The KSZ9893 3-Port Gigabit Ethernet Switch can be controlled via SPI,
I²C or MDIO (very limited and not supported by this driver). While there
is already a compatible entry for the SPI bus, it was missing for I²C.
Signed-off-by: Helmut Grohne <helmut.grohne@intenta.de>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Whenever tcp_try_rmem_schedule() returns an error, we are under
trouble and should make sure to wakeup readers so that they
can drain socket queues and eventually make room.

Fixes: 03f45c88 ("tcp: avoid extra wakeups for SO_RCVLOWAT users")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

When no full socket is available, skbs are sent over a per-netns
control socket. Its sk_mark is temporarily adjusted to match that
of the real (request or timewait) socket or to reflect an incoming
skb, so that the outgoing skb inherits this in __ip_make_skb.

Introduction of the socket cookie mark field broke this. Now the
skb is set through the cookie and cork:

<caller>		# init sockc.mark from sk_mark or cmsg
ip_append_data
  ip_setup_cork		# convert sockc.mark to cork mark
ip_push_pending_frames
  ip_finish_skb
    __ip_make_skb	# set skb->mark to cork mark

But I missed these special control sockets. Update all callers of
__ip(6)_make_skb that were originally missed.

For IPv6, the same two icmp(v6) paths are affected. The third
case is not, as commit 92e55f41 ("tcp: don't annotate
mark on control socket from tcp_v6_send_response()") replaced
the ctl_sk->sk_mark with passing the mark field directly as a
function argument. That commit predates the commit that
introduced the bug.

Fixes: c6af0c22 ("ip: support SO_MARK cmsg")
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reported-by: Martin KaFai Lau <kafai@fb.com>
Reviewed-by: Martin KaFai Lau <kafai@fb.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Whenever cookie_init_timestamp() has been used to encode
ECN,SACK,WSCALE options, we can not remove the TS option in the SYNACK.

Otherwise, tcp_synack_options() will still advertize options like WSCALE
that we can not deduce later when receiving the packet from the client
to complete 3WHS.

Note that modern linux TCP stacks wont use MD5+TS+SACK in a SYN packet,
but we can not know for sure that all TCP stacks have the same logic.

Before the fix a tcpdump would exhibit this wrong exchange :

10:12:15.464591 IP C > S: Flags [S], seq 4202415601, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 456965269 ecr 0,nop,wscale 8], length 0
10:12:15.464602 IP S > C: Flags [S.], seq 253516766, ack 4202415602, win 65535, options [nop,nop,md5 valid,mss 1400,nop,nop,sackOK,nop,wscale 8], length 0
10:12:15.464611 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid], length 0
10:12:15.464678 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid], length 12
10:12:15.464685 IP S > C: Flags [.], ack 13, win 65535, options [nop,nop,md5 valid], length 0

After this patch the exchange looks saner :

11:59:59.882990 IP C > S: Flags [S], seq 517075944, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508483 ecr 0,nop,wscale 8], length 0
11:59:59.883002 IP S > C: Flags [S.], seq 1902939253, ack 517075945, win 65535, options [nop,nop,md5 valid,mss 1400,sackOK,TS val 1751508479 ecr 1751508483,nop,wscale 8], length 0
11:59:59.883012 IP C > S: Flags [.], ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 0
11:59:59.883114 IP C > S: Flags [P.], seq 1:13, ack 1, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508479], length 12
11:59:59.883122 IP S > C: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508483 ecr 1751508483], length 0
11:59:59.883152 IP S > C: Flags [P.], seq 1:13, ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508483], length 12
11:59:59.883170 IP C > S: Flags [.], ack 13, win 256, options [nop,nop,md5 valid,nop,nop,TS val 1751508484 ecr 1751508484], length 0

Of course, no SACK block will ever be added later, but nothing should break.
Technically, we could remove the 4 nops included in MD5+TS options,
but again some stacks could break seeing not conventional alignment.

Fixes: 4957faad ("TCPCT part 1g: Responder Cookie => Initiator")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Florian Westphal <fw@strlen.de>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In testing with mprds enabled, Oracle Cluster nodes after reboot were
not able to communicate with others nodes and so failed to rejoin
the cluster. Peers with lower IP address initiated connection but the
node could not respond as it choose a different path and could not
initiate a connection as it had a higher IP address.

