Long standing KCSAN issues are caused by data-race around
some dev->stats changes.
Most performance critical paths already use per-cpu
variables, or per-queue ones.
It is reasonable (and more correct) to use atomic operations
for the slow paths.
This patch adds an union for each field of net_device_stats,
so that we can convert paths that are not yet protected
by a spinlock or a mutex.
netdev_stats_to_stats64() no longer has an #if BITS_PER_LONG==64
Note that the memcpy() we were using on 64bit arches
had no provision to avoid load-tearing,
while atomic_long_read() is providing the needed protection
at no cost.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adopt atomic64_try_cmpxchg() and remove the loop,
to make the intent more obvious.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adopting atomic_try_cmpxchg() makes the code cleaner.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Adopt atomic_long_try_cmpxchg() in mm_account_pinned_pages()
as it is slightly more efficient.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
RFC 2863 says:
The lowerLayerDown state is also a refinement on the down state.
This new state indicates that this interface runs "on top of" one or
more other interfaces (see ifStackTable) and that this interface is
down specifically because one or more of these lower-layer interfaces
are down.
DSA interfaces are virtual network devices, stacked on top of the DSA
master, but they have a physical MAC, with a PHY that reports a real
link status.
But since DSA (perhaps improperly) uses an iflink to describe the
relationship to its master since commit c084080151 ("dsa: set ->iflink
on slave interfaces to the ifindex of the parent"), default_operstate()
will misinterpret this to mean that every time the carrier of a DSA
interface is not ok, it is because of the master being not ok.
In fact, since commit c0a8a9c274 ("net: dsa: automatically bring user
ports down when master goes down"), DSA cannot even in theory be in the
lowerLayerDown state, because it just calls dev_close_many(), thereby
going down, when the master goes down.
We could revert the commit that creates an iflink between a DSA user
port and its master, especially since now we have an alternative
IFLA_DSA_MASTER which has less side effects. But there may be tooling in
use which relies on the iflink, which has existed since 2009.
We could also probably do something local within DSA to overwrite what
rfc2863_policy() did, in a way similar to hsr_set_operstate(), but this
seems like a hack.
What seems appropriate is to follow the iflink, and check the carrier
status of that interface as well. If that's down too, yes, keep
reporting lowerLayerDown, otherwise just down.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
We will soon introduce an optional per-netns hash table
for UDP.
This means we cannot use udp_table directly in most places.
Instead, access it via net->ipv4.udp_table.
The access will be valid only while initialising udp_table
itself and creating/destroying each netns.
Signed-off-by: Kuniyuki Iwashima <kuniyu@amazon.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
If a metadata dst is present with the type METADATA_HW_PORT_MUX on a dsa cpu
port netdev, assume that it carries the port number and that there is no DSA
tag present in the skb data.
Signed-off-by: Felix Fietkau <nbd@nbd.name>
Reviewed-by: Vladimir Oltean <olteanv@gmail.com>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
For queueing packets in XDP we want to add a new redirect map type with
support for 64-bit indexes. To prepare fore this, expand the width of the
'key' argument to the bpf_redirect_map() helper. Since BPF registers are
always 64-bit, this should be safe to do after the fact.
Acked-by: Song Liu <song@kernel.org>
Reviewed-by: Stanislav Fomichev <sdf@google.com>
Signed-off-by: Toke Høiland-Jørgensen <toke@redhat.com>
Link: https://lore.kernel.org/r/20221108140601.149971-3-toke@redhat.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Instead of having to pass multiple arguments that describe the register,
pass the bpf_reg_state into the btf_struct_access callback. Currently,
all call sites simply reuse the btf and btf_id of the reg they want to
check the access of. The only exception to this pattern is the callsite
in check_ptr_to_map_access, hence for that case create a dummy reg to
simulate PTR_TO_BTF_ID access.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20221114191547.1694267-8-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
This adds a new flow_rule_match_arp function that allows drivers
to be able to dissect ARP frames.
Signed-off-by: Steen Hegelund <steen.hegelund@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Andrii Nakryiko says:
====================
bpf-next 2022-11-11
We've added 49 non-merge commits during the last 9 day(s) which contain
a total of 68 files changed, 3592 insertions(+), 1371 deletions(-).
