Based on review, `lock' is only acquired in hwbm_pool_add() which is
invoked via ->probe(), ->resume() and ->ndo_change_mtu(). Based on this
the lock can become a mutex and there is no need to disable interrupts
during the procedure.
Now that the lock is a mutex, hwbm_pool_add() no longer invokes
hwbm_pool_refill() in an atomic context so we can pass GFP_KERNEL to
hwbm_pool_refill() and remove the `gfp' argument from hwbm_pool_add().

Cc: Thomas Petazzoni <thomas.petazzoni@bootlin.com>
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

__netdev_alloc_skb() can be used from any context and is used by NAPI
and non-NAPI drivers. Non-NAPI drivers use it in interrupt context and
NAPI drivers use it during initial allocation (->ndo_open() or
->ndo_change_mtu()). Some NAPI drivers share the same function for the
initial allocation and the allocation in their NAPI callback.

The interrupts are disabled in order to ensure locked access from every
context to `netdev_alloc_cache'.

Let __netdev_alloc_skb() check if interrupts are disabled. If they are, use
`netdev_alloc_cache'. Otherwise disable BH and use `napi_alloc_cache.page'.
The IRQ check is cheaper compared to disabling & enabling interrupts and
memory allocation with disabled interrupts does not work on -RT.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

netdev_alloc_frag() can be used from any context and is used by NAPI
and non-NAPI drivers. Non-NAPI drivers use it in interrupt context
and NAPI drivers use it during initial allocation (->ndo_open() or
->ndo_change_mtu()). Some NAPI drivers share the same function for the
initial allocation and the allocation in their NAPI callback.

The interrupts are disabled in order to ensure locked access from every
context to `netdev_alloc_cache'.

Let netdev_alloc_frag() check if interrupts are disabled. If they are,
use `netdev_alloc_cache' otherwise disable BH and invoke
__napi_alloc_frag() for the allocation. The IRQ check is cheaper
compared to disabling & enabling interrupts and memory allocation with
disabled interrupts does not work on -RT.
Signed-off-by: NSebastian Andrzej Siewior <bigeasy@linutronix.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: 6146e6a4 ("[PKTGEN]: Removes thread_{un,}lock() macros.")
Reported-and-tested-by: NMatteo Croce <mcroce@redhat.com>
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This reverts commit e9919a24.

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: e9919a24 ("fib_rules: return 0 directly if an exactly same rule exists when NLM_F_EXCL not supplied")
Reported-by: NNathan Chancellor <natechancellor@gmail.com>
Reported-by: NYaro Slav <yaro330@gmail.com>
Reported-by: NMaciej Żenczykowski <zenczykowski@gmail.com>
Signed-off-by: NHangbin Liu <liuhangbin@gmail.com>
Reviewed-by: NNathan Chancellor <natechancellor@gmail.com>
Tested-by: NNathan Chancellor <natechancellor@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: NVivien Didelot <vivien.didelot@gmail.com>
Reviewed-by: NMichal Kubecek <mkubecek@suse.cz>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

As Eric noted, the current wrapper for ptype func hook inside
__netif_receive_skb_list_ptype() has no chance of avoiding the indirect
call: we enter such code path only for protocols other than ipv4 and
ipv6.

Instead we can wrap the list_func invocation.

v1 -> v2:
 - use the correct fix tag

Fixes: f5737cba ("net: use indirect calls helpers for ptype hook")
Suggested-by: NEric Dumazet <eric.dumazet@gmail.com>
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Acked-by: NEdward Cree <ecree@solarflare.com>
Reviewed-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use helpers to access fib_nh and fib_nhs fields of a fib_info. Drop the
fib_dev macro which is an alias for the first nexthop. Replacements:

  fi->fib_dev    --> fib_info_nh(fi, 0)->fib_nh_dev
  fi->fib_nh     --> fib_info_nh(fi, 0)
  fi->fib_nh[i]  --> fib_info_nh(fi, i)
  fi->fib_nhs    --> fib_info_num_path(fi)

where fib_info_nh(fi, i) returns fi->fib_nh[nhsel] and fib_info_num_path
returns fi->fib_nhs.

