linux-stable.git/net/core, branch v5.1.9

neighbor: Call __ipv4_neigh_lookup_noref in neigh_xmit

2019-06-11T10:19:13+00:00

[ Upstream commit 4b2a2bfeb3f056461a90bd621e8bd7d03fa47f60 ]

Commit cd9ff4de0107 changed the key for IFF_POINTOPOINT devices to
INADDR_ANY but neigh_xmit which is used for MPLS encapsulations was not
updated to use the altered key. The result is that every packet Tx does
a lookup on the gateway address which does not find an entry, a new one
is created only to find the existing one in the table right before the
insert since arp_constructor was updated to reset the primary key. This
is seen in the allocs and destroys counters:
    ip -s -4 ntable show | head -10 | grep alloc

which increase for each packet showing the unnecessary overhread.

Fix by having neigh_xmit use __ipv4_neigh_lookup_noref for NEIGH_ARP_TABLE.

Fixes: cd9ff4de0107 ("ipv4: Make neigh lookup keys for loopback/point-to-point devices be INADDR_ANY")
Reported-by: Alan Maguire 
Signed-off-by: David Ahern 
Tested-by: Alan Maguire 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

neighbor: Reset gc_entries counter if new entry is released before insert

2019-06-11T10:19:12+00:00

[ Upstream commit 64c6f4bbca748c3b2101469a76d88b7cd1c00476 ]

Ian and Alan both reported seeing overflows after upgrades to 5.x kernels:
  neighbour: arp_cache: neighbor table overflow!

Alan's mpls script helped get to the bottom of this bug. When a new entry
is created the gc_entries counter is bumped in neigh_alloc to check if a
new one is allowed to be created. ___neigh_create then searches for an
existing entry before inserting the just allocated one. If an entry
already exists, the new one is dropped in favor of the existing one. In
this case the cleanup path needs to drop the gc_entries counter. There
is no memory leak, only a counter leak.

Fixes: 58956317c8d ("neighbor: Improve garbage collection")
Reported-by: Ian Kumlien 
Reported-by: Alan Maguire 
Signed-off-by: David Ahern 
Tested-by: Alan Maguire 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

Revert "fib_rules: return 0 directly if an exactly same rule exists when NLM_F_EXCL not supplied"

2019-06-11T10:19:12+00:00

[ Upstream commit 4970b42d5c362bf873982db7d93245c5281e58f4 ]

This reverts commit e9919a24d3022f72bcadc407e73a6ef17093a849.

Nathan reported the new behaviour breaks Android, as Android just add
new rules and delete old ones.

If we return 0 without adding dup rules, Android will remove the new
added rules and causing system to soft-reboot.

Fixes: e9919a24d302 ("fib_rules: return 0 directly if an exactly same rule exists when NLM_F_EXCL not supplied")
Reported-by: Nathan Chancellor 
Reported-by: Yaro Slav 
Reported-by: Maciej Żenczykowski 
Signed-off-by: Hangbin Liu 
Reviewed-by: Nathan Chancellor 
Tested-by: Nathan Chancellor 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

pktgen: do not sleep with the thread lock held.

2019-06-11T10:19:11+00:00

[ Upstream commit 720f1de4021f09898b8c8443f3b3e995991b6e3a ]

Currently, the process issuing a "start" command on the pktgen procfs
interface, acquires the pktgen thread lock and never release it, until
all pktgen threads are completed. The above can blocks indefinitely any
other pktgen command and any (even unrelated) netdevice removal - as
the pktgen netdev notifier acquires the same lock.

The issue is demonstrated by the following script, reported by Matteo:

ip -b - <<'EOF'
	link add type dummy
	link add type veth
	link set dummy0 up
EOF
modprobe pktgen
echo reset >/proc/net/pktgen/pgctrl
{
	echo rem_device_all
	echo add_device dummy0
} >/proc/net/pktgen/kpktgend_0
echo count 0 >/proc/net/pktgen/dummy0
echo start >/proc/net/pktgen/pgctrl &
sleep 1
rmmod veth

Fix the above releasing the thread lock around the sleep call.