The main changes are:
1) Veristat tool improvements to support custom filtering, sorting, and replay
of results, from Andrii Nakryiko.
2) BPF verifier precision tracking fixes and improvements,
from Andrii Nakryiko.
3) Lots of new BPF documentation for various BPF maps, from Dave Tucker,
Donald Hunter, Maryam Tahhan, Bagas Sanjaya.
4) BTF dedup improvements and libbpf's hashmap interface clean ups, from
Eduard Zingerman.
5) Fix veth driver panic if XDP program is attached before veth_open, from
John Fastabend.
6) BPF verifier clean ups and fixes in preparation for follow up features,
from Kumar Kartikeya Dwivedi.
7) Add access to hwtstamp field from BPF sockops programs,
from Martin KaFai Lau.
8) Various fixes for BPF selftests and samples, from Artem Savkov,
Domenico Cerasuolo, Kang Minchul, Rong Tao, Yang Jihong.
9) Fix redirection to tunneling device logic, preventing skb->len == 0, from
Stanislav Fomichev.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (49 commits)
selftests/bpf: fix veristat's singular file-or-prog filter
selftests/bpf: Test skops->skb_hwtstamp
selftests/bpf: Fix incorrect ASSERT in the tcp_hdr_options test
bpf: Add hwtstamp field for the sockops prog
selftests/bpf: Fix xdp_synproxy compilation failure in 32-bit arch
bpf, docs: Document BPF_MAP_TYPE_ARRAY
docs/bpf: Document BPF map types QUEUE and STACK
docs/bpf: Document BPF ARRAY_OF_MAPS and HASH_OF_MAPS
docs/bpf: Document BPF_MAP_TYPE_CPUMAP map
docs/bpf: Document BPF_MAP_TYPE_LPM_TRIE map
libbpf: Hashmap.h update to fix build issues using LLVM14
bpf: veth driver panics when xdp prog attached before veth_open
selftests: Fix test group SKIPPED result
selftests/bpf: Tests for btf_dedup_resolve_fwds
libbpf: Resolve unambigous forward declarations
libbpf: Hashmap interface update to allow both long and void* keys/values
samples/bpf: Fix sockex3 error: Missing BPF prog type
selftests/bpf: Fix u32 variable compared with less than zero
Documentation: bpf: Escape underscore in BPF type name prefix
selftests/bpf: Use consistent build-id type for liburandom_read.so
...
====================
Link: https://lore.kernel.org/r/20221111233733.1088228-1-andrii@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
We can remove a conditional test in gro_list_prepare()
by comparing vlan_all fields of the two skbs.
Notes:
While comparing the vlan_proto is not strictly needed,
because part of the following compare_ether_header() call,
using 32bit word is actually faster than using 16bit values.
napi_reuse_skb() makes sure to clear skb->vlan_all,
as it already calls __vlan_hwaccel_clear_tag()
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
skb->vlan_present seems redundant.
We can instead derive it from this boolean expression:
vlan_present = skb->vlan_proto != 0 || skb->vlan_tci != 0
Add a new union, to access both fields in a single load/store
when possible.
union {
u32 vlan_all;
struct {
__be16 vlan_proto;
__u16 vlan_tci;
};
};
This allows following patch to remove a conditional test in GRO stack.
Note:
We move remcsum_offload to keep TC_AT_INGRESS_MASK
and SKB_MONO_DELIVERY_TIME_MASK unchanged.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Acked-by: Yonghong Song <yhs@fb.com>
Acked-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The bpf-tc prog has already been able to access the
skb_hwtstamps(skb)->hwtstamp. This patch extends the same hwtstamp
access to the sockops prog.
In sockops, the skb is also available to the bpf prog during
the BPF_SOCK_OPS_PARSE_HDR_OPT_CB event. There is a use case
that the hwtstamp will be useful to the sockops prog to better
measure the one-way-delay when the sender has put the tx
timestamp in the tcp header option.
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Andrii Nakryiko <andrii@kernel.org>
Acked-by: Yonghong Song <yhs@fb.com>
Link: https://lore.kernel.org/bpf/20221107230420.4192307-2-martin.lau@linux.dev
When a devlink port is unregistered, its type is expected to be unset or
otherwise a WARNING is generated [1]. This was supposed to be handled by
cited commit by clearing the type upon 'NETDEV_PRE_UNINIT'.