With this patch, when a node sends out a packet and the selected path
is down, all other paths are also checked and any down paths are
re-connected.
Reviewed-by: Ka-cheong Poon <ka-cheong.poon@oracle.com>
Reviewed-by: David Edmondson <david.edmondson@oracle.com>
Signed-off-by: Somasundaram Krishnasamy <somasundaram.krishnasamy@oracle.com>
Signed-off-by: Rao Shoaib <rao.shoaib@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

My prior fix went a bit too far, according to Herbert and Mathieu.

Since we accept that concurrent TCP MD5 lookups might see inconsistent
keys, we can use READ_ONCE()/WRITE_ONCE() instead of smp_rmb()/smp_wmb()

Clearing all key->key[] is needed to avoid possible KMSAN reports,
if key->keylen is increased. Since tcp_md5_do_add() is not fast path,
using __GFP_ZERO to clear all struct tcp_md5sig_key is simpler.

data_race() was added in linux-5.8 and will prevent KCSAN reports,
this can safely be removed in stable backports, if data_race() is
not yet backported.

v2: use data_race() both in tcp_md5_hash_key() and tcp_md5_do_add()

Fixes: 6a2febec ("tcp: md5: add missing memory barriers in tcp_md5_do_add()/tcp_md5_hash_key()")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Cc: Herbert Xu <herbert@gondor.apana.org.au>
Cc: Marco Elver <elver@google.com>
Reviewed-by: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Acked-by: Herbert Xu <herbert@gondor.apana.org.au>
Signed-off-by: David S. Miller <davem@davemloft.net>

A potential deadlock can occur during registering or unregistering a
new generic netlink family between the main nl_table_lock and the
cb_lock where each thread wants the lock held by the other, as
demonstrated below.

1) Thread 1 is performing a netlink_bind() operation on a socket. As part
   of this call, it will call netlink_lock_table(), incrementing the
   nl_table_users count to 1.
2) Thread 2 is registering (or unregistering) a genl_family via the
   genl_(un)register_family() API. The cb_lock semaphore will be taken for
   writing.
3) Thread 1 will call genl_bind() as part of the bind operation to handle
   subscribing to GENL multicast groups at the request of the user. It will
   attempt to take the cb_lock semaphore for reading, but it will fail and
   be scheduled away, waiting for Thread 2 to finish the write.
4) Thread 2 will call netlink_table_grab() during the (un)registration
   call. However, as Thread 1 has incremented nl_table_users, it will not
   be able to proceed, and both threads will be stuck waiting for the
   other.

genl_bind() is a noop, unless a genl_family implements the mcast_bind()
function to handle setting up family-specific multicast operations. Since
no one in-tree uses this functionality as Cong pointed out, simply removing
the genl_bind() function will remove the possibility for deadlock, as there
is no attempt by Thread 1 above to take the cb_lock semaphore.

Fixes: c380d9a7 ("genetlink: pass multicast bind/unbind to families")
Suggested-by: Cong Wang <xiyou.wangcong@gmail.com>
Acked-by: Johannes Berg <johannes.berg@intel.com>
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Sean Tranchetti <stranche@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

get_dev_cap and set_resources_state functions may return a positive
value because of hardware failure, and the positive return value
can not be passed to ERR_PTR directly.

Fixes: 7dd29ee1 ("hinic: add sriov feature support")
Signed-off-by: Luo bin <luobin9@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This code assumes that the user passed in enough data for a
qrtr_hdr_v1 or qrtr_hdr_v2 struct, but it's not necessarily true.  If
the buffer is too small then it will read beyond the end.
Reported-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org>
Reported-by: syzbot+b8fe393f999a291a9ea6@syzkaller.appspotmail.com
Fixes: 194ccc88 ("net: qrtr: Support decoding incoming v2 packets")
Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

MD5 keys are read with RCU protection, and tcp_md5_do_add()
might update in-place a prior key.

Normally, typical RCU updates would allocate a new piece
of memory. In this case only key->key and key->keylen might
be updated, and we do not care if an incoming packet could
see the old key, the new one, or some intermediate value,
since changing the key on a live flow is known to be problematic
anyway.