Move the existing fib_info_nhc to nexthop.h and define the new ones
there. A later patch adds a check if a fib_info uses a nexthop object,
and defining the helpers in nexthop.h avoid circular header
dependencies.

After this all remaining open coded references to fi->fib_nhs and
fi->fib_nh are in:
- fib_create_info and helpers used to lookup an existing fib_info
  entry, and
- the netdev event functions fib_sync_down_dev and fib_sync_up.

The latter two will not be reused for nexthops, and the fib_create_info
will be updated to handle a nexthop in a fib_info.
Signed-off-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

If strparser gets cornered into starting a new message from
an sk_buff which already has frags, it will allocate a new
skb to become the "wrapper" around the fragments of the
message.

This new skb does not inherit any metadata fields.  In case
of TLS offload this may lead to unnecessarily re-encrypting
the message, as skb->decrypted is not set for the wrapper skb.

Try to be conservative and copy all fields of old skb
strparser's user may reasonably need.
Signed-off-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: NDirk van der Merwe <dirk.vandermerwe@netronome.com>
Reviewed-by: NSimon Horman <simon.horman@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Introduce a function to be called from drivers during flash. It sends
notification to userspace about flash update progress.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Reviewed-by: NIdo Schimmel <idosch@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This flag is not used by any caller, remove it.
Signed-off-by: NStanislav Fomichev <sdf@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ifa_list is protected by rcu, yet code doesn't reflect this.

Add the __rcu annotations and fix up all places that are now reported by
sparse.

I've done this in the same commit to not add intermediate patches that
result in new warnings.
Reported-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NFlorian Westphal <fw@strlen.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Most bpf map types doing similar checks and bytes to pages
conversion during memory allocation and charging.

Let's unify these checks by moving them into bpf_map_charge_init().
Signed-off-by: NRoman Gushchin <guro@fb.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

In order to unify the existing memlock charging code with the
memcg-based memory accounting, which will be added later, let's
rework the current scheme.

Currently the following design is used:
  1) .alloc() callback optionally checks if the allocation will likely
     succeed using bpf_map_precharge_memlock()
  2) .alloc() performs actual allocations
  3) .alloc() callback calculates map cost and sets map.memory.pages
  4) map_create() calls bpf_map_init_memlock() which sets map.memory.user
     and performs actual charging; in case of failure the map is
     destroyed
  <map is in use>
  1) bpf_map_free_deferred() calls bpf_map_release_memlock(), which
     performs uncharge and releases the user
  2) .map_free() callback releases the memory

The scheme can be simplified and made more robust:
  1) .alloc() calculates map cost and calls bpf_map_charge_init()
  2) bpf_map_charge_init() sets map.memory.user and performs actual
    charge
  3) .alloc() performs actual allocations
  <map is in use>
  1) .map_free() callback releases the memory
  2) bpf_map_charge_finish() performs uncharge and releases the user

The new scheme also allows to reuse bpf_map_charge_init()/finish()
functions for memcg-based accounting. Because charges are performed
before actual allocations and uncharges after freeing the memory,
no bogus memory pressure can be created.

In cases when the map structure is not available (e.g. it's not
created yet, or is already destroyed), on-stack bpf_map_memory
structure is used. The charge can be transferred with the
bpf_map_charge_move() function.
Signed-off-by: NRoman Gushchin <guro@fb.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Group "user" and "pages" fields of bpf_map into the bpf_map_memory
structure. Later it can be extended with "memcg" and other related
information.

The main reason for a such change (beside cosmetics) is to pass
bpf_map_memory structure to charging functions before the actual
allocation of bpf_map.
Signed-off-by: NRoman Gushchin <guro@fb.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Socket local storage maps lack the memlock precharge check,
which is performed before the memory allocation for
most other bpf map types.