The assumption was that no other events can be generated for the netdev
after this event, but this proved to be wrong. After the event is
generated, netdev_wait_allrefs_any() will rebroadcast a
'NETDEV_UNREGISTER' until the netdev's reference count drops to 1. This
causes devlink to set the port type back to Ethernet.
Fix by only setting and clearing the port type upon 'NETDEV_POST_INIT'
and 'NETDEV_PRE_UNINIT', respectively. For all other events, preserve
the port type.
[1]
WARNING: CPU: 0 PID: 11 at net/core/devlink.c:9998 devl_port_unregister+0x2f6/0x390 net/core/devlink.c:9998
Modules linked in:
CPU: 1 PID: 11 Comm: kworker/u4:1 Not tainted 6.1.0-rc3-next-20221107-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 10/26/2022
Workqueue: netns cleanup_net
RIP: 0010:devl_port_unregister+0x2f6/0x390 net/core/devlink.c:9998
[...]
Call Trace:
<TASK>
__nsim_dev_port_del+0x1bb/0x240 drivers/net/netdevsim/dev.c:1433
nsim_dev_port_del_all drivers/net/netdevsim/dev.c:1443 [inline]
nsim_dev_reload_destroy+0x171/0x510 drivers/net/netdevsim/dev.c:1660
nsim_dev_reload_down+0x6b/0xd0 drivers/net/netdevsim/dev.c:968
devlink_reload+0x1c2/0x6b0 net/core/devlink.c:4501
devlink_pernet_pre_exit+0x104/0x1c0 net/core/devlink.c:12609
ops_pre_exit_list net/core/net_namespace.c:159 [inline]
cleanup_net+0x451/0xb10 net/core/net_namespace.c:594
process_one_work+0x9bf/0x1710 kernel/workqueue.c:2289
worker_thread+0x665/0x1080 kernel/workqueue.c:2436
kthread+0x2e4/0x3a0 kernel/kthread.c:376
ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:308
</TASK>
Fixes: 02a68a47ea ("net: devlink: track netdev with devlink_port assigned")
Reported-by: syzbot+85e47e1a08b3e159b159@syzkaller.appspotmail.com
Reported-by: syzbot+c2ca18f0fccdd1f09c66@syzkaller.appspotmail.com
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Link: https://lore.kernel.org/r/20221110085150.520800-1-idosch@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Add packet traps for 802.1X operation. The "eapol" control trap is used
to trap EAPOL packets and is required for the correct operation of the
control plane. The "locked_port" drop trap can be enabled to gain
visibility into packets that were dropped by the device due to the
locked bridge port check.
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The notifier block tracking netdev changes in devlink is registered
during devlink_alloc() per-net, it is then unregistered
in devlink_free(). When devlink moves from net namespace to another one,
the notifier block needs to move along.
Fix this by adding forgotten call to move the block.
Reported-by: Ido Schimmel <idosch@idosch.org>
Fixes: 02a68a47ea ("net: devlink: track netdev with devlink_port assigned")
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Tested-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, net_dev() netdev notifier variant follows the netdev with
per-net notifier from namespace to namespace. This is implemented
by move_netdevice_notifiers_dev_net() helper.
For devlink it is needed to re-register per-net notifier during
devlink reload. Introduce a new helper called
move_netdevice_notifier_net() and share the unregister/register code
with existing move_netdevice_notifiers_dev_net() helper.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
The callbacks in struct class namespace() and get_ownership() do not
modify the struct device passed to them, so mark the pointer as constant
and fix up all callbacks in the kernel to have the correct function
signature.
This helps make it more obvious what calls and callbacks do, and do not,
modify structures passed to them.
Cc: "Rafael J. Wysocki" <rafael@kernel.org>
Link: https://lore.kernel.org/r/20221001165426.2690912-1-gregkh@linuxfoundation.org
Signed-off-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
rxrpc changes
David Howells says:
====================
rxrpc: Increasing SACK size and moving away from softirq, part 1
AF_RXRPC has some issues that need addressing:
(1) The SACK table has a maximum capacity of 255, but for modern networks
that isn't sufficient. This is hard to increase in the upstream code
because of the way the application thread is coupled to the softirq
and retransmission side through a ring buffer. Adjustments to the rx
protocol allows a capacity of up to 8192, and having a ring
sufficiently large to accommodate that would use an excessive amount
of memory as this is per-call.