Additionally we must prevent racing with forcefull rmmod - as the
thread lock no more protects from them. Instead, acquire a self-reference
before waiting for any thread. As a side effect, running

rmmod pktgen

while some thread is running now fails with "module in use" error,
before this patch such command hanged indefinitely.

Note: the issue predates the commit reported in the fixes tag, but
this fix can't be applied before the mentioned commit.

v1 -> v2:
 - no need to check for thread existence after flipping the lock,
   pktgen threads are freed only at net exit time
 -

Fixes: 6146e6a43b35 ("[PKTGEN]: Removes thread_{un,}lock() macros.")
Reported-and-tested-by: Matteo Croce 
Signed-off-by: Paolo Abeni 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

ethtool: fix potential userspace buffer overflow

2019-06-11T10:19:09+00:00

[ Upstream commit 0ee4e76937d69128a6a66861ba393ebdc2ffc8a2 ]

ethtool_get_regs() allocates a buffer of size ops->get_regs_len(),
and pass it to the kernel driver via ops->get_regs() for filling.

There is no restriction about what the kernel drivers can or cannot do
with the open ethtool_regs structure. They usually set regs->version
and ignore regs->len or set it to the same size as ops->get_regs_len().

But if userspace allocates a smaller buffer for the registers dump,
we would cause a userspace buffer overflow in the final copy_to_user()
call, which uses the regs.len value potentially reset by the driver.

To fix this, make this case obvious and store regs.len before calling
ops->get_regs(), to only copy as much data as requested by userspace,
up to the value returned by ops->get_regs_len().

While at it, remove the redundant check for non-null regbuf.

Signed-off-by: Vivien Didelot 
Reviewed-by: Michal Kubecek 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net: correct zerocopy refcnt with udp MSG_MORE

2019-06-04T05:59:45+00:00

[ Upstream commit 100f6d8e09905c59be45b6316f8f369c0be1b2d8 ]

TCP zerocopy takes a uarg reference for every skb, plus one for the
tcp_sendmsg_locked datapath temporarily, to avoid reaching refcnt zero
as it builds, sends and frees skbs inside its inner loop.

UDP and RAW zerocopy do not send inside the inner loop so do not need
the extra sock_zerocopy_get + sock_zerocopy_put pair. Commit
52900d22288ed ("udp: elide zerocopy operation in hot path") introduced
extra_uref to pass the initial reference taken in sock_zerocopy_alloc
to the first generated skb.

But, sock_zerocopy_realloc takes this extra reference at the start of
every call. With MSG_MORE, no new skb may be generated to attach the
extra_uref to, so refcnt is incorrectly 2 with only one skb.

Do not take the extra ref if uarg && !tcp, which implies MSG_MORE.
Update extra_uref accordingly.

This conditional assignment triggers a false positive may be used
uninitialized warning, so have to initialize extra_uref at define.

Changes v1->v2: fix typo in Fixes SHA1

Fixes: 52900d22288e7 ("udp: elide zerocopy operation in hot path")
Reported-by: syzbot 
Diagnosed-by: Eric Dumazet 
Signed-off-by: Willem de Bruijn 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

net-gro: fix use-after-free read in napi_gro_frags()

2019-06-04T05:59:41+00:00

[ Upstream commit a4270d6795b0580287453ea55974d948393e66ef ]

If a network driver provides to napi_gro_frags() an
skb with a page fragment of exactly 14 bytes, the call
to gro_pull_from_frag0() will 'consume' the fragment
by calling skb_frag_unref(skb, 0), and the page might
be freed and reused.

Reading eth->h_proto at the end of napi_frags_skb() might
read mangled data, or crash under specific debugging features.

BUG: KASAN: use-after-free in napi_frags_skb net/core/dev.c:5833 [inline]
BUG: KASAN: use-after-free in napi_gro_frags+0xc6f/0xd10 net/core/dev.c:5841
Read of size 2 at addr ffff88809366840c by task syz-executor599/8957