We only want to make sure that in the case key->keylen
is changed, cpus in tcp_md5_hash_key() wont try to use
uninitialized data, or crash because key->keylen was
read twice to feed sg_init_one() and ahash_request_set_crypt()

Fixes: 9ea88a15 ("tcp: md5: check md5 signature without socket lock")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Mathieu Desnoyers <mathieu.desnoyers@efficios.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The flow is allocated in qrtr_tx_wait, but not freed when qrtr node
is released. (*slot) becomes NULL after radix_tree_iter_delete is
called in __qrtr_node_release. The fix is to save (*slot) to a
vairable and then free it.

This memory leak is catched when kmemleak is enabled in kernel,
the report looks like below:

unreferenced object 0xffffa0de69e08420 (size 32):
  comm "kworker/u16:3", pid 176, jiffies 4294918275 (age 82858.876s)
  hex dump (first 32 bytes):
    00 00 00 00 00 00 00 00 28 84 e0 69 de a0 ff ff  ........(..i....
    28 84 e0 69 de a0 ff ff 03 00 00 00 00 00 00 00  (..i............
  backtrace:
    [<00000000e252af0a>] qrtr_node_enqueue+0x38e/0x400 [qrtr]
    [<000000009cea437f>] qrtr_sendmsg+0x1e0/0x2a0 [qrtr]
    [<000000008bddbba4>] sock_sendmsg+0x5b/0x60
    [<0000000003beb43a>] qmi_send_message.isra.3+0xbe/0x110 [qmi_helpers]
    [<000000009c9ae7de>] qmi_send_request+0x1c/0x20 [qmi_helpers]
Signed-off-by: Carl Huang <cjhuang@codeaurora.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

t4_prep_fw goto bye tag with positive return value when something
bad happened and which can not free resource in adap_init0.
so fix it to return negative value.

Fixes: 16e47624 ("cxgb4: Add new scheme to update T4/T5 firmware")
Reported-by: Hulk Robot <hulkci@huawei.com>
Signed-off-by: Li Heng <liheng40@huawei.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Daniel Borkmann says:

====================
pull-request: bpf 2020-06-30

The following pull-request contains BPF updates for your *net* tree.

We've added 28 non-merge commits during the last 9 day(s) which contain
a total of 35 files changed, 486 insertions(+), 232 deletions(-).

The main changes are:

1) Fix an incorrect verifier branch elimination for PTR_TO_BTF_ID pointer
   types, from Yonghong Song.

2) Fix UAPI for sockmap and flow_dissector progs that were ignoring various
   arguments passed to BPF_PROG_{ATTACH,DETACH}, from Lorenz Bauer & Jakub Sitnicki.

3) Fix broken AF_XDP DMA hacks that are poking into dma-direct and swiotlb
   internals and integrate it properly into DMA core, from Christoph Hellwig.

4) Fix RCU splat from recent changes to avoid skipping ingress policy when
   kTLS is enabled, from John Fastabend.

5) Fix BPF ringbuf map to enforce size to be the power of 2 in order for its
   position masking to work, from Andrii Nakryiko.

6) Fix regression from CAP_BPF work to re-allow CAP_SYS_ADMIN for loading
   of network programs, from Maciej Żenczykowski.

7) Fix libbpf section name prefix for devmap progs, from Jesper Dangaard Brouer.

8) Fix formatting in UAPI documentation for BPF helpers, from Quentin Monnet.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Add two tests for PTR_TO_BTF_ID vs. null ptr comparison,
one for PTR_TO_BTF_ID in the ctx structure and the
other for PTR_TO_BTF_ID after one level pointer chasing.
In both cases, the test ensures condition is not
removed.

For example, for this test
 struct bpf_fentry_test_t {
     struct bpf_fentry_test_t *a;
 };
 int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
 {
     if (arg == 0)
         test7_result = 1;
     return 0;
 }
Before the previous verifier change, we have xlated codes:
  int test7(long long unsigned int * ctx):
  ; int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
     0: (79) r1 = *(u64 *)(r1 +0)
  ; int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
     1: (b4) w0 = 0
     2: (95) exit
After the previous verifier change, we have:
  int test7(long long unsigned int * ctx):
  ; int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
     0: (79) r1 = *(u64 *)(r1 +0)
  ; if (arg == 0)
     1: (55) if r1 != 0x0 goto pc+4
  ; test7_result = 1;
     2: (18) r1 = map[id:6][0]+48
     4: (b7) r2 = 1
     5: (7b) *(u64 *)(r1 +0) = r2
  ; int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
     6: (b4) w0 = 0
     7: (95) exit
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20200630171241.2523875-1-yhs@fb.com

Wenbo reported an issue in [1] where a checking of null
pointer is evaluated as always false. In this particular
case, the program type is tp_btf and the pointer to
compare is a PTR_TO_BTF_ID.