(2) Processing ACKs in softirq mode causes the ACKs get conflated, with
only the most recent being considered. Whilst this has the upside
that the retransmission algorithm only needs to deal with the most
recent ACK, it causes DATA transmission for a call to be very bursty
because DATA packets cannot be transmitted in softirq mode. Rather
transmission must be delegated to either the application thread or a
workqueue, so there tend to be sudden bursts of traffic for any
particular call due to scheduling delays.
(3) All crypto in a single call is done in series; however, each DATA
packet is individually encrypted so encryption and decryption of large
calls could be parallelised if spare CPU resources are available.
This is the first of a number of sets of patches that try and address them.
The overall aims of these changes include:
(1) To get rid of the TxRx ring and instead pass the packets round in
queues (eg. sk_buff_head). On the Tx side, each ACK packet comes with
a SACK table that can be parsed as-is, so there's no particular need
to maintain our own; we just have to refer to the ACK.
On the Rx side, we do need to maintain a SACK table with one bit per
entry - but only if packets go missing - and we don't want to have to
perform a complex transformation to get the information into an ACK
packet.
(2) To try and move almost all processing of received packets out of the
softirq handler and into a high-priority kernel I/O thread. Only the
transferral of packets would be left there. I would still use the
encap_rcv hook to receive packets as there's a noticeable performance
drop from letting the UDP socket put the packets into its own queue
and then getting them out of there.
(3) To make the I/O thread also do all the transmission. The app thread
would be responsible for packaging the data into packets and then
buffering them for the I/O thread to transmit. This would make it
easier for the app thread to run ahead of the I/O thread, and would
mean the I/O thread is less likely to have to wait around for a new
packet to come available for transmission.
(4) To logically partition the socket/UAPI/KAPI side of things from the
I/O side of things. The local endpoint, connection, peer and call
objects would belong to the I/O side. The socket side would not then
touch the private internals of calls and suchlike and would not change
their states. It would only look at the send queue, receive queue and
a way to pass a message to cause an abort.
(5) To remove as much locking, synchronisation, barriering and atomic ops
as possible from the I/O side. Exclusion would be achieved by
limiting modification of state to the I/O thread only. Locks would
still need to be used in communication with the UDP socket and the
AF_RXRPC socket API.
(6) To provide crypto offload kernel threads that, when there's slack in
the system, can see packets that need crypting and provide
parallelisation in dealing with them.
(7) To remove the use of system timers. Since each timer would then send
a poke to the I/O thread, which would then deal with it when it had
the opportunity, there seems no point in using system timers if,
instead, a list of timeouts can be sensibly consulted. An I/O thread
only then needs to schedule with a timeout when it is idle.
(8) To use zero-copy sendmsg to send packets. This would make use of the
I/O thread being the sole transmitter on the socket to manage the
dead-reckoning sequencing of the completion notifications. There is a
problem with zero-copy, though: the UDP socket doesn't handle running
out of option memory very gracefully.
With regard to this first patchset, the changes made include:
(1) Some fixes, including a fallback for proc_create_net_single_write(),
setting ack.bufferSize to 0 in ACK packets and a fix for rxrpc
congestion management, which shouldn't be saving the cwnd value
between calls.
(2) Improvements in rxrpc tracepoints, including splitting the timer
tracepoint into a set-timer and a timer-expired trace.
(3) Addition of a new proc file to display some stats.
(4) Some code cleanups, including removing some unused bits and
unnecessary header inclusions.
(5) A change to the recently added UDP encap_err_rcv hook so that it has
the same signature as {ip,ipv6}_icmp_error(), and then just have rxrpc
point its UDP socket's hook directly at those.
(6) Definition of a new struct, rxrpc_txbuf, that is used to hold
transmissible packets of DATA and ACK type in a single 2KiB block
rather than using an sk_buff. This allows the buffer to be on a
number of queues simultaneously more easily, and also guarantees that
the entire block is in a single unit for zerocopy purposes and that
the data payload is aligned for in-place crypto purposes.
(7) ACK txbufs are allocated at proposal and queued for later transmission
rather than being stored in a single place in the rxrpc_call struct,
which means only a single ACK can be pending transmission at a time.