Let's add it in order to unify all map types.
Signed-off-by: NRoman Gushchin <guro@fb.com>
Acked-by: NSong Liu <songliubraving@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

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: 52900d22 ("udp: elide zerocopy operation in hot path")
Reported-by: Nsyzbot <syzkaller@googlegroups.com>
Diagnosed-by: NEric Dumazet <edumazet@google.com>
Signed-off-by: NWillem de Bruijn <willemb@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: eca4205f ("ethtool: add ethtool_rx_flow_spec to flow_rule structure translator")
Signed-off-by: NMaxime Chevallier <maxime.chevallier@bootlin.com>
Acked-by: NPablo Neira Ayuso <pablo@gnumonks.org>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: a50e233c ("net-gro: restore frag0 optimization")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: Nsyzbot <syzkaller@googlegroups.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 283c16a2 ("indirect call wrappers: helpers to speed-up
indirect calls of builtin") introduces some macros to avoid doing
indirect calls.

Use these helpers to remove two indirect calls in the L4 checksum
calculation for devices which don't have hardware support for it.

As a test I generate packets with pktgen out to a dummy interface
with HW checksumming disabled, to have the checksum calculated in
every sent packet.
The packet rate measured with an i7-6700K CPU and a single pktgen
thread raised from 6143 to 6608 Kpps, an increase by 7.5%
Suggested-by: NDavide Caratti <dcaratti@redhat.com>
Signed-off-by: NMatteo Croce <mcroce@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license as published by
  the free software foundation either version 2 of the license or at
  your option any later version

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-or-later

has been chosen to replace the boilerplate/reference in 3029 file(s).
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: NAllison Randal <allison@lohutok.net>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190527070032.746973796@linutronix.deSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When a device is stacked like (team, bonding, failsafe or netvsc) the
XDP generic program for the parent device was not called.

Move the call to XDP generic inside __netif_receive_skb_core where
it can be done multiple times for stacked case.

Fixes: d4455169 ("net: xdp: support xdp generic on virtual devices")
Signed-off-by: NStephen Hemminger <sthemmin@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

One of the more common cases of allocation size calculations is finding
the size of a structure that has a zero-sized array at the end, along
with memory for some number of elements for that array. For example:

struct foo {
   int stuff;
   struct boo entry[];
};

instance = kzalloc(sizeof(struct foo) + count * sizeof(struct boo), GFP_KERNEL);

Instead of leaving these open-coded and prone to type mistakes, we can
now use the new struct_size() helper:

instance = kzalloc(struct_size(instance, entry, count), GFP_KERNEL);

This code was detected with the help of Coccinelle.
Signed-off-by: NGustavo A. R. Silva <gustavo@embeddedor.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Based on 1 normalized pattern(s):

  this program is free software you can redistribute it and or modify
  it under the terms of the gnu general public license as published by
  the free software foundation either version 2 of the license or at
  your option any later version this program is distributed in the
  hope that it will be useful but without any warranty without even
  the implied warranty of merchantability or fitness for a particular
  purpose see the gnu general public license for more details you
  should have received a copy of the gnu general public license along
  with this program if not write to the free software foundation inc
  675 mass ave cambridge ma 02139 usa

extracted by the scancode license scanner the SPDX license identifier

  GPL-2.0-or-later

has been chosen to replace the boilerplate/reference in 441 file(s).
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Reviewed-by: Michael Ellerman <mpe@ellerman.id.au> (powerpc)
Reviewed-by: NRichard Fontana <rfontana@redhat.com>
Reviewed-by: NAllison Randal <allison@lohutok.net>
Reviewed-by: NKate Stewart <kstewart@linuxfoundation.org>
Cc: linux-spdx@vger.kernel.org
Link: https://lkml.kernel.org/r/20190520071858.739733335@linutronix.deSigned-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Prevent misbehavior of drivers who would not set port type for longer
period of time. Drivers should always set port type. Do WARN if that
happens.

Note that it is perfectly fine to temporarily not have the type set,
during initialization and port type change.
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add tracepoint to __neigh_create to enable debugging of new entries.
Signed-off-by: NDavid Ahern <dsahern@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add SPDX license identifiers to all files which:

 - Have no license information of any form

 - Have MODULE_LICENCE("GPL*") inside which was used in the initial
   scan/conversion to ignore the file

These files fall under the project license, GPL v2 only. The resulting SPDX
license identifier is:

  GPL-2.0-only
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

Add SPDX license identifiers to all files which:

 - Have no license information of any form

 - Have EXPORT_.*_SYMBOL_GPL inside which was used in the
   initial scan/conversion to ignore the file

These files fall under the project license, GPL v2 only. The resulting SPDX
license identifier is:

  GPL-2.0-only
Signed-off-by: NThomas Gleixner <tglx@linutronix.de>
Signed-off-by: NGreg Kroah-Hartman <gregkh@linuxfoundation.org>

When host is under high stress, it is very possible thread
running netdev_wait_allrefs() returns from msleep(250)
10 seconds late.

