0200 Subject: xfrm: fix stack access out of bounds with CONFIG_XFRM_SUB_POLICY When CONFIG_XFRM_SUB_POLICY=y, xfrm_dst stores a copy of the flowi for that dst. Unfortunately, the code that allocates and fills this copy doesn't care about what type of flowi (flowi, flowi4, flowi6) gets passed. In multiple code paths (from raw_sendmsg, from TCP when replying to a FIN, in vxlan, geneve, and gre), the flowi that gets passed to xfrm is actually an on-stack flowi4, so we end up reading stuff from the stack past the end of the flowi4 struct. Since xfrm_dst->origin isn't used anywhere following commit ca116922afa8 ("xfrm: Eliminate "fl" and "pol" args to xfrm_bundle_ok()."), just get rid of it. xfrm_dst->partner isn't used either, so get rid of that too. Fixes: 9d6ec938019c ("ipv4: Use flowi4 in public route lookup interfaces.") Signed-off-by: Sabrina Dubroca Signed-off-by: Steffen Klassert --- include/net/xfrm.h | 10 ---------- net/xfrm/xfrm_policy.c | 47 ----------------------------------------------- 2 files changed, 57 deletions(-) (limited to 'net') diff --git a/include/net/xfrm.h b/include/net/xfrm.h index 6793a30c66b1..7e7e2b0d2915 100644 --- a/include/net/xfrm.h +++ b/include/net/xfrm.h @@ -979,10 +979,6 @@ struct xfrm_dst { struct flow_cache_object flo; struct xfrm_policy *pols[XFRM_POLICY_TYPE_MAX]; int num_pols, num_xfrms; -#ifdef CONFIG_XFRM_SUB_POLICY - struct flowi *origin; - struct xfrm_selector *partner; -#endif u32 xfrm_genid; u32 policy_genid; u32 route_mtu_cached; @@ -998,12 +994,6 @@ static inline void xfrm_dst_destroy(struct xfrm_dst *xdst) dst_release(xdst->route); if (likely(xdst->u.dst.xfrm)) xfrm_state_put(xdst->u.dst.xfrm); -#ifdef CONFIG_XFRM_SUB_POLICY - kfree(xdst->origin); - xdst->origin = NULL; - kfree(xdst->partner); - xdst->partner = NULL; -#endif } #endif diff --git a/net/xfrm/xfrm_policy.c b/net/xfrm/xfrm_policy.c index b00a1d5a7f52..ed4e52d95172 100644 --- a/net/xfrm/xfrm_policy.c +++ b/net/xfrm/xfrm_policy.c @@ -1797,43 +1797,6 @@ free_dst: goto out; } -#ifdef CONFIG_XFRM_SUB_POLICY -static int xfrm_dst_alloc_copy(void **target, const void *src, int size) -{ - if (!*target) { - *target = kmalloc(size, GFP_ATOMIC); - if (!*target) - return -ENOMEM; - } - - memcpy(*target, src, size); - return 0; -} -#endif - -static int xfrm_dst_update_parent(struct dst_entry *dst, - const struct xfrm_selector *sel) -{ -#ifdef CONFIG_XFRM_SUB_POLICY - struct xfrm_dst *xdst = (struct xfrm_dst *)dst; - return xfrm_dst_alloc_copy((void **)&(xdst->partner), - sel, sizeof(*sel)); -#else - return 0; -#endif -} - -static int xfrm_dst_update_origin(struct dst_entry *dst, - const struct flowi *fl) -{ -#ifdef CONFIG_XFRM_SUB_POLICY - struct xfrm_dst *xdst = (struct xfrm_dst *)dst; - return xfrm_dst_alloc_copy((void **)&(xdst->origin), fl, sizeof(*fl)); -#else - return 0; -#endif -} - static int xfrm_expand_policies(const struct flowi *fl, u16 family, struct xfrm_policy **pols, int *num_pols, int *num_xfrms) @@ -1905,16 +1868,6 @@ xfrm_resolve_and_create_bundle(struct xfrm_policy **pols, int num_pols, xdst = (struct xfrm_dst *)dst; xdst->num_xfrms = err; - if (num_pols > 1) - err = xfrm_dst_update_parent(dst, &pols[1]->selector); - else - err = xfrm_dst_update_origin(dst, fl); - if (unlikely(err)) { - dst_free(dst); - XFRM_INC_STATS(net, LINUX_MIB_XFRMOUTBUNDLECHECKERROR); - return ERR_PTR(err); - } - xdst->num_pols = num_pols; memcpy(xdst->pols, pols, sizeof(struct xfrm_policy *) * num_pols); xdst->policy_genid = atomic_read(&pols[0]->genid); -- cgit v1.2.3 From d90c902449a7561f1b1d58ba5a0d11728ce8b0b2 Mon Sep 17 00:00:00 2001 From: Steffen Klassert Date: Fri, 5 May 2017 07:40:42 +0200 Subject: af_key: Fix slab-out-of-bounds in pfkey_compile_policy. The sadb_x_sec_len is stored in the unit 'byte divided by eight'. So we have to multiply this value by eight before we can do size checks. Otherwise we may get a slab-out-of-bounds when we memcpy the user sec_ctx. Fixes: df71837d502 ("[LSM-IPSec]: Security association restriction.") Reported-by: Andrey Konovalov Tested-by: Andrey Konovalov Signed-off-by: Steffen Klassert --- net/key/af_key.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net') diff --git a/net/key/af_key.c b/net/key/af_key.c index c1950bb14735..512dc43d0ce6 100644 --- a/net/key/af_key.c +++ b/net/key/af_key.c @@ -3285,7 +3285,7 @@ static struct xfrm_policy *pfkey_compile_policy(struct sock *sk, int opt, p += pol->sadb_x_policy_len*8; sec_ctx = (struct sadb_x_sec_ctx *)p; if (len < pol->sadb_x_policy_len*8 + - sec_ctx->sadb_x_sec_len) { + sec_ctx->sadb_x_sec_len*8) { *dir = -EINVAL; goto out; } -- cgit v1.2.3 From 2c1497bbc8fdee897341ab48ee9c9209b421b8c0 Mon Sep 17 00:00:00 2001 From: Ilan Tayari Date: Mon, 8 May 2017 10:30:18 +0300 Subject: xfrm: Fix NETDEV_DOWN with IPSec offload Upon NETDEV_DOWN event, all xfrm_state objects which are bound to the device are flushed. The condition for this is wrong, though, testing dev->hw_features instead of dev->features. If a device has non-user-modifiable NETIF_F_HW_ESP, then its xfrm_state objects are not flushed, causing a crash later on after the device is deleted. Check dev->features instead of dev->hw_features. Fixes: d77e38e612a0 ("xfrm: Add an IPsec hardware offloading API") Signed-off-by: Ilan Tayari Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_device.