The queue is then drained at various points. This allows the ACK
generation code to be simplified.
(8) The Rx ring buffer is removed. When a jumbo packet is received (which
comprises a number of ordinary DATA packets glued together), it used
to be pointed to by the ring multiple times, with an annotation in a
side ring indicating which subpacket was in that slot - but this is no
longer possible. Instead, the packet is cloned once for each
subpacket, barring the last, and the range of data is set in the skb
private area. This makes it easier for the subpackets in a jumbo
packet to be decrypted in parallel.
(9) The Tx ring buffer is removed. The side annotation ring that held the
SACK information is also removed. Instead, in the event of packet
loss, the SACK data attached an ACK packet is parsed.
(10) Allocate an skcipher request when needed in the rxkad security class
rather than caching one in the rxrpc_call struct. This deals with a
race between externally-driven call disconnection getting rid of the
skcipher request and sendmsg/recvmsg trying to use it because they
haven't seen the completion yet. This is also needed to support
parallelisation as the skcipher request cannot be used by two or more
threads simultaneously.
(11) Call udp_sendmsg() and udpv6_sendmsg() directly rather than going
through kernel_sendmsg() so that we can provide our own iterator
(zerocopy explicitly doesn't work with a KVEC iterator). This also
lets us avoid the overhead of the security hook.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Allow a network interface to be renamed when the interface
is up.
As described in the netconsole documentation [1], when netconsole is
used as a built-in, it will bring up the specified interface as soon as
possible. As a result, user space will not be able to rename the
interface since the kernel disallows renaming of interfaces that are
administratively up unless the 'IFF_LIVE_RENAME_OK' private flag was set
by the kernel.
The original solution [2] to this problem was to add a new parameter to
the netconsole configuration parameters that allows renaming of
the interface used by netconsole while it is administratively up.
However, during the discussion that followed, it became apparent that we
have no reason to keep the current restriction and instead we should
allow user space to rename interfaces regardless of their administrative
state:
1. The restriction was put in place over 20 years ago when renaming was
only possible via IOCTL and before rtnetlink started notifying user
space about such changes like it does today.
2. The 'IFF_LIVE_RENAME_OK' flag was added over 3 years ago in version
5.2 and no regressions were reported.
3. In-kernel listeners to 'NETDEV_CHANGENAME' do not seem to care about
the administrative state of interface.
Therefore, allow user space to rename running interfaces by removing the
restriction and the associated 'IFF_LIVE_RENAME_OK' flag. Help in
possible triage by emitting a message to the kernel log that an
interface was renamed while UP.
[1] https://www.kernel.org/doc/Documentation/networking/netconsole.rst
[2] https://lore.kernel.org/netdev/20221102002420.2613004-1-andy.ren@getcruise.com/
Signed-off-by: Andy Ren <andy.ren@getcruise.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
Improve the tracking of which packets need to be transmitted by saving the
last ACK packet that we receive that has a populated soft-ACK table rather
than marking packets. Then we can step through the soft-ACK table and look
at the packets we've transmitted beyond that to determine which packets we
might want to retransmit.
We also look at the highest serial number that has been acked to try and
guess which packets we've transmitted the peer is likely to have seen. If
necessary, we send a ping to retrieve that number.
One downside that might be a problem is that we can't then compare the
previous acked/unacked state so easily in rxrpc_input_soft_acks() - which
is a potential problem for the slow-start algorithm.
Signed-off-by: David Howells <dhowells@redhat.com>
cc: Marc Dionne <marc.dionne@auristor.com>
cc: linux-afs@lists.infradead.org
We currently have two forms of operations - small ops and "full" ops
(or just ops). The former does not have pointers for some of the less
commonly used features (namely dump start/done and policy).
The "full" ops, however, still don't contain all the necessary
information. In particular the policy is per command ID, while
do and dump often accept different attributes. It's also not
possible to define different pre_doit and post_doit callbacks
for different commands within the family.
At the same time a lot of commands do not support dumping and
therefore all the dump-related information is wasted space.
Create a new command representation which can hold info about
a do implementation or a dump implementation, but not both at
the same time.