This leads to these messages in the syslog :

[...] unregister_netdevice: waiting for syz_tun to become free. Usage count = 0

If the device refcount is zero, the wait is over.
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: Nsyzbot <syzkaller@googlegroups.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Plumb it through from the flow_dissector.
Signed-off-by: NEdward Cree <ecree@solarflare.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: NDan Winship <danw@redhat.com>
Fixes: d8a5ec67 ("[NET]: netlink support for moving devices between network namespaces.")
Signed-off-by: NSabrina Dubroca <sd@queasysnail.net>
Acked-by: NNicolas Dichtel <nicolas.dichtel@6wind.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When converting a skb to msg->sg we forget to set the size after the
latest ktls/tls code conversion. This patch can be reached by doing
a redir into ingress path from BPF skb sock recv hook. Then trying to
read the size fails.

Fix this by setting the size.

Fixes: 604326b4 ("bpf, sockmap: convert to generic sk_msg interface")
Signed-off-by: NJohn Fastabend <john.fastabend@gmail.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>

If we try to call strp_done on a parser that has never been
initialized, because the sockmap user is only using TX side for
example we get the following error.

  [  883.422081] WARNING: CPU: 1 PID: 208 at kernel/workqueue.c:3030 __flush_work+0x1ca/0x1e0
  ...
  [  883.422095] Workqueue: events sk_psock_destroy_deferred
  [  883.422097] RIP: 0010:__flush_work+0x1ca/0x1e0

This had been wrapped in a 'if (psock->parser.enabled)' logic which
was broken because the strp_done() was never actually being called
because we do a strp_stop() earlier in the tear down logic will
set parser.enabled to false. This could result in a use after free
if work was still in the queue and was resolved by the patch here,
1d79895a ("sk_msg: Always cancel strp work before freeing the
psock"). However, calling strp_stop(), done by the patch marked in
the fixes tag, only is useful if we never initialized a strp parser
program and never initialized the strp to start with. Because if
we had initialized a stream parser strp_stop() would have been called
by sk_psock_drop() earlier in the tear down process.  By forcing the
strp to stop we get past the WARNING in strp_done that checks
the stopped flag but calling cancel_work_sync on work that has never
been initialized is also wrong and generates the warning above.

To fix check if the parser program exists. If the program exists
then the strp work has been initialized and must be sync'd and
cancelled before free'ing any structures. If no program exists we
never initialized the stream parser in the first place so skip the
sync/cancel logic implemented by strp_done.

Finally, remove the strp_done its not needed and in the case where we
are using the stream parser has already been called.

Fixes: e8e34377 ("bpf: Stop the psock parser before canceling its work")
Signed-off-by: NJohn Fastabend <john.fastabend@gmail.com>
Signed-off-by: NDaniel Borkmann <daniel@iogearbox.net>

Various things in eBPF really require us to disable preemption
before running an eBPF program.

syzbot reported :