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net') diff --git a/net/xfrm/xfrm_device.c b/net/xfrm/xfrm_device.c index 8ec8a3fcf8d4..574e6f32f94f 100644 --- a/net/xfrm/xfrm_device.c +++ b/net/xfrm/xfrm_device.c @@ -170,7 +170,7 @@ static int xfrm_dev_feat_change(struct net_device *dev) static int xfrm_dev_down(struct net_device *dev) { - if (dev->hw_features & NETIF_F_HW_ESP) + if (dev->features & NETIF_F_HW_ESP) xfrm_dev_state_flush(dev_net(dev), dev, true); xfrm_garbage_collect(dev_net(dev)); -- cgit v1.2.3 From 5667c86acf021e6dcf02584408b4484a273ac68f Mon Sep 17 00:00:00 2001 From: Rajkumar Manoharan Date: Sun, 14 May 2017 21:41:55 -0700 Subject: mac80211: strictly check mesh address extension mode Mesh forwarding path checks for address extension mode to fetch appropriate proxied address and MPP address. Existing condition that looks for 6 address format is not strict enough so that frames with improper values are processed and invalid entries are added into MPP table. Fix that by adding a stricter check before processing the packet. Per IEEE Std 802.11s-2011 spec. Table 7-6g1 lists address extension mode 0x3 as reserved one. And also Table Table 9-13 does not specify 0x3 as valid address field. Fixes: 9b395bc3be1c ("mac80211: verify that skb data is present") Signed-off-by: Rajkumar Manoharan Signed-off-by: Johannes Berg --- net/mac80211/rx.c | 3 ++- net/wireless/util.c | 10 ++++++---- 2 files changed, 8 insertions(+), 5 deletions(-) (limited to 'net') diff --git a/net/mac80211/rx.c b/net/mac80211/rx.c index 35f4c7d7a500..1f75280ba26c 100644 --- a/net/mac80211/rx.c +++ b/net/mac80211/rx.c @@ -2492,7 +2492,8 @@ ieee80211_rx_h_mesh_fwding(struct ieee80211_rx_data *rx) if (is_multicast_ether_addr(hdr->addr1)) { mpp_addr = hdr->addr3; proxied_addr = mesh_hdr->eaddr1; - } else if (mesh_hdr->flags & MESH_FLAGS_AE_A5_A6) { + } else if ((mesh_hdr->flags & MESH_FLAGS_AE) == + MESH_FLAGS_AE_A5_A6) { /* has_a4 already checked in ieee80211_rx_mesh_check */ mpp_addr = hdr->addr4; proxied_addr = mesh_hdr->eaddr2; diff --git a/net/wireless/util.c b/net/wireless/util.c index 7198373e2920..4992f1025c9d 100644 --- a/net/wireless/util.c +++ b/net/wireless/util.c @@ -454,6 +454,8 @@ int ieee80211_data_to_8023_exthdr(struct sk_buff *skb, struct ethhdr *ehdr, if (iftype == NL80211_IFTYPE_MESH_POINT) skb_copy_bits(skb, hdrlen, &mesh_flags, 1); + mesh_flags &= MESH_FLAGS_AE; + switch (hdr->frame_control & cpu_to_le16(IEEE80211_FCTL_TODS | IEEE80211_FCTL_FROMDS)) { case cpu_to_le16(IEEE80211_FCTL_TODS): @@ -469,9 +471,9 @@ int ieee80211_data_to_8023_exthdr(struct sk_buff *skb, struct ethhdr *ehdr, iftype != NL80211_IFTYPE_STATION)) return -1; if (iftype == NL80211_IFTYPE_MESH_POINT) { - if (mesh_flags & MESH_FLAGS_AE_A4) + if (mesh_flags == MESH_FLAGS_AE_A4) return -1; - if (mesh_flags & MESH_FLAGS_AE_A5_A6) { + if (mesh_flags == MESH_FLAGS_AE_A5_A6) { skb_copy_bits(skb, hdrlen + offsetof(struct ieee80211s_hdr, eaddr1), tmp.h_dest, 2 * ETH_ALEN); @@ -487,9 +489,9 @@ int ieee80211_data_to_8023_exthdr(struct sk_buff *skb, struct ethhdr *ehdr, ether_addr_equal(tmp.h_source, addr))) return -1; if (iftype == NL80211_IFTYPE_MESH_POINT) { - if (mesh_flags & MESH_FLAGS_AE_A5_A6) + if (mesh_flags == MESH_FLAGS_AE_A5_A6) return -1; - if (mesh_flags & MESH_FLAGS_AE_A4) + if (mesh_flags == MESH_FLAGS_AE_A4) skb_copy_bits(skb, hdrlen + offsetof(struct ieee80211s_hdr, eaddr1), tmp.h_source, ETH_ALEN); -- cgit v1.2.3 From a486cd23661c9387fb076c3f6ae8b2aa9d20d54a Mon Sep 17 00:00:00 2001 From: Antony Antony Date: Fri, 19 May 2017 12:47:00 +0200 Subject: xfrm: fix state migration copy replay sequence numbers During xfrm migration copy replay and preplay sequence numbers from the previous state. Here is a tcpdump output showing the problem. 10.0.10.46 is running vanilla kernel, is the IKE/IPsec responder. After the migration it sent wrong sequence number, reset to 1. The migration is from 10.0.0.52 to 10.0.0.53. IP 10.0.0.52.4500 > 10.0.10.46.4500: UDP-encap: ESP(spi=0x43ef462d,seq=0x7cf), length 136 IP 10.0.10.46.4500 > 10.0.0.52.4500: UDP-encap: ESP(spi=0xca1c282d,seq=0x7cf), length 136 IP 10.0.0.52.4500 > 10.0.10.46.4500: UDP-encap: ESP(spi=0x43ef462d,seq=0x7d0), length 136 IP 10.0.10.46.4500 > 10.0.0.52.4500: UDP-encap: ESP(spi=0xca1c282d,seq=0x7d0), length 136 IP 10.0.0.53.4500 > 10.0.10.46.4500: NONESP-encap: isakmp: child_sa inf2[I] IP 10.0.10.46.4500 > 10.0.0.53.4500: NONESP-encap: isakmp: child_sa inf2[R] IP 10.0.0.53.4500 > 10.0.10.46.4500: NONESP-encap: isakmp: child_sa inf2[I] IP 10.0.10.46.4500 > 10.0.0.53.4500: NONESP-encap: isakmp: child_sa inf2[R] IP 10.0.0.53.4500 > 10.0.10.46.4500: UDP-encap: ESP(spi=0x43ef462d,seq=0x7d1), length 136 NOTE: next sequence is wrong 0x1 IP 10.0.10.46.4500 > 10.0.0.53.4500: UDP-encap: ESP(spi=0xca1c282d,seq=0x1), length 136 IP 10.0.0.53.4500 > 10.0.10.46.4500: UDP-encap: ESP(spi=0x43ef462d,seq=0x7d2), length 136 IP 10.0.10.46.4500 > 10.0.0.53.4500: UDP-encap: ESP(spi=0xca1c282d,seq=0x2), length 136 Signed-off-by: Antony Antony Reviewed-by: Richard Guy Briggs Signed-off-by: Steffen Klassert --- net/xfrm/xfrm_state.c | 2 ++ 1 file changed, 2 insertions(+) (limited to 'net') diff --git a/net/xfrm/xfrm_state.c b/net/xfrm/xfrm_state.c index fc3c5aa38754..2e291bc5f1fc 100644 --- a/net/xfrm/xfrm_state.c +++ b/net/xfrm/xfrm_state.c @@ -1383,6 +1383,8 @@ static struct xfrm_state *xfrm_state_clone(struct xfrm_state *orig) x->curlft.add_time = orig->curlft.add_time; x->km.state = orig->km.state; x->km.seq = orig->km.seq; + x->replay = orig->replay; + x->preplay = orig->preplay; return x; -- cgit v1.2.3 From 1b57b6210f4e52904393be97c62122aae69bc8aa Mon Sep 17 00:00:00 2001 From: Arend Van Spriel Date: Tue, 23 May 2017 09:58:07 +0100 Subject: cfg80211: make cfg80211_sched_scan_results() work from atomic context Drivers should be able to call cfg80211_sched_scan_results() from atomic context. However, with the introduction of multiple scheduled scan feature this requirement was not taken into account resulting in regression shown below. [ 119.021594] BUG: scheduling while atomic: irq/47-iwlwifi/517/0x00000200 [ 119.021604] Modules linked in: [...] [ 119.021759] CPU: 1 PID: 517 Comm: irq/47-iwlwifi Not tainted 4.12.0-rc2-t440s-20170522+ #1 [ 119.021763] Hardware name: LENOVO 20AQS03H00/20AQS03H00, BIOS GJET91WW (2.41 ) 09/21/2016 [ 119.021766] Call Trace: [ 119.021778] ? dump_stack+0x5c/0x84 [ 119.021784] ? __schedule_bug+0x4c/0x70 [ 119.021792] ? __schedule+0x496/0x5c0 [ 119.021798] ? schedule+0x2d/0x80 [ 119.021804] ? schedule_preempt_disabled+0x5/0x10 [ 119.021810] ? __mutex_lock.isra.0+0x18e/0x4c0 [ 119.021817] ? __wake_up+0x2f/0x50 [ 119.021833] ? cfg80211_sched_scan_results+0x19/0x60 [cfg80211] [ 119.021844] ? cfg80211_sched_scan_results+0x19/0x60 [cfg80211] [ 119.021859] ? iwl_mvm_rx_lmac_scan_iter_complete_notif+0x17/0x30 [iwlmvm] [ 119.021869] ? iwl_pcie_rx_handle+0x2a9/0x7e0 [iwlwifi] [ 119.021878] ? iwl_pcie_irq_handler+0x17c/0x730 [iwlwifi] [ 119.021884] ? irq_forced_thread_fn+0x60/0x60 [ 119.021887] ? irq_thread_fn+0x16/0x40 [ 119.021892] ? irq_thread+0x109/0x180 [ 119.021896] ? wake_threads_waitq+0x30/0x30 [ 119.021901] ? kthread+0xf2/0x130 [ 119.021905] ? irq_thread_dtor+0x90/0x90 [ 119.021910] ? kthread_create_on_node+0x40/0x40 [ 119.021915] ? ret_from_fork+0x26/0x40 Fixes: b34939b98369 ("cfg80211: add request id to cfg80211_sched_scan_*() api") Reported-by: Sander Eikelenboom Signed-off-by: Arend van Spriel Signed-off-by: Johannes Berg --- net/wireless/scan.c | 8 ++++---- 1 file changed, 4 insertions(+), 4 deletions(-) (limited to 'net') diff --git a/net/wireless/scan.c b/net/wireless/scan.c index 14d5f0c8c45f..9f0901f3e42b 100644 --- a/net/wireless/scan.c +++ b/net/wireless/scan.c @@ -322,9 +322,9 @@ cfg80211_find_sched_scan_req(struct cfg80211_registered_device *rdev, u64 reqid) { struct cfg80211_sched_scan_request *pos; - ASSERT_RTNL(); + WARN_ON_ONCE(!rcu_read_lock_held() && !lockdep_rtnl_is_held()); - list_for_each_entry(pos, &rdev->sched_scan_req_list, list) { + list_for_each_entry_rcu(pos, &rdev->sched_scan_req_list, list) { if (pos->reqid == reqid) return pos; } @@ -398,13 +398,13 @@ void cfg80211_sched_scan_results(struct wiphy *wiphy, u64 reqid) trace_cfg80211_sched_scan_results(wiphy, reqid); /* ignore if we're not scanning */ - rtnl_lock(); + rcu_read_lock(); request = cfg80211_find_sched_scan_req(rdev, reqid); if (request) { request->report_results = true; queue_work(cfg80211_wq, &rdev->sched_scan_res_wk); } - rtnl_unlock(); + rcu_read_unlock(); } EXPORT_SYMBOL(cfg80211_sched_scan_results); -- cgit v1.2.3 From 3ab2137915aea0ce7b3ec02e0f260ecc0f1c289d Mon Sep 17 00:00:00 2001 From: Xin Long Date: Tue, 23 May 2017 13:28:54 +0800 Subject: sctp: fix stream update when processing dupcookie Since commit 3dbcc105d556 ("sctp: alloc stream info when initializing asoc"), stream and stream.out info are always alloced when creating an asoc. So it's not correct to check !asoc->stream before updating stream info when processing dupcookie, but would be better to check asoc state instead. Fixes: 3dbcc105d556 ("sctp: alloc stream info when initializing asoc") Signed-off-by: Xin Long Acked-by: Neil Horman Acked-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/associola.c | 4 +++- 1 file changed, 3 insertions(+), 1 deletion(-) (limited to 'net') diff --git a/net/sctp/associola.c b/net/sctp/associola.c index a9708da28eb5..95238284c422 100644 --- a/net/sctp/associola.c +++ b/net/sctp/associola.c @@ -1176,7 +1176,9 @@ void sctp_assoc_update(struct sctp_association *asoc, asoc->ctsn_ack_point = asoc->next_tsn - 1; asoc->adv_peer_ack_point = asoc->ctsn_ack_point; - if (!asoc->stream) { + + if (sctp_state(asoc, COOKIE_WAIT)) { + sctp_stream_free(asoc->stream); asoc->stream = new->stream; new->stream = NULL; } -- cgit v1.2.3 From 7e06297768886337707f5833942b3bd524a6d3d5 Mon Sep 17 00:00:00 2001 From: Xin Long Date: Tue, 23 May 2017 13:28:55 +0800 Subject: sctp: set new_asoc temp when processing dupcookie After sctp changed to use transport hashtable, a transport would be added into global hashtable when adding the peer to an asoc, then the asoc can be got by searching the transport in the hashtbale. The problem is when processing dupcookie in sctp_sf_do_5_2_4_dupcook, a new asoc would be created. A peer with the same addr and port as the one in the old asoc might be added into the new asoc, but fail to be added into the hashtable, as they also belong to the same sk. It causes that sctp's dupcookie processing can not really work. Since the new asoc will be freed after copying it's information to the old asoc, it's more like a temp asoc. So this patch is to fix it by setting it as a temp asoc to avoid adding it's any transport into the hashtable and also avoid allocing assoc_id. An extra thing it has to do is to also alloc stream info for any temp asoc, as sctp dupcookie process needs it to update old asoc. But I don't think it would hurt something, as a temp asoc would always be freed after finishing processing cookie echo packet. Reported-by: Jianwen Ji Signed-off-by: Xin Long Acked-by: Neil Horman Acked-by: Vlad Yasevich Signed-off-by: David S. Miller --- net/sctp/sm_make_chunk.c | 13 ++++--------- net/sctp/sm_statefuns.c | 3 +++ 2 files changed, 7 insertions(+), 9 deletions(-) (limited to 'net') diff --git a/net/sctp/sm_make_chunk.c b/net/sctp/sm_make_chunk.c index 8a08f13469c4..92e332e17391 100644 --- a/net/sctp/sm_make_chunk.c +++ b/net/sctp/sm_make_chunk.c @@ -2454,16 +2454,11 @@ int sctp_process_init(struct sctp_association *asoc, struct sctp_chunk *chunk, * stream sequence number shall be set to 0. */ - /* Allocate storage for the negotiated streams if it is not a temporary - * association. - */ - if (!asoc->temp) { - if (sctp_stream_init(asoc, gfp)) - goto clean_up; + if (sctp_stream_init(asoc, gfp)) + goto clean_up; - if (sctp_assoc_set_id(asoc, gfp)) - goto clean_up; - } + if (!asoc->temp && sctp_assoc_set_id(asoc, gfp)) + goto clean_up; /* ADDIP Section 4.1 ASCONF Chunk Procedures * diff --git a/net/sctp/sm_statefuns.c b/net/sctp/sm_statefuns.c index 4f5e6cfc7f60..f863b5573e42 100644 --- a/net/sctp/sm_statefuns.c +++ b/net/sctp/sm_statefuns.c @@ -2088,6 +2088,9 @@ sctp_disposition_t sctp_sf_do_5_2_4_dupcook(struct net *net, } } + /* Set temp so that it won't be added into hashtable */ + new_asoc->temp = 1; + /* Compare the tie_tag in cookie with the verification tag of * current association. */ -- cgit v1.2.3 From 0ff50e83b5122e836ca492fefb11656b225ac29c Mon Sep 17 00:00:00 2001 From: Alexander Potapenko Date: Tue, 23 May 2017 13:20:28 +0200 Subject: net: rtnetlink: bail out from rtnl_fdb_dump() on parse error rtnl_fdb_dump() failed to check the result of nlmsg_parse(), which led to contents of |ifm| being uninitialized because nlh->nlmsglen was too small to accommodate |ifm|. The uninitialized data may affect some branches and result in unwanted effects, although kernel data