Use this new representation on the command execution path
(genl_family_rcv_msg) as we either run a do or a dump and
don't have to create a "full" op there.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
Commit 4ff09db1b7 ("bpf: net: Change sk_getsockopt() to take the
sockptr_t argument") made it possible to call sk_getsockopt()
with both user and kernel address space buffers through the use of
the sockptr_t type. Unfortunately at the time of conversion the
security_socket_getpeersec_stream() LSM hook was written to only
accept userspace buffers, and in a desire to avoid having to change
the LSM hook the commit author simply passed the sockptr_t's
userspace buffer pointer. Since the only sk_getsockopt() callers
at the time of conversion which used kernel sockptr_t buffers did
not allow SO_PEERSEC, and hence the
security_socket_getpeersec_stream() hook, this was acceptable but
also very fragile as future changes presented the possibility of
silently passing kernel space pointers to the LSM hook.
There are several ways to protect against this, including careful
code review of future commits, but since relying on code review to
catch bugs is a recipe for disaster and the upstream eBPF maintainer
is "strongly against defensive programming", this patch updates the
LSM hook, and all of the implementations to support sockptr_t and
safely handle both user and kernel space buffers.
Acked-by: Casey Schaufler <casey@schaufler-ca.com>
Acked-by: John Johansen <john.johansen@canonical.com>
Signed-off-by: Paul Moore <paul@paul-moore.com>
Now that kptr_off_tab has been refactored into btf_record, and can hold
more than one specific field type, accomodate bpf_spin_lock and
bpf_timer as well.
While they don't require any more metadata than offset, having all
special fields in one place allows us to share the same code for
allocated user defined types and handle both map values and these
allocated objects in a similar fashion.
As an optimization, we still keep spin_lock_off and timer_off offsets in
the btf_record structure, just to avoid having to find the btf_field
struct each time their offset is needed. This is mostly needed to
manipulate such objects in a map value at runtime. It's ok to hardcode
just one offset as more than one field is disallowed.
Signed-off-by: Kumar Kartikeya Dwivedi <memxor@gmail.com>
Link: https://lore.kernel.org/r/20221103191013.1236066-8-memxor@gmail.com
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Since commit 3dcbdb134f ("net: gso: Fix skb_segment splat when
splitting gso_size mangled skb having linear-headed frag_list"), it is
allowed to change gso_size of a GRO packet. However, that commit assumes
that "checking the first list_skb member suffices; i.e if either of the
list_skb members have non head_frag head, then the first one has too".
It turns out this assumption does not hold. We've seen BUG_ON being hit
in skb_segment when skbs on the frag_list had differing head_frag with
the vmxnet3 driver. This happens because __netdev_alloc_skb and
__napi_alloc_skb can return a skb that is page backed or kmalloced
depending on the requested size. As the result, the last small skb in
the GRO packet can be kmalloced.
There are three different locations where this can be fixed:
(1) We could check head_frag in GRO and not allow GROing skbs with
different head_frag. However, that would lead to performance
regression on normal forward paths with unmodified gso_size, where
!head_frag in the last packet is not a problem.
(2) Set a flag in bpf_skb_net_grow and bpf_skb_net_shrink indicating
that NETIF_F_SG is undesirable. That would need to eat a bit in
sk_buff. Furthermore, that flag can be unset when all skbs on the
frag_list are page backed. To retain good performance,
bpf_skb_net_grow/shrink would have to walk the frag_list.
(3) Walk the frag_list in skb_segment when determining whether
NETIF_F_SG should be cleared. This of course slows things down.
This patch implements (3). To limit the performance impact in
skb_segment, the list is walked only for skbs with SKB_GSO_DODGY set
that have gso_size changed. Normal paths thus will not hit it.
We could check only the last skb but since we need to walk the whole
list anyway, let's stay on the safe side.
Fixes: 3dcbdb134f ("net: gso: Fix skb_segment splat when splitting gso_size mangled skb having linear-headed frag_list")
Signed-off-by: Jiri Benc <jbenc@redhat.com>
Reviewed-by: Willem de Bruijn <willemb@google.com>
Link: https://lore.kernel.org/r/e04426a6a91baf4d1081e1b478c82b5de25fdf21.1667407944.git.jbenc@redhat.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Expose devlink port handle related to netdev over rtnetlink. Introduce a
new nested IFLA attribute to carry the info. Call into devlink code to
fill-up the nest with existing devlink attributes that are used over
devlink netlink.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Use newly introduced devlink_port pointer instead of getting it calling
to ndo_get_devlink_port op.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
By the time port unregister is called. There should be no type set. Make
sure that the driver cleared it before and warn in case it didn't. This
enforces symmetricity with type set and port register.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
To avoid a need to take RTNL mutex in port_fill() function, benefit from
the introduce infrastructure that tracks netdevice notifier events.