The current verifier considers PTR_TO_BTF_ID always
reprents a non-null pointer, hence all PTR_TO_BTF_ID compares
to 0 will be evaluated as always not-equal, which resulted
in the branch elimination.

For example,
 struct bpf_fentry_test_t {
     struct bpf_fentry_test_t *a;
 };
 int BPF_PROG(test7, struct bpf_fentry_test_t *arg)
 {
     if (arg == 0)
         test7_result = 1;
     return 0;
 }
 int BPF_PROG(test8, struct bpf_fentry_test_t *arg)
 {
     if (arg->a == 0)
         test8_result = 1;
     return 0;
 }

In above bpf programs, both branch arg == 0 and arg->a == 0
are removed. This may not be what developer expected.

The bug is introduced by Commit cac616db ("bpf: Verifier
track null pointer branch_taken with JNE and JEQ"),
where PTR_TO_BTF_ID is considered to be non-null when evaluting
pointer vs. scalar comparison. This may be added
considering we have PTR_TO_BTF_ID_OR_NULL in the verifier
as well.

PTR_TO_BTF_ID_OR_NULL is added to explicitly requires
a non-NULL testing in selective cases. The current generic
pointer tracing framework in verifier always
assigns PTR_TO_BTF_ID so users does not need to
check NULL pointer at every pointer level like a->b->c->d.

We may not want to assign every PTR_TO_BTF_ID as
PTR_TO_BTF_ID_OR_NULL as this will require a null test
before pointer dereference which may cause inconvenience
for developers. But we could avoid branch elimination
to preserve original code intention.

This patch simply removed PTR_TO_BTD_ID from reg_type_not_null()
in verifier, which prevented the above branches from being eliminated.

 [1]: https://lore.kernel.org/bpf/79dbb7c0-449d-83eb-5f4f-7af0cc269168@fb.com/T/

Fixes: cac616db ("bpf: Verifier track null pointer branch_taken with JNE and JEQ")
Reported-by: Wenbo Zhang <ethercflow@gmail.com>
Signed-off-by: Yonghong Song <yhs@fb.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Link: https://lore.kernel.org/bpf/20200630171240.2523722-1-yhs@fb.com

Alex Elder says:

====================
net: ipa: three bug fixes

This series contains three bug fixes for the Qualcomm IPA driver.
In practice these bugs are unlikke.y to be harmful, but they do
represent incorrect code.

Version 2 adds "Fixes" tags to two of the patches and fixes a typo
in one (found by checkpatch.pl).
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Create a new function ipa_cmd_tag_process() that simply allocates a
transaction, adds a tag process command to it to clear the hardware
pipeline, and commits the transaction.

Call it in from ipa_endpoint_suspend(), after suspending the modem
endpoints but before suspending the AP command TX and AP LAN RX
endpoints (which are used by the tag sequence).
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

The AP LAN RX endpoint should not have download checksum offload
enabled.

The receive handler does properly accommodate the trailer that's
added by the hardware, but we ignore it.

Fixes: 1ed7d0c0 ("soc: qcom: ipa: configuration data")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

In gsi_channel_stop(), there's a check to see if the channel might
have entered STOPPED state since a previous call, which might have
timed out before stopping completed.

That check actually belongs in gsi_channel_stop_command(), which is
called repeatedly by gsi_channel_stop() for RX channels.

Fixes: 650d1603 ("soc: qcom: ipa: the generic software interface")
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

When support for short preambles was added, it incorrectly keyed its
decision off state->speed instead of state->interface.  state->speed
is not guaranteed to be correct for in-band modes, which can lead to
short preambles being unexpectedly disabled.

Fix this by keying off the interface mode, which is the only way that
mvneta can operate at 2.5Gbps.