doesn't seem to leak to the userspace directly. The bug has been detected with KMSAN and syzkaller. For the record, here is the KMSAN report: ================================================================== BUG: KMSAN: use of unitialized memory in rtnl_fdb_dump+0x5dc/0x1000 CPU: 0 PID: 1039 Comm: probe Not tainted 4.11.0-rc5+ #2727 Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS Bochs 01/01/2011 Call Trace: __dump_stack lib/dump_stack.c:16 dump_stack+0x143/0x1b0 lib/dump_stack.c:52 kmsan_report+0x12a/0x180 mm/kmsan/kmsan.c:1007 __kmsan_warning_32+0x66/0xb0 mm/kmsan/kmsan_instr.c:491 rtnl_fdb_dump+0x5dc/0x1000 net/core/rtnetlink.c:3230 netlink_dump+0x84f/0x1190 net/netlink/af_netlink.c:2168 __netlink_dump_start+0xc97/0xe50 net/netlink/af_netlink.c:2258 netlink_dump_start ./include/linux/netlink.h:165 rtnetlink_rcv_msg+0xae9/0xb40 net/core/rtnetlink.c:4094 netlink_rcv_skb+0x339/0x5a0 net/netlink/af_netlink.c:2339 rtnetlink_rcv+0x83/0xa0 net/core/rtnetlink.c:4110 netlink_unicast_kernel net/netlink/af_netlink.c:1272 netlink_unicast+0x13b7/0x1480 net/netlink/af_netlink.c:1298 netlink_sendmsg+0x10b8/0x10f0 net/netlink/af_netlink.c:1844 sock_sendmsg_nosec net/socket.c:633 sock_sendmsg net/socket.c:643 ___sys_sendmsg+0xd4b/0x10f0 net/socket.c:1997 __sys_sendmsg net/socket.c:2031 SYSC_sendmsg+0x2c6/0x3f0 net/socket.c:2042 SyS_sendmsg+0x87/0xb0 net/socket.c:2038 do_syscall_64+0x102/0x150 arch/x86/entry/common.c:285 entry_SYSCALL64_slow_path+0x25/0x25 arch/x86/entry/entry_64.S:246 RIP: 0033:0x401300 RSP: 002b:00007ffc3b0e6d58 EFLAGS: 00000246 ORIG_RAX: 000000000000002e RAX: ffffffffffffffda RBX: 00000000004002b0 RCX: 0000000000401300 RDX: 0000000000000000 RSI: 00007ffc3b0e6d80 RDI: 0000000000000003 RBP: 00007ffc3b0e6e00 R08: 000000000000000b R09: 0000000000000004 R10: 000000000000000d R11: 0000000000000246 R12: 0000000000000000 R13: 00000000004065a0 R14: 0000000000406630 R15: 0000000000000000 origin: 000000008fe00056 save_stack_trace+0x59/0x60 arch/x86/kernel/stacktrace.c:59 kmsan_save_stack_with_flags mm/kmsan/kmsan.c:352 kmsan_internal_poison_shadow+0xb1/0x1a0 mm/kmsan/kmsan.c:247 kmsan_poison_shadow+0x6d/0xc0 mm/kmsan/kmsan.c:260 slab_alloc_node mm/slub.c:2743 __kmalloc_node_track_caller+0x1f4/0x390 mm/slub.c:4349 __kmalloc_reserve net/core/skbuff.c:138 __alloc_skb+0x2cd/0x740 net/core/skbuff.c:231 alloc_skb ./include/linux/skbuff.h:933 netlink_alloc_large_skb net/netlink/af_netlink.c:1144 netlink_sendmsg+0x934/0x10f0 net/netlink/af_netlink.c:1819 sock_sendmsg_nosec net/socket.c:633 sock_sendmsg net/socket.c:643 ___sys_sendmsg+0xd4b/0x10f0 net/socket.c:1997 __sys_sendmsg net/socket.c:2031 SYSC_sendmsg+0x2c6/0x3f0 net/socket.c:2042 SyS_sendmsg+0x87/0xb0 net/socket.c:2038 do_syscall_64+0x102/0x150 arch/x86/entry/common.c:285 return_from_SYSCALL_64+0x0/0x6a arch/x86/entry/entry_64.S:246 ================================================================== and the reproducer: ================================================================== #include #include #include #include int main() { int sock = socket(PF_NETLINK, SOCK_DGRAM | SOCK_NONBLOCK, 0); struct msghdr msg; memset(&msg, 0, sizeof(msg)); char nlmsg_buf[32]; memset(nlmsg_buf, 0, sizeof(nlmsg_buf)); struct nlmsghdr *nlmsg = nlmsg_buf; nlmsg->nlmsg_len = 0x11; nlmsg->nlmsg_type = 0x1e; // RTM_NEWROUTE = RTM_BASE + 0x0e // type = 0x0e = 1110b // kind = 2 nlmsg->nlmsg_flags = 0x101; // NLM_F_ROOT | NLM_F_REQUEST nlmsg->nlmsg_seq = 0; nlmsg->nlmsg_pid = 0; nlmsg_buf[16] = (char)7; struct iovec iov; iov.iov_base = nlmsg_buf; iov.iov_len = 17; msg.msg_iov = &iov; msg.msg_iovlen = 1; sendmsg(sock, &msg, 0); return 0; } ================================================================== Signed-off-by: Alexander Potapenko Signed-off-by: David S. Miller --- net/core/rtnetlink.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/core/rtnetlink.c b/net/core/rtnetlink.c index 49a279a7cc15..9e2c0a7cb325 100644 --- a/net/core/rtnetlink.c +++ b/net/core/rtnetlink.c @@ -3231,8 +3231,11 @@ static int rtnl_fdb_dump(struct sk_buff *skb, struct netlink_callback *cb) int err = 0; int fidx = 0; - if (nlmsg_parse(cb->nlh, sizeof(struct ifinfomsg), tb, - IFLA_MAX, ifla_policy, NULL) == 0) { + err = nlmsg_parse(cb->nlh, sizeof(struct ifinfomsg), tb, + IFLA_MAX, ifla_policy, NULL); + if (err < 0) { + return -EINVAL; + } else if (err == 0) { if (tb[IFLA_MASTER]) br_idx = nla_get_u32(tb[IFLA_MASTER]); } -- cgit v1.2.3 From 7c3f1875c66fbc19762760097cabc91849ea0bbb Mon Sep 17 00:00:00 2001 From: Roman Kapl Date: Wed, 24 May 2017 10:22:22 +0200 Subject: net: move somaxconn init from sysctl code The default value for somaxconn is set in sysctl_core_net_init(), but this function is not called when kernel is configured without CONFIG_SYSCTL. This results in the kernel not being able to accept TCP connections, because the backlog has zero size. Usually, the user ends up with: "TCP: request_sock_TCP: Possible SYN flooding on port 7. Dropping request. Check SNMP counters." If SYN cookies are not enabled the connection is rejected. Before ef547f2ac16 (tcp: remove max_qlen_log), the effects were less severe, because the backlog was always at least eight slots long. Signed-off-by: Roman Kapl Signed-off-by: David S. Miller --- net/core/net_namespace.c | 19 +++++++++++++++++++ net/core/sysctl_net_core.c | 2 -- 2 files changed, 19 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/core/net_namespace.c b/net/core/net_namespace.c index 1934efd4a9d4..26bbfababff2 100644 --- a/net/core/net_namespace.c +++ b/net/core/net_namespace.c @@ -315,6 +315,25 @@ out_undo: goto out; } +static int __net_init net_defaults_init_net(struct net *net) +{ + net->core.sysctl_somaxconn = SOMAXCONN; + return 0; +} + +static struct pernet_operations net_defaults_ops = { + .init = net_defaults_init_net, +}; + +static __init int net_defaults_init(void) +{ + if (register_pernet_subsys(&net_defaults_ops)) + panic("Cannot initialize net default settings"); + + return 0; +} + +core_initcall(net_defaults_init); #ifdef CONFIG_NET_NS static struct ucounts *inc_net_namespaces(struct user_namespace *ns) diff --git a/net/core/sysctl_net_core.c b/net/core/sysctl_net_core.c index ea23254b2457..b7cd9aafe99e 100644 --- a/net/core/sysctl_net_core.c +++ b/net/core/sysctl_net_core.c @@ -479,8 +479,6 @@ static __net_init int sysctl_core_net_init(struct net *net) { struct ctl_table *tbl; - net->core.sysctl_somaxconn = SOMAXCONN; - tbl = netns_core_table; if (!net_eq(net, &init_net)) { tbl = kmemdup(tbl, sizeof(netns_core_table), GFP_KERNEL); -- cgit v1.2.3 From ba615f675281d76fd19aa03558777f81fb6b6084 Mon Sep 17 00:00:00 2001 From: Wei Wang Date: Wed, 24 May 2017 09:59:31 -0700 Subject: tcp: avoid fastopen API to be used on AF_UNSPEC Fastopen API should be used to perform fastopen operations on the TCP socket. It does not make sense to use fastopen API to perform disconnect by calling it with AF_UNSPEC. The fastopen data path is also prone to race conditions and bugs when using with AF_UNSPEC. One issue reported and analyzed by Vegard Nossum is as follows: +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Thread A: Thread B: ------------------------------------------------------------------------ sendto() - tcp_sendmsg() - sk_stream_memory_free() = 0 - goto wait_for_sndbuf - sk_stream_wait_memory() - sk_wait_event() // sleep | sendto(flags=MSG_FASTOPEN, dest_addr=AF_UNSPEC) | - tcp_sendmsg() | - tcp_sendmsg_fastopen() | - __inet_stream_connect() | - tcp_disconnect() //because of AF_UNSPEC | - tcp_transmit_skb()// send RST | - return 0; // no reconnect! | - sk_stream_wait_connect() | - sock_error() | - xchg(&sk->sk_err, 0) | - return -ECONNRESET - ... // wake up, see sk->sk_err == 0 - skb_entail() on TCP_CLOSE socket If the connection is reopened then we will send a brand new SYN packet after thread A has already queued a buffer. At this point I think the socket internal state (sequence numbers etc.) becomes messed up. When the new connection is closed, the FIN-ACK is rejected because the sequence number is outside the window. The other side tries to retransmit, but __tcp_retransmit_skb() calls tcp_trim_head() on an empty skb which corrupts the skb data length and hits a BUG() in copy_and_csum_bits(). +++++++++++++++++++++++++++++++++++++++++++++++++++++++++ Hence, this patch adds a check for AF_UNSPEC in the fastopen data path and return EOPNOTSUPP to user if such case happens. Fixes: cf60af03ca4e7 ("tcp: Fast Open client - sendmsg(MSG_FASTOPEN)") Reported-by: Vegard Nossum Signed-off-by: Wei Wang Signed-off-by: Eric Dumazet Signed-off-by: David S. Miller --- net/ipv4/tcp.c | 7 +++++-- 1 file changed, 5 insertions(+), 2 deletions(-) (limited to 'net') diff --git a/net/ipv4/tcp.c b/net/ipv4/tcp.c index 842b575f8fdd..59792d283ff8 100644 --- a/net/ipv4/tcp.c +++ b/net/ipv4/tcp.c @@ -1084,9 +1084,12 @@ static int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, { struct tcp_sock *tp = tcp_sk(sk); struct inet_sock *inet = inet_sk(sk); + struct sockaddr *uaddr = msg->msg_name; int err, flags; - if (!(sysctl_tcp_fastopen & TFO_CLIENT_ENABLE)) + if (!(sysctl_tcp_fastopen & TFO_CLIENT_ENABLE) || + (uaddr && msg->msg_namelen >= sizeof(uaddr->sa_family) && + uaddr->sa_family == AF_UNSPEC)) return -EOPNOTSUPP; if (tp->fastopen_req) return -EALREADY; /* Another Fast Open is in progress */ @@ -1108,7 +1111,7 @@ static int tcp_sendmsg_fastopen(struct sock *sk, struct msghdr *msg, } } flags = (msg->msg_flags & MSG_DONTWAIT) ? O_NONBLOCK : 0; - err = __inet_stream_connect(sk->sk_socket, msg->msg_name, + err = __inet_stream_connect(sk->sk_socket, uaddr, msg->msg_namelen, flags, 1); /* fastopen_req could already be freed in __inet_stream_connect * if the connection times out or gets rst -- cgit v1.2.3 From 5990baaa6d7b437dfcf58b7021ca56b1d6b35869 Mon Sep 17 00:00:00 2001 From: Ihar Hrachyshka Date: Wed, 24 May 2017 15:19:35 -0700 Subject: arp: fixed -Wuninitialized compiler warning Commit 7d472a59c0e5ec117220a05de6b370447fb6cb66 ("arp: always override existing neigh entries with gratuitous ARP") introduced a compiler warning: net/ipv4/arp.c:880:35: warning: 'addr_type' may be used uninitialized in this function [-Wmaybe-uninitialized] While the code logic seems to be correct and doesn't allow the variable to be used uninitialized, and the warning is not consistently reproducible, it's still worth fixing it for other people not to waste time looking at the warning in case it pops up in the build environment. Yes, compiler is probably at fault, but we will need to accommodate. Fixes: 7d472a59c0e5 ("arp: always override existing neigh entries with gratuitous ARP") Signed-off-by: Ihar Hrachyshka Signed-off-by: David S. Miller --- net/ipv4/arp.c | 2 +- 1 file changed, 1 insertion(+), 1 deletion(-) (limited to 'net') diff --git a/net/ipv4/arp.c b/net/ipv4/arp.c index ae96e6f3e0cb..e9f3386a528b 100644 --- a/net/ipv4/arp.c +++ b/net/ipv4/arp.c @@ -863,8 +863,8 @@ static int arp_process(struct net *net, struct sock *sk, struct sk_buff *skb) n = __neigh_lookup(&arp_tbl, &sip, dev, 0); + addr_type = -1; if (n || IN_DEV_ARP_ACCEPT(in_dev)) { - addr_type = -1; is_garp = arp_is_garp(net, dev, &addr_type, arp->ar_op, sip, tip, sha, tha); } -- cgit v1.2.3 From 41703a731066fde79c3e5ccf3391cf77a98aeda5 Mon Sep 17 00:00:00 2001 From: Daniel Borkmann Date: Thu, 25 May 2017 01:05:07 +0200 Subject: bpf: add bpf_clone_redirect to bpf_helper_changes_pkt_data The bpf_clone_redirect() still needs to be listed in bpf_helper_changes_pkt_data() since we call into bpf_try_make_head_writable() from there, thus we need to invalidate prior pkt regs as well. Fixes: 36bbef52c7eb ("bpf: direct packet write and access for helpers for clsact progs") Signed-off-by: Daniel Borkmann Acked-by: Alexei Starovoitov Signed-off-by: David S. Miller --- net/core/filter.c | 1 + 1 file changed, 1 insertion(+) (limited to 'net') diff --git a/net/core/filter.c b/net/core/filter.c index a253a6197e6b..a6bb95fa87b2 100644 --- a/net/core/filter.c +++ b/net/core/filter.c @@ -2281,6 +2281,7 @@ bool bpf_helper_changes_pkt_data(void *func) func == bpf_skb_change_head || func == bpf_skb_change_tail || func == bpf_skb_pull_data || + func == bpf_clone_redirect || func == bpf_l3_csum_replace || func == bpf_l4_csum_replace || func == bpf_xdp_adjust_head) -- cgit v1.2.3 From 0908cf4dfef35fc6ac12329007052ebe93ff1081 Mon Sep 17 00:00:00 2001 From: linzhang Date: Thu, 25 May 2017 14:07:18 +0800 Subject: net: llc: add lock_sock in llc_ui_bind to avoid a race condition There is a race condition in llc_ui_bind if two or more processes/threads try to bind a same socket. If more processes/threads bind a same socket success that will lead to two problems, one is this action is not what we expected, another is will lead to kernel in unstable status or oops(in my simple test case, cause llc2.ko can't unload). The current code is test SOCK_ZAPPED bit to avoid a process to bind a same socket twice but that is can't avoid more processes/threads try to bind a same socket at the same time. So, add lock_sock in llc_ui_bind like others, such as llc_ui_connect. Signed-off-by: Lin Zhang Signed-off-by: David S. Miller --- net/llc/af_llc.c | 3 +++ 1 file changed, 3 insertions(+) (limited to 'net') diff --git a/net/llc/af_llc.c b/net/llc/af_llc.c index 8364fe5b59e4..c38d16f22d2a 100644 --- a/net/llc/af_llc.c +++ b/net/llc/af_llc.c @@ -311,6 +311,8 @@ static int llc_ui_bind(struct socket *sock, struct sockaddr *uaddr, int addrlen) int rc = -EINVAL; dprintk("%s: binding %02X\n", __func__, addr->sllc_sap); + + lock_sock(sk); if (unlikely(!sock_flag(sk, SOCK_ZAPPED) || addrlen != sizeof(*addr))) goto out; rc = -EAFNOSUPPORT; @@ -382,6 +384,7 @@ static int llc_ui_bind(struct socket *sock, struct sockaddr *uaddr, int addrlen) out_put: llc_sap_put(sap); out: + release_sock(sk); return rc; } -- cgit v1.2.3 From 804ec7ebe8ea003999ca8d1bfc499edc6a9e07df Mon Sep 17 00:00:00 2001 From: Davide Caratti Date: Thu, 25 May 2017 19:14:56 +0200 Subject: sctp: fix ICMP processing if skb is non-linear sometimes ICMP replies to INIT chunks are ignored by the client, even if the encapsulated SCTP headers match an open socket. This happens when the ICMP packet is carried by a paged skb: use skb_header_pointer() to read packet contents beyond the SCTP header, so that chunk header and initiate tag are validated correctly. v2: - don't use skb_header_pointer() to read the transport header, since icmp_socket_deliver() already puts these 8 bytes in the linear area. - change commit message to make specific reference to INIT chunks. Signed-off-by: Davide Caratti Acked-by: Marcelo Ricardo Leitner Acked-by: Vlad Yasevich Reviewed-by: Xin Long Signed-off-by: David S. Miller --- net/sctp/input.c | 16 +++++++++------- 1 file changed, 9 insertions(+), 7 deletions(-) (limited to 'net') diff --git a/net/sctp/input.c b/net/sctp/input.c index 0e06a278d2a9..ba9ad32fc447 100644 --- a/net/sctp/input.c +++ b/net/sctp/input.c @@ -473,15 +473,14 @@ struct sock *sctp_err_lookup(struct net *net, int family, struct sk_buff *skb, struct sctp_association **app, struct sctp_transport **tpp) { + struct sctp_init_chunk *chunkhdr, _chunkhdr; union sctp_addr saddr; union sctp_addr daddr; struct sctp_af *af; struct sock *sk = NULL; struct sctp_association *asoc; struct sctp_transport *transport = NULL; - struct sctp_init_chunk *chunkhdr; __u32 vtag = ntohl(sctphdr->vtag); - int len = skb->len - ((void *)sctphdr - (void *)skb->data); *app = NULL; *tpp = NULL; @@ -516,13 +515,16 @@ struct sock *sctp_err_lookup(struct net *net, int family, struct sk_buff *skb, * discard the packet. */ if (vtag == 0) { - chunkhdr = (void *)sctphdr + sizeof(struct sctphdr); - if (len < sizeof(struct sctphdr) + sizeof(sctp_chunkhdr_t) - + sizeof(__be32) || + /* chunk header + first 4 octects of init header */ + chunkhdr = skb_header_pointer(skb, skb_transport_offset(skb) + + sizeof(struct sctphdr), + sizeof(struct sctp_chunkhdr) + + sizeof(__be32), &_chunkhdr); + if (!chunkhdr || chunkhdr->chunk_hdr.type != SCTP_CID_INIT || - ntohl(chunkhdr->init_hdr.init_tag) != asoc->c.my_vtag) { + ntohl(chunkhdr->init_hdr.init_tag) != asoc->c.my_vtag) goto out; - } + } else if (vtag != asoc->c.peer_vtag) { goto out; } -- cgit v1.2.3 From 0e9a709560dbcfbace8bf4019dc5298619235891 Mon Sep 17 00:00:00 2001 From: Peter Dawson Date: Fri, 26 May 2017 06:35:18 +1000 Subject: ip6_tunnel, ip6_gre: fix setting of DSCP on encapsulated packets This fix addresses two problems in the way the DSCP field is formulated on the encapsulating header of IPv6 tunnels. Bugzilla: https://bugzilla.kernel.org/show_bug.cgi?id=195661 1) The IPv6 tunneling code was manipulating the DSCP field of the encapsulating packet using the 32b flowlabel. Since the flowlabel is only the lower 20b it was incorrect to assume that the upper 12b containing the DSCP and ECN fields would remain intact when formulating the encapsulating header. This fix handles the 'inherit' and 'fixed-value' DSCP cases explicitly using the extant dsfield u8 variable. 2) The use of INET_ECN_encapsulate(0, dsfield) in ip6_tnl_xmit was incorrect and resulted in the DSCP value always being set to 0. Commit 90427ef5d2a4 ("ipv6: fix flow labels when the traffic class is non-0") caused the regression by masking out the flowlabel which exposed the incorrect handling of the DSCP portion of the flowlabel in ip6_tunnel and ip6_gre. Fixes: 90427ef5d2a4 ("ipv6: fix flow labels when the traffic class is non-0") Signed-off-by: Peter Dawson Signed-off-by: David S. Miller --- net/ipv6/ip6_gre.c | 13 +++++++------ net/ipv6/ip6_tunnel.c | 21 +++++++++++++-------- 2 files changed, 20 insertions(+), 14 deletions(-) (limited to 'net') diff --git a/net/ipv6/ip6_gre.c b/net/ipv6/ip6_gre.c index 8d128ba79b66..0c5b4caa1949 100644 --- a/net/ipv6/ip6_gre.c +++ b/net/ipv6/ip6_gre.c @@ -537,11 +537,10 @@ static inline int ip6gre_xmit_ipv4(struct sk_buff *skb, struct net_device *dev) memcpy(&fl6, &t->fl.u.ip6, sizeof(fl6)); - dsfield = ipv4_get_dsfield(iph); - if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS) - fl6.flowlabel |= htonl((__u32)iph->tos << IPV6_TCLASS_SHIFT) - & IPV6_TCLASS_MASK; + dsfield = ipv4_get_dsfield(iph); + else + dsfield = ip6_tclass(t->parms.flowinfo); if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK) fl6.flowi6_mark = skb->mark; else @@ -598,9 +597,11 @@ static inline int ip6gre_xmit_ipv6(struct sk_buff *skb, struct net_device *dev) memcpy(&fl6, &t->fl.u.ip6, sizeof(fl6)); - dsfield = ipv6_get_dsfield(ipv6h); if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS) - fl6.flowlabel |= (*(__be32 *) ipv6h & IPV6_TCLASS_MASK); + dsfield = ipv6_get_dsfield(ipv6h); + else + dsfield = ip6_tclass(t->parms.flowinfo); + if (t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL) fl6.flowlabel |= ip6_flowlabel(ipv6h); if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK) diff --git a/net/ipv6/ip6_tunnel.c b/net/ipv6/ip6_tunnel.c index 6eb2ae507500..7ae6c503f1ca 100644 --- a/net/ipv6/ip6_tunnel.c +++ b/net/ipv6/ip6_tunnel.c @@ -1196,7 +1196,7 @@ route_lookup: skb_push(skb, sizeof(struct ipv6hdr)); skb_reset_network_header(skb); ipv6h = ipv6_hdr(skb); - ip6_flow_hdr(ipv6h, INET_ECN_encapsulate(0, dsfield), + ip6_flow_hdr(ipv6h, dsfield, ip6_make_flowlabel(net, skb, fl6->flowlabel, true, fl6)); ipv6h->hop_limit = hop_limit; ipv6h->nexthdr = proto; @@ -1231,8 +1231,6 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) if (tproto != IPPROTO_IPIP && tproto != 0) return -1; - dsfield = ipv4_get_dsfield(iph); - if (t->parms.collect_md) { struct ip_tunnel_info *tun_info; const struct ip_tunnel_key *key; @@ -1246,6 +1244,7 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) fl6.flowi6_proto = IPPROTO_IPIP; fl6.daddr = key->u.ipv6.dst; fl6.flowlabel = key->label; + dsfield = ip6_tclass(key->label); } else { if (!(t->parms.flags & IP6_TNL_F_IGN_ENCAP_LIMIT)) encap_limit = t->parms.encap_limit; @@ -1254,8 +1253,9 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) fl6.flowi6_proto = IPPROTO_IPIP; if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS) - fl6.flowlabel |= htonl((__u32)iph->tos << IPV6_TCLASS_SHIFT) - & IPV6_TCLASS_MASK; + dsfield = ipv4_get_dsfield(iph); + else + dsfield = ip6_tclass(t->parms.flowinfo); if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK) fl6.flowi6_mark = skb->mark; else @@ -1267,6 +1267,8 @@ ip4ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) if (iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6)) return -1; + dsfield = INET_ECN_encapsulate(dsfield, ipv4_get_dsfield(iph)); + skb_set_inner_ipproto(skb, IPPROTO_IPIP); err = ip6_tnl_xmit(skb, dev, dsfield, &fl6, encap_limit, &mtu, @@ -1300,8 +1302,6 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) ip6_tnl_addr_conflict(t, ipv6h)) return -1; - dsfield = ipv6_get_dsfield(ipv6h); - if (t->parms.collect_md) { struct ip_tunnel_info *tun_info; const struct ip_tunnel_key *key; @@ -1315,6 +1315,7 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) fl6.flowi6_proto = IPPROTO_IPV6; fl6.daddr = key->u.ipv6.dst; fl6.flowlabel = key->label; + dsfield = ip6_tclass(key->label); } else { offset = ip6_tnl_parse_tlv_enc_lim(skb, skb_network_header(skb)); /* ip6_tnl_parse_tlv_enc_lim() might have reallocated skb->head */ @@ -1337,7 +1338,9 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) fl6.flowi6_proto = IPPROTO_IPV6; if (t->parms.flags & IP6_TNL_F_USE_ORIG_TCLASS) - fl6.flowlabel |= (*(__be32 *)ipv6h & IPV6_TCLASS_MASK); + dsfield = ipv6_get_dsfield(ipv6h); + else + dsfield = ip6_tclass(t->parms.flowinfo); if (t->parms.flags & IP6_TNL_F_USE_ORIG_FLOWLABEL) fl6.flowlabel |= ip6_flowlabel(ipv6h); if (t->parms.flags & IP6_TNL_F_USE_ORIG_FWMARK) @@ -1351,6 +1354,8 @@ ip6ip6_tnl_xmit(struct sk_buff *skb, struct net_device *dev) if (iptunnel_handle_offloads(skb, SKB_GSO_IPXIP6)) return -1; + dsfield = INET_ECN_encapsulate(dsfield, ipv6_get_dsfield(ipv6h)); + skb_set_inner_ipproto(skb, IPPROTO_IPV6); err = ip6_tnl_xmit(skb, dev, dsfield, &fl6, encap_limit, &mtu, -- cgit v1.2.3 From 3fb07daff8e99243366a081e5129560734de4ada Mon Sep 17 00:00:00 2001 From: Eric Dumazet Date: Thu, 25 May 2017 14:27:35 -0700 Subject: ipv4: add reference counting to metrics Andrey Konovalov reported crashes in ipv4_mtu() I could reproduce the issue with KASAN kernels, between 10.246.7.151 and 10.246.7.152 : 1) 20 concurrent netperf -t TCP_RR -H 10.246.7.152 -l 1000 & 2) At the same time run following loop : while : do ip ro add 10.246.7.152 dev eth0 src 10.246.7.151 mtu 1500 ip ro del 10.246.7.152 dev eth0 src 10.246.7.151 mtu 1500 done Cong Wang attempted to add back rt->fi in commit 82486aa6f1b9 ("ipv4: restore rt->fi for reference counting") but this proved to add some issues that were complex to solve. Instead, I suggested to add a refcount to the metrics themselves, being a standalone object (in particular, no reference to other objects) I tried to make this patch as small as possible to ease its backport, instead of being super clean. Note that we believe that only ipv4 dst need to take care of the metric refcount. But if this is wrong, this patch adds the basic infrastructure to extend this to other families. Many thanks to Julian Anastasov for reviewing this patch, and Cong Wang for his efforts on this problem. Fixes: 2860583fe840 ("ipv4: Kill rt->fi") Signed-off-by: Eric Dumazet Reported-by: Andrey Konovalov Reviewed-by: Julian Anastasov Acked-by: Cong Wang Signed-off-by: David S. Miller --- include/net/dst.h | 8 +++++++- include/net/ip_fib.h | 10 +++++----- net/core/dst.c | 23 ++++++++++++++--------- net/ipv4/fib_semantics.c | 17 ++++++++++------- net/ipv4/route.c | 10 +++++++++- 5 files changed, 45 insertions(+), 23 deletions(-) (limited to 'net') diff --git a/include/net/dst.h b/include/net/dst.h index 049af33da3b6..cfc043784166 100644 --- a/include/net/dst.h +++ b/include/net/dst.h @@ -107,10 +107,16 @@ struct dst_entry { }; }; +struct dst_metrics { + u32 metrics[RTAX_MAX]; + atomic_t refcnt; +}; +extern const struct dst_metrics dst_default_metrics; + u32 *dst_cow_metrics_generic(struct dst_entry *dst, unsigned long old); -extern const u32 dst_default_metrics[]; #define DST_METRICS_READ_ONLY 0x1UL +#define DST_METRICS_REFCOUNTED 0x2UL #define DST_METRICS_FLAGS 0x3UL #define __DST_METRICS_PTR(Y) \ ((u32 *)((Y) & ~DST_METRICS_FLAGS)) diff --git a/include/net/ip_fib.h b/include/net/ip_fib.h index 