Store the ifindex and ifname upon register and change name events.
Remove the rtnl_held bool propagated down to port_fill() function as it
is no longer needed.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
It is ensured by the netdevice notifier event processing, that only
netdev pointers from the same net namespaces are filled. Remove the
net namespace check from devlink_nl_port_fill() as it is no longer
needed.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Since devlink_port_type_eth_set() should no longer be called by any
driver with netdev pointer as it should rather use
SET_NETDEV_DEVLINK_PORT, remove the netdev arg. Add a warn to
type_clear()
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Currently, ethernet drivers are using devlink_port_type_eth_set() and
devlink_port_type_clear() to set devlink port type and link to related
netdev.
Instead of calling them directly, let the driver use
SET_NETDEV_DEVLINK_PORT macro to assign devlink_port pointer and let
devlink to track it. Note the devlink port pointer is static during
the time netdevice is registered.
In devlink code, use per-namespace netdev notifier to track
the netdevices with devlink_port assigned and change the internal
devlink_port type and related type pointer accordingly.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Follow-up patch is going to introduce a netdevice notifier event
processing which is called with RTNL mutex held. Processing of this will
eventually lead to call to port_notity() and port_fill() which currently
takes RTNL mutex internally. So as a temporary solution, propagate a
bool indicating if the mutex is already held. This will go away in one
of the follow-up patches.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As __devlink_port_type_set() is going to be called directly from netdevice
notifier event handle in one of the follow-up patches, move the
port_type_netdev_checks() call there.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
As __devlink_port_type_set() is going to be called directly from netdevice
notifier event handle in one of the follow-up patches, move the
port_type_warn_schedule() call there.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Instead of storing type_dev as a void pointer, convert it to union and
use it to store either struct net_device or struct ib_device pointer.
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Hosts that support 802.1X authentication are able to authenticate
themselves by exchanging EAPOL frames with an authenticator (Ethernet
bridge, in this case) and an authentication server. Access to the
network is only granted by the authenticator to successfully
authenticated hosts.
The above is implemented in the bridge using the "locked" bridge port
option. When enabled, link-local frames (e.g., EAPOL) can be locally
received by the bridge, but all other frames are dropped unless the host
is authenticated. That is, unless the user space control plane installed
an FDB entry according to which the source address of the frame is
located behind the locked ingress port. The entry can be dynamic, in
which case learning needs to be enabled so that the entry will be
refreshed by incoming traffic.
There are deployments in which not all the devices connected to the
authenticator (the bridge) support 802.1X. Such devices can include
printers and cameras. One option to support such deployments is to
unlock the bridge ports connecting these devices, but a slightly more
secure option is to use MAB. When MAB is enabled, the MAC address of the
connected device is used as the user name and password for the
authentication.
For MAB to work, the user space control plane needs to be notified about
MAC addresses that are trying to gain access so that they will be
compared against an allow list. This can be implemented via the regular
learning process with the sole difference that learned FDB entries are
installed with a new "locked" flag indicating that the entry cannot be
used to authenticate the device. The flag cannot be set by user space,
but user space can clear the flag by replacing the entry, thereby
authenticating the device.
Locked FDB entries implement the following semantics with regards to
roaming, aging and forwarding:
1. Roaming: Locked FDB entries can roam to unlocked (authorized) ports,
in which case the "locked" flag is cleared. FDB entries cannot roam
to locked ports regardless of MAB being enabled or not. Therefore,
locked FDB entries are only created if an FDB entry with the given {MAC,
VID} does not already exist. This behavior prevents unauthenticated
devices from disrupting traffic destined to already authenticated
devices.
2. Aging: Locked FDB entries age and refresh by incoming traffic like
regular entries.
3. Forwarding: Locked FDB entries forward traffic like regular entries.
If user space detects an unauthorized MAC behind a locked port and
wishes to prevent traffic with this MAC DA from reaching the host, it
can do so using tc or a different mechanism.