Fixes: da58a931 ("net: mvneta: Add support for 2500Mbps SGMII")
Signed-off-by: Russell King <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jason A. Donenfeld says:

====================
support AF_PACKET for layer 3 devices

Hans reported that packets injected by a correct-looking and trivial
libpcap-based program were not being accepted by wireguard. In
investigating that, I noticed that a few devices weren't properly
handling AF_PACKET-injected packets, and so this series introduces a bit
of shared infrastructure to support that.

The basic problem begins with socket(AF_PACKET, SOCK_RAW,
htons(ETH_P_ALL)) sockets. When sendto is called, AF_PACKET examines the
headers of the packet with this logic:

static void packet_parse_headers(struct sk_buff *skb, struct socket *sock)
{
    if ((!skb->protocol || skb->protocol == htons(ETH_P_ALL)) &&
        sock->type == SOCK_RAW) {
        skb_reset_mac_header(skb);
        skb->protocol = dev_parse_header_protocol(skb);
    }

    skb_probe_transport_header(skb);
}

The middle condition there triggers, and we jump to
dev_parse_header_protocol. Note that this is the only caller of
dev_parse_header_protocol in the kernel, and I assume it was designed
for this purpose:

static inline __be16 dev_parse_header_protocol(const struct sk_buff *skb)
{
    const struct net_device *dev = skb->dev;

    if (!dev->header_ops || !dev->header_ops->parse_protocol)
        return 0;
    return dev->header_ops->parse_protocol(skb);
}

Since AF_PACKET already knows which netdev the packet is going to, the
dev_parse_header_protocol function can see if that netdev has a way it
prefers to figure out the protocol from the header. This, again, is the
only use of parse_protocol in the kernel. At the moment, it's only used
with ethernet devices, via eth_header_parse_protocol. This makes sense,
as mostly people are used to AF_PACKET-injecting ethernet frames rather
than layer 3 frames. But with nothing in place for layer 3 netdevs, this
function winds up returning 0, and skb->protocol then is set to 0, and
then by the time it hits the netdev's ndo_start_xmit, the driver doesn't
know what to do with it.

This is a problem because drivers very much rely on skb->protocol being
correct, and routinely reject packets where it's incorrect. That's why
having this parsing happen for injected packets is quite important. In
wireguard, ipip, and ipip6, for example, packets from AF_PACKET are just
dropped entirely. For tun devices, it's sort of uglier, with the tun
"packet information" header being passed to userspace containing a bogus
protocol value. Some userspace programs are ill-equipped to deal with
that. (But of course, that doesn't happen with tap devices, which
benefit from the similar shared infrastructure for layer 2 netdevs,
further motiviating this patchset for layer 3 netdevs.)

This patchset addresses the issue by first adding a layer 3 header parse
function, much akin to the existing one for layer 2 packets, and then
adds a shared header_ops structure that, also much akin to the existing
one for layer 2 packets. Then it wires it up to a few immediate places
that stuck out as requiring it, and does a bit of cleanup.

This patchset seems like it's fixing real bugs, so it might be
appropriate for stable. But they're also very old bugs, so if you'd
rather not backport to stable, that'd make sense to me too.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

The xfrm interface uses skb->protocol to determine packet type, and
bails out if it's not set. For AF_PACKET injection, we need to support
its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and xfrmi rejects the
skb. So, this wires up the ip_tunnel handler for layer 3 packets for
that case.
Reported-by: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Sit uses skb->protocol to determine packet type, and bails out if it's
not set. For AF_PACKET injection, we need to support its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and sit rejects the
skb. So, this wires up the ip_tunnel handler for layer 3 packets for
that case.
Reported-by: Willem de Bruijn <willemdebruijn.kernel@gmail.com>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Vti uses skb->protocol to determine packet type, and bails out if it's
not set. For AF_PACKET injection, we need to support its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and vti rejects the
skb. So, this wires up the ip_tunnel handler for layer 3 packets for
that case.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The tun driver passes up skb->protocol to userspace in the form of PI headers.
For AF_PACKET injection, we need to support its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and the tun driver
then gives userspace bogus values that it can't deal with.