CPU: 1 PID: 8957 Comm: syz-executor599 Not tainted 5.2.0-rc1+ #32
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 __dump_stack lib/dump_stack.c:77 [inline]
 dump_stack+0x172/0x1f0 lib/dump_stack.c:113
 print_address_description.cold+0x7c/0x20d mm/kasan/report.c:188
 __kasan_report.cold+0x1b/0x40 mm/kasan/report.c:317
 kasan_report+0x12/0x20 mm/kasan/common.c:614
 __asan_report_load_n_noabort+0xf/0x20 mm/kasan/generic_report.c:142
 napi_frags_skb net/core/dev.c:5833 [inline]
 napi_gro_frags+0xc6f/0xd10 net/core/dev.c:5841
 tun_get_user+0x2f3c/0x3ff0 drivers/net/tun.c:1991
 tun_chr_write_iter+0xbd/0x156 drivers/net/tun.c:2037
 call_write_iter include/linux/fs.h:1872 [inline]
 do_iter_readv_writev+0x5f8/0x8f0 fs/read_write.c:693
 do_iter_write fs/read_write.c:970 [inline]
 do_iter_write+0x184/0x610 fs/read_write.c:951
 vfs_writev+0x1b3/0x2f0 fs/read_write.c:1015
 do_writev+0x15b/0x330 fs/read_write.c:1058

Fixes: a50e233c50db ("net-gro: restore frag0 optimization")
Signed-off-by: Eric Dumazet 
Reported-by: syzbot 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

ethtool: Check for vlan etype or vlan tci when parsing flow_rule

2019-06-04T05:59:40+00:00

[ Upstream commit b73484b2fc0d0ba84a13e9d86eb4adcae718163b ]

When parsing an ethtool flow spec to build a flow_rule, the code checks
if both the vlan etype and the vlan tci are specified by the user to add
a FLOW_DISSECTOR_KEY_VLAN match.

However, when the user only specified a vlan etype or a vlan tci, this
check silently ignores these parameters.

For example, the following rule :

ethtool -N eth0 flow-type udp4 vlan 0x0010 action -1 loc 0

will result in no error being issued, but the equivalent rule will be
created and passed to the NIC driver :

ethtool -N eth0 flow-type udp4 action -1 loc 0

In the end, neither the NIC driver using the rule nor the end user have
a way to know that these keys were dropped along the way, or that
incorrect parameters were entered.

This kind of check should be left to either the driver, or the ethtool
flow spec layer.

This commit makes so that ethtool parameters are forwarded as-is to the
NIC driver.

Since none of the users of ethtool_rx_flow_rule_create are using the
VLAN dissector, I don't think this qualifies as a regression.

Fixes: eca4205f9ec3 ("ethtool: add ethtool_rx_flow_spec to flow_rule structure translator")
Signed-off-by: Maxime Chevallier 
Acked-by: Pablo Neira Ayuso 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

flow_offload: support CVLAN match

2019-05-25T16:16:15+00:00

[ Upstream commit bae9ed69029c7d499c57485593b2faae475fd704 ]

Plumb it through from the flow_dissector.

Signed-off-by: Edward Cree 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman

rtnetlink: always put IFLA_LINK for links with a link-netnsid

2019-05-25T16:16:13+00:00

[ Upstream commit feadc4b6cf42a53a8a93c918a569a0b7e62bd350 ]

Currently, nla_put_iflink() doesn't put the IFLA_LINK attribute when
iflink == ifindex.

In some cases, a device can be created in a different netns with the
same ifindex as its parent. That device will not dump its IFLA_LINK
attribute, which can confuse some userspace software that expects it.
For example, if the last ifindex created in init_net and foo are both
8, these commands will trigger the issue:

    ip link add parent type dummy                   # ifindex 9
    ip link add link parent netns foo type macvlan  # ifindex 9 in ns foo

So, in case a device puts the IFLA_LINK_NETNSID attribute in a dump,
always put the IFLA_LINK attribute as well.

Thanks to Dan Winship for analyzing the original OpenShift bug down to
the missing netlink attribute.

v2: change Fixes tag, it's been here forever, as Nicolas Dichtel said
    add Nicolas' ack
v3: change Fixes tag
    fix subject typo, spotted by Edward Cree

Analyzed-by: Dan Winship 
Fixes: d8a5ec672768 ("[NET]: netlink support for moving devices between network namespaces.")
Signed-off-by: Sabrina Dubroca 
Acked-by: Nicolas Dichtel 
Signed-off-by: David S. Miller 
Signed-off-by: Greg Kroah-Hartman