6692c5758b33..f7f6aa789c61 100644 --- a/include/net/ip_fib.h +++ b/include/net/ip_fib.h @@ -114,11 +114,11 @@ struct fib_info { __be32 fib_prefsrc; u32 fib_tb_id; u32 fib_priority; - u32 *fib_metrics; -#define fib_mtu fib_metrics[RTAX_MTU-1] -#define fib_window fib_metrics[RTAX_WINDOW-1] -#define fib_rtt fib_metrics[RTAX_RTT-1] -#define fib_advmss fib_metrics[RTAX_ADVMSS-1] + struct dst_metrics *fib_metrics; +#define fib_mtu fib_metrics->metrics[RTAX_MTU-1] +#define fib_window fib_metrics->metrics[RTAX_WINDOW-1] +#define fib_rtt fib_metrics->metrics[RTAX_RTT-1] +#define fib_advmss fib_metrics->metrics[RTAX_ADVMSS-1] int fib_nhs; #ifdef CONFIG_IP_ROUTE_MULTIPATH int fib_weight; diff --git a/net/core/dst.c b/net/core/dst.c index 960e503b5a52..6192f11beec9 100644 --- a/net/core/dst.c +++ b/net/core/dst.c @@ -151,13 +151,13 @@ int dst_discard_out(struct net *net, struct sock *sk, struct sk_buff *skb) } EXPORT_SYMBOL(dst_discard_out); -const u32 dst_default_metrics[RTAX_MAX + 1] = { +const struct dst_metrics dst_default_metrics = { /* This initializer is needed to force linker to place this variable * into const section. Otherwise it might end into bss section. * We really want to avoid false sharing on this variable, and catch * any writes on it. */ - [RTAX_MAX] = 0xdeadbeef, + .refcnt = ATOMIC_INIT(1), }; void dst_init(struct dst_entry *dst, struct dst_ops *ops, @@ -169,7 +169,7 @@ void dst_init(struct dst_entry *dst, struct dst_ops *ops, if (dev) dev_hold(dev); dst->ops = ops; - dst_init_metrics(dst, dst_default_metrics, true); + dst_init_metrics(dst, dst_default_metrics.metrics, true); dst->expires = 0UL; dst->path = dst; dst->from = NULL; @@ -314,25 +314,30 @@ EXPORT_SYMBOL(dst_release); u32 *dst_cow_metrics_generic(struct dst_entry *dst, unsigned long old) { - u32 *p = kmalloc(sizeof(u32) * RTAX_MAX, GFP_ATOMIC); + struct dst_metrics *p = kmalloc(sizeof(*p), GFP_ATOMIC); if (p) { - u32 *old_p = __DST_METRICS_PTR(old); + struct dst_metrics *old_p = (struct dst_metrics *)__DST_METRICS_PTR(old); unsigned long prev, new; - memcpy(p, old_p, sizeof(u32) * RTAX_MAX); + atomic_set(&p->refcnt, 1); + memcpy(p->metrics, old_p->metrics, sizeof(p->metrics)); new = (unsigned long) p; prev = cmpxchg(&dst->_metrics, old, new); if (prev != old) { kfree(p); - p = __DST_METRICS_PTR(prev); + p = (struct dst_metrics *)__DST_METRICS_PTR(prev); if (prev & DST_METRICS_READ_ONLY) p = NULL; + } else if (prev & DST_METRICS_REFCOUNTED) { + if (atomic_dec_and_test(&old_p->refcnt)) + kfree(old_p); } } - return p; + BUILD_BUG_ON(offsetof(struct dst_metrics, metrics) != 0); + return (u32 *)p; } EXPORT_SYMBOL(dst_cow_metrics_generic); @@ -341,7 +346,7 @@ void __dst_destroy_metrics_generic(struct dst_entry *dst, unsigned long old) { unsigned long prev, new; - new = ((unsigned long) dst_default_metrics) | DST_METRICS_READ_ONLY; + new = ((unsigned long) &dst_default_metrics) | DST_METRICS_READ_ONLY; prev = cmpxchg(&dst->_metrics, old, new); if (prev == old) kfree(__DST_METRICS_PTR(old)); diff --git a/net/ipv4/fib_semantics.c b/net/ipv4/fib_semantics.c index da449ddb8cc1..ad9ad4aab5da 100644 --- a/net/ipv4/fib_semantics.c +++ b/net/ipv4/fib_semantics.c @@ -203,6 +203,7 @@ static void rt_fibinfo_free_cpus(struct rtable __rcu * __percpu *rtp) static void free_fib_info_rcu(struct rcu_head *head) { struct fib_info *fi = container_of(head, struct fib_info, rcu); + struct dst_metrics *m; change_nexthops(fi) { if (nexthop_nh->nh_dev) @@ -213,8 +214,9 @@ static void free_fib_info_rcu(struct rcu_head *head) rt_fibinfo_free(&nexthop_nh->nh_rth_input); } endfor_nexthops(fi); - if (fi->fib_metrics != (u32 *) dst_default_metrics) - kfree(fi->fib_metrics); + m = fi->fib_metrics; + if (m != &dst_default_metrics && atomic_dec_and_test(&m->refcnt)) + kfree(m); kfree(fi); } @@ -971,11 +973,11 @@ fib_convert_metrics(struct fib_info *fi, const struct fib_config *cfg) val = 255; if (type == RTAX_FEATURES && (val & ~RTAX_FEATURE_MASK)) return -EINVAL; - fi->fib_metrics[type - 1] = val; + fi->fib_metrics->metrics[type - 1] = val; } if (ecn_ca) - fi->fib_metrics[RTAX_FEATURES - 1] |= DST_FEATURE_ECN_CA; + fi->fib_metrics->metrics[RTAX_FEATURES - 1] |= DST_FEATURE_ECN_CA; return 0; } @@ -1033,11 +1035,12 @@ struct fib_info *fib_create_info(struct fib_config *cfg) goto failure; fib_info_cnt++; if (cfg->fc_mx) { - fi->fib_metrics = kzalloc(sizeof(u32) * RTAX_MAX, GFP_KERNEL); + fi->fib_metrics = kzalloc(sizeof(*fi->fib_metrics), GFP_KERNEL); if (!fi->fib_metrics) goto failure; + atomic_set(&fi->fib_metrics->refcnt, 1); } else - fi->fib_metrics = (u32 *) dst_default_metrics; + fi->fib_metrics = (struct dst_metrics *)&dst_default_metrics; fi->fib_net = net; fi->fib_protocol = cfg->fc_protocol; @@ -1238,7 +1241,7 @@ int fib_dump_info(struct sk_buff *skb, u32 portid, u32 seq, int event, if (fi->fib_priority && nla_put_u32(skb, RTA_PRIORITY, fi->fib_priority)) goto nla_put_failure; - if (rtnetlink_put_metrics(skb, fi->fib_metrics) < 0) + if (rtnetlink_put_metrics(skb, fi->fib_metrics->metrics) < 0) goto nla_put_failure; if (fi->fib_prefsrc && diff --git a/net/ipv4/route.c b/net/ipv4/route.c index 655d9eebe43e..6883b3d4ba8f 100644 --- a/net/ipv4/route.c +++ b/net/ipv4/route.c @@ -1385,8 +1385,12 @@ static void rt_add_uncached_list(struct rtable *rt) static void ipv4_dst_destroy(struct dst_entry *dst) { + struct dst_metrics *p = (struct dst_metrics *)DST_METRICS_PTR(dst); struct rtable *rt = (struct rtable *) dst; + if (p != &dst_default_metrics && atomic_dec_and_test(&p->refcnt)) + kfree(p); + if (!list_empty(&rt->rt_uncached)) { struct uncached_list *ul = rt->rt_uncached_list; @@ -1438,7 +1442,11 @@ static void rt_set_nexthop(struct rtable *rt, __be32 daddr, rt->rt_gateway = nh->nh_gw; rt->rt_uses_gateway = 1; } - dst_init_metrics(&rt->dst, fi->fib_metrics, true); + dst_init_metrics(&rt->dst, fi->fib_metrics->metrics, true); + if (fi->fib_metrics != &dst_default_metrics) { + rt->dst._metrics |= DST_METRICS_REFCOUNTED; + atomic_inc(&fi->fib_metrics->refcnt); + } #ifdef CONFIG_IP_ROUTE_CLASSID rt->dst.tclassid = nh->nh_tclassid; #endif -- cgit v1.2.3