Note that this isn't the case with tap, because tap already benefits
from the shared infrastructure for ethernet headers. But with tun,
there's nothing.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Now that wg_examine_packet_protocol has been added for general
consumption as ip_tunnel_parse_protocol, it's possible to remove
wg_examine_packet_protocol and simply use the new
ip_tunnel_parse_protocol function directly.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

WireGuard uses skb->protocol to determine packet type, and bails out if
it's not set or set to something it's not expecting. For AF_PACKET
injection, we need to support its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and wireguard then
rejects the skb. So, this wires up the ip_tunnel handler for layer 3
packets for that case.
Reported-by: Hans Wippel <ndev@hwipl.net>
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Ipip uses skb->protocol to determine packet type, and bails out if it's
not set. For AF_PACKET injection, we need to support its call chain of:

    packet_sendmsg -> packet_snd -> packet_parse_headers ->
      dev_parse_header_protocol -> parse_protocol

Without a valid parse_protocol, this returns zero, and ipip rejects the
skb. So, this wires up the ip_tunnel handler for layer 3 packets for
that case.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Some devices that take straight up layer 3 packets benefit from having a
shared header_ops so that AF_PACKET sockets can inject packets that are
recognized. This shared infrastructure will be used by other drivers
that currently can't inject packets using AF_PACKET. It also exposes the
parser function, as it is useful in standalone form too.
Signed-off-by: Jason A. Donenfeld <Jason@zx2c4.com>
Acked-by: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Iterating over BPF links attached to network namespace in pre_exit hook is
not safe, even if there is just one. Once link gets auto-detached, that is
its back-pointer to net object is set to NULL, the link can be released and
freed without waiting on netns_bpf_mutex, effectively causing the list
element we are operating on to be freed.

This leads to use-after-free when trying to access the next element on the
list, as reported by KASAN. Bug can be triggered by destroying a network
namespace, while also releasing a link attached to this network namespace.

| ==================================================================
| BUG: KASAN: use-after-free in netns_bpf_pernet_pre_exit+0xd9/0x130
| Read of size 8 at addr ffff888119e0d778 by task kworker/u8:2/177
|
| CPU: 3 PID: 177 Comm: kworker/u8:2 Not tainted 5.8.0-rc1-00197-ga0c04c9d1008-dirty #776
| Hardware name: QEMU Standard PC (i440FX + PIIX, 1996), BIOS ?-20190727_073836-buildvm-ppc64le-16.ppc.fedoraproject.org-3.fc31 04/01/2014
| Workqueue: netns cleanup_net
| Call Trace:
|  dump_stack+0x9e/0xe0
|  print_address_description.constprop.0+0x3a/0x60
|  ? netns_bpf_pernet_pre_exit+0xd9/0x130
|  kasan_report.cold+0x1f/0x40
|  ? netns_bpf_pernet_pre_exit+0xd9/0x130
|  netns_bpf_pernet_pre_exit+0xd9/0x130
|  cleanup_net+0x30b/0x5b0
|  ? unregister_pernet_device+0x50/0x50
|  ? rcu_read_lock_bh_held+0xb0/0xb0
|  ? _raw_spin_unlock_irq+0x24/0x50
|  process_one_work+0x4d1/0xa10
|  ? lock_release+0x3e0/0x3e0
|  ? pwq_dec_nr_in_flight+0x110/0x110
|  ? rwlock_bug.part.0+0x60/0x60
|  worker_thread+0x7a/0x5c0
|  ? process_one_work+0xa10/0xa10
|  kthread+0x1e3/0x240
|  ? kthread_create_on_node+0xd0/0xd0
|  ret_from_fork+0x1f/0x30
|
| Allocated by task 280:
|  save_stack+0x1b/0x40
|  __kasan_kmalloc.constprop.0+0xc2/0xd0
|  netns_bpf_link_create+0xfe/0x650
|  __do_sys_bpf+0x153a/0x2a50
|  do_syscall_64+0x59/0x300
|  entry_SYSCALL_64_after_hwframe+0x44/0xa9
|
| Freed by task 198:
|  save_stack+0x1b/0x40
|  __kasan_slab_free+0x12f/0x180
|  kfree+0xed/0x350
|  process_one_work+0x4d1/0xa10
|  worker_thread+0x7a/0x5c0
|  kthread+0x1e3/0x240
|  ret_from_fork+0x1f/0x30
|
| The buggy address belongs to the object at ffff888119e0d700
|  which belongs to the cache kmalloc-192 of size 192
| The buggy address is located 120 bytes inside of
|  192-byte region [ffff888119e0d700, ffff888119e0d7c0)
| The buggy address belongs to the page:
| page:ffffea0004678340 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0
| flags: 0x2fffe0000000200(slab)
| raw: 02fffe0000000200 ffffea00045ba8c0 0000000600000006 ffff88811a80ea80
| raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
| page dumped because: kasan: bad access detected
|
| Memory state around the buggy address:
|  ffff888119e0d600: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
|  ffff888119e0d680: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
| >ffff888119e0d700: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
|                                                                 ^
|  ffff888119e0d780: fb fb fb fb fb fb fb fb fc fc fc fc fc fc fc fc
|  ffff888119e0d800: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
| ==================================================================

Remove the "fast-path" for releasing a link that got auto-detached by a
dying network namespace to fix it. This way as long as link is on the list
and netns_bpf mutex is held, we have a guarantee that link memory can be
accessed.

An alternative way to fix this issue would be to safely iterate over the
list of links and ensure there is no access to link object after detaching
it. But, at the moment, optimizing synchronization overhead on link release
without a workload in mind seems like an overkill.

Fixes: ab53cad9 ("bpf, netns: Keep a list of attached bpf_link's")
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20200630164541.1329993-1-jakub@cloudflare.com

Lorenz Bauer says:

====================
Both sockmap and flow_dissector ingnore various arguments passed to
BPF_PROG_ATTACH and BPF_PROG_DETACH. We can fix the attach case by
checking that the unused arguments are zero. I considered requiring
target_fd to be -1 instead of 0, but this leads to a lot of churn
in selftests. There is also precedent in that bpf_iter already
expects 0 for a similar field. I think that we can come up with a
work around for fd 0 should we need to in the future.

The detach case is more problematic: both cgroups and lirc2 verify
that attach_bpf_fd matches the currently attached program. This
way you need access to the program fd to be able to remove it.
Neither sockmap nor flow_dissector do this. flow_dissector even
has a check for CAP_NET_ADMIN because of this. The patch set
addresses this by implementing the desired behaviour.

There is a possibility for user space breakage: any callers that
don't provide the correct fd will fail with ENOENT. For sockmap
the risk is low: even the selftests assume that sockmap works
the way I described. For flow_dissector the story is less
straightforward, and the selftests use a variety of arguments.

I've includes fixes tags for the oldest commits that allow an easy
backport, however the behaviour dates back to when sockmap and
flow_dissector were introduced. What is the best way to handle these?

This set is based on top of Jakub's work "bpf, netns: Prepare
for multi-prog attachment" available at
https://lore.kernel.org/bpf/87k0zwmhtb.fsf@cloudflare.com/T/

Since v1:
- Adjust selftests
- Implement detach behaviour
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

Calling bpf_prog_detach is incorrect, since it takes target_fd as
its argument. The intention here is to pass it as attach_bpf_fd,
so use bpf_prog_detach2 and pass zero for target_fd.

Fixes: 06716e04 ("selftests/bpf: Extend test_flow_dissector to cover link creation")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-7-lmb@cloudflare.com

Pass 0 as target_fd when attaching and detaching flow dissector.
Additionally, pass the expected program when detaching.

Fixes: 1f043f87 ("selftests/bpf: Add tests for attaching bpf_link to netns")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-6-lmb@cloudflare.com

The sockmap code currently ignores the value of attach_bpf_fd when
detaching a program. This is contrary to the usual behaviour of
checking that attach_bpf_fd represents the currently attached
program.

Ensure that attach_bpf_fd is indeed the currently attached
program. It turns out that all sockmap selftests already do this,
which indicates that this is unlikely to cause breakage.

Fixes: 604326b4 ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-5-lmb@cloudflare.com

Using BPF_PROG_ATTACH on a sockmap program currently understands no
flags or replace_bpf_fd, but accepts any value. Return EINVAL instead.

Fixes: 604326b4 ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-4-lmb@cloudflare.com

Using BPF_PROG_DETACH on a flow dissector program supports neither
attach_flags nor attach_bpf_fd. Yet no value is enforced for them.

Enforce that attach_flags are zero, and require the current program
to be passed via attach_bpf_fd. This allows us to remove the check
for CAP_SYS_ADMIN, since userspace can now no longer remove
arbitrary flow dissector programs.

Fixes: b27f7bb5 ("flow_dissector: Move out netns_bpf prog callbacks")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-3-lmb@cloudflare.com

Using BPF_PROG_ATTACH on a flow dissector program supports neither
target_fd, attach_flags or replace_bpf_fd but accepts any value.

Enforce that all of them are zero. This is fine for replace_bpf_fd
since its presence is indicated by BPF_F_REPLACE. It's more
problematic for target_fd, since zero is a valid fd. Should we
want to use the flag later on we'd have to add an exception for
fd 0. The alternative is to force a value like -1. This requires
more changes to tests. There is also precedent for using 0,
since bpf_iter uses this for target_fd as well.

Fixes: b27f7bb5 ("flow_dissector: Move out netns_bpf prog callbacks")
Signed-off-by: Lorenz Bauer <lmb@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Link: https://lore.kernel.org/bpf/20200629095630.7933-2-lmb@cloudflare.com

Jakub Sitnicki says:

====================
This patch set prepares ground for link-based multi-prog attachment for
future netns attach types, with BPF_SK_LOOKUP attach type in mind [0].

Two changes are needed in order to attach and run a series of BPF programs:

  1) an bpf_prog_array of programs to run (patch #2), and
  2) a list of attached links to keep track of attachments (patch #3).

Nothing changes for BPF flow_dissector. Just as before only one program can
be attached to netns.

In v3 I've simplified patch #2 that introduces bpf_prog_array to take
advantage of the fact that it will hold at most one program for now.

In particular, I'm no longer using bpf_prog_array_copy. It turned out to be
less suitable for link operations than I thought as it fails to append the
same BPF program.

bpf_prog_array_replace_item is also gone, because we know we always want to
replace the first element in prog_array.

Naturally the code that handles bpf_prog_array will need change once
more when there is a program type that allows multi-prog attachment. But I
feel it will be better to do it gradually and present it together with
tests that actually exercise multi-prog code paths.

[0] https://lore.kernel.org/bpf/20200511185218.1422406-1-jakub@cloudflare.com/

v2 -> v3:
- Don't check if run_array is null in link update callback. (Martin)
- Allow updating the link with the same BPF program. (Andrii)
- Add patch #4 with a test for the above case.
- Kill bpf_prog_array_replace_item. Access the run_array directly.
- Switch from bpf_prog_array_copy() to bpf_prog_array_alloc(1, ...).
- Replace rcu_deref_protected & RCU_INIT_POINTER with rcu_replace_pointer.
- Drop Andrii's Ack from patch #2. Code changed.

v1 -> v2:

- Show with a (void) cast that bpf_prog_array_replace_item() return value
  is ignored on purpose. (Andrii)
- Explain why bpf-cgroup cannot replace programs in bpf_prog_array based
  on bpf_prog pointer comparison in patch #2 description. (Andrii)
====================
Signed-off-by: Alexei Starovoitov <ast@kernel.org>

This case, while not particularly useful, is worth covering because we
expect the operation to succeed as opposed when re-attaching the same
program directly with PROG_ATTACH.

While at it, update the tests summary that fell out of sync when tests
extended to cover links.
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20200625141357.910330-5-jakub@cloudflare.com

To support multi-prog link-based attachments for new netns attach types, we
need to keep track of more than one bpf_link per attach type. Hence,
convert net->bpf.links into a list, that currently can be either empty or
have just one item.

Instead of reusing bpf_prog_list from bpf-cgroup, we link together
bpf_netns_link's themselves. This makes list management simpler as we don't
have to allocate, initialize, and later release list elements. We can do
this because multi-prog attachment will be available only for bpf_link, and
we don't need to build a list of programs attached directly and indirectly
via links.

No functional changes intended.
Signed-off-by: Jakub Sitnicki <jakub@cloudflare.com>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Acked-by: Andrii Nakryiko <andriin@fb.com>
Acked-by: Martin KaFai Lau <kafai@fb.com>
Link: https://lore.kernel.org/bpf/20200625141357.910330-4-jakub@cloudflare.com