BUG: assuming atomic context at net/core/flow_dissector.c:737
in_atomic(): 0, irqs_disabled(): 0, pid: 24710, name: syz-executor.3
2 locks held by syz-executor.3/24710:
 #0: 00000000e81a4bf1 (&tfile->napi_mutex){+.+.}, at: tun_get_user+0x168e/0x3ff0 drivers/net/tun.c:1850
 #1: 00000000254afebd (rcu_read_lock){....}, at: __skb_flow_dissect+0x1e1/0x4bb0 net/core/flow_dissector.c:822
CPU: 1 PID: 24710 Comm: syz-executor.3 Not tainted 5.1.0+ #6
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
 __cant_sleep kernel/sched/core.c:6165 [inline]
 __cant_sleep.cold+0xa3/0xbb kernel/sched/core.c:6142
 bpf_flow_dissect+0xfe/0x390 net/core/flow_dissector.c:737
 __skb_flow_dissect+0x362/0x4bb0 net/core/flow_dissector.c:853
 skb_flow_dissect_flow_keys_basic include/linux/skbuff.h:1322 [inline]
 skb_probe_transport_header include/linux/skbuff.h:2500 [inline]
 skb_probe_transport_header include/linux/skbuff.h:2493 [inline]
 tun_get_user+0x2cfe/0x3ff0 drivers/net/tun.c:1940
 tun_chr_write_iter+0xbd/0x156 drivers/net/tun.c:2037
 call_write_iter include/linux/fs.h:1872 [inline]
 do_iter_readv_writev+0x5fd/0x900 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
 __do_sys_writev fs/read_write.c:1131 [inline]
 __se_sys_writev fs/read_write.c:1128 [inline]
 __x64_sys_writev+0x75/0xb0 fs/read_write.c:1128
 do_syscall_64+0x103/0x670 arch/x86/entry/common.c:298
 entry_SYSCALL_64_after_hwframe+0x49/0xbe

Fixes: d58e468b ("flow_dissector: implements flow dissector BPF hook")
Signed-off-by: NEric Dumazet <edumazet@google.com>
Reported-by: Nsyzbot <syzkaller@googlegroups.com>
Cc: Petar Penkov <ppenkov@google.com>
Cc: Stanislav Fomichev <sdf@google.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

With commit 153380ec ("fib_rules: Added NLM_F_EXCL support to
fib_nl_newrule") we now able to check if a rule already exists. But this
only works with iproute2. For other tools like libnl, NetworkManager,
it still could add duplicate rules with only NLM_F_CREATE flag, like

[localhost ~ ]# ip rule
0:      from all lookup local
32766:  from all lookup main
32767:  from all lookup default
100000: from 192.168.7.5 lookup 5
100000: from 192.168.7.5 lookup 5

As it doesn't make sense to create two duplicate rules, let's just return
0 if the rule exists.

Fixes: 153380ec ("fib_rules: Added NLM_F_EXCL support to fib_nl_newrule")
Reported-by: NThomas Haller <thaller@redhat.com>
Signed-off-by: NHangbin Liu <liuhangbin@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This avoids an indirect call per RX IPv6/IPv4 packet.
Note that we don't want to use the indirect calls helper for taps.
Signed-off-by: NPaolo Abeni <pabeni@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Use PTR_ERR_OR_ZERO rather than if(IS_ERR(...)) + PTR_ERR
Signed-off-by: NYueHaibing <yuehaibing@huawei.com>
Acked-by: NMartin KaFai Lau <kafai@fb.com>
Signed-off-by: NAlexei Starovoitov <ast@kernel.org>

Commit cd9ff4de 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: cd9ff4de ("ipv4: Make neigh lookup keys for loopback/point-to-point devices be INADDR_ANY")
Reported-by: NAlan Maguire <alan.maguire@oracle.com>
Signed-off-by: NDavid Ahern <dsahern@gmail.com>
Tested-by: NAlan Maguire <alan.maguire@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

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: 58956317 ("neighbor: Improve garbage collection")
Reported-by: NIan Kumlien <ian.kumlien@gmail.com>
Reported-by: NAlan Maguire <alan.maguire@oracle.com>
Signed-off-by: NDavid Ahern <dsahern@gmail.com>
Tested-by: NAlan Maguire <alan.maguire@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The devlink health reporters create/destroy and user commands currently
use the devlink->lock as a locking mechanism. Different reporters have
different rules in the driver and are being created/destroyed during
different stages of driver load/unload/running. So during execution of a
reporter recover the flow can go through another reporter's destroy and
create. Such flow leads to deadlock trying to lock a mutex already
held.

With the new locking mechanism the different reporters share mutex lock
only to protect access to shared reporters list.
Added refcount per reporter, to protect the reporters from destroy while
being used.
Signed-off-by: NMoshe Shemesh <moshe@mellanox.com>
Signed-off-by: NJiri Pirko <jiri@mellanox.com>
Acked-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>