Enable the above behavior using a new bridge port option called "mab".
It can only be enabled on a bridge port that is both locked and has
learning enabled. Locked FDB entries are flushed from the port once MAB
is disabled. A new option is added because there are pure 802.1X
deployments that are not interested in notifications about locked FDB
entries.
Signed-off-by: Hans J. Schultz <netdev@kapio-technology.com>
Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Acked-by: Nikolay Aleksandrov <razor@blackwall.org>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Daniel Borkmann says:
====================
bpf 2022-11-04
We've added 8 non-merge commits during the last 3 day(s) which contain
a total of 10 files changed, 113 insertions(+), 16 deletions(-).
The main changes are:
1) Fix memory leak upon allocation failure in BPF verifier's stack state
tracking, from Kees Cook.
2) Fix address leakage when BPF progs release reference to an object,
from Youlin Li.
3) Fix BPF CI breakage from buggy in.h uapi header dependency,
from Andrii Nakryiko.
4) Fix bpftool pin sub-command's argument parsing, from Pu Lehui.
5) Fix BPF sockmap lockdep warning by cancelling psock work outside
of socket lock, from Cong Wang.
6) Follow-up for BPF sockmap to fix sk_forward_alloc accounting,
from Wang Yufen.
bpf-for-netdev
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf:
selftests/bpf: Add verifier test for release_reference()
bpf: Fix wrong reg type conversion in release_reference()
bpf, sock_map: Move cancel_work_sync() out of sock lock
tools/headers: Pull in stddef.h to uapi to fix BPF selftests build in CI
net/ipv4: Fix linux/in.h header dependencies
bpftool: Fix NULL pointer dereference when pin {PROG, MAP, LINK} without FILE
bpf, sockmap: Fix the sk->sk_forward_alloc warning of sk_stream_kill_queues
bpf, verifier: Fix memory leak in array reallocation for stack state
====================
Link: https://lore.kernel.org/r/20221104000445.30761-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
syzkaller managed to trigger another case where skb->len == 0
when we enter __dev_queue_xmit:
WARNING: CPU: 0 PID: 2470 at include/linux/skbuff.h:2576 skb_assert_len include/linux/skbuff.h:2576 [inline]
WARNING: CPU: 0 PID: 2470 at include/linux/skbuff.h:2576 __dev_queue_xmit+0x2069/0x35e0 net/core/dev.c:4295
Call Trace:
dev_queue_xmit+0x17/0x20 net/core/dev.c:4406
__bpf_tx_skb net/core/filter.c:2115 [inline]
__bpf_redirect_no_mac net/core/filter.c:2140 [inline]
__bpf_redirect+0x5fb/0xda0 net/core/filter.c:2163
____bpf_clone_redirect net/core/filter.c:2447 [inline]
bpf_clone_redirect+0x247/0x390 net/core/filter.c:2419
bpf_prog_48159a89cb4a9a16+0x59/0x5e
bpf_dispatcher_nop_func include/linux/bpf.h:897 [inline]
__bpf_prog_run include/linux/filter.h:596 [inline]
bpf_prog_run include/linux/filter.h:603 [inline]
bpf_test_run+0x46c/0x890 net/bpf/test_run.c:402
bpf_prog_test_run_skb+0xbdc/0x14c0 net/bpf/test_run.c:1170
bpf_prog_test_run+0x345/0x3c0 kernel/bpf/syscall.c:3648
__sys_bpf+0x43a/0x6c0 kernel/bpf/syscall.c:5005
__do_sys_bpf kernel/bpf/syscall.c:5091 [inline]
__se_sys_bpf kernel/bpf/syscall.c:5089 [inline]
__x64_sys_bpf+0x7c/0x90 kernel/bpf/syscall.c:5089
do_syscall_64+0x54/0x70 arch/x86/entry/common.c:48
entry_SYSCALL_64_after_hwframe+0x61/0xc6
The reproducer doesn't really reproduce outside of syzkaller
environment, so I'm taking a guess here. It looks like we
do generate correct ETH_HLEN-sized packet, but we redirect
the packet to the tunneling device. Before we do so, we
__skb_pull l2 header and arrive again at skb->len == 0.
Doesn't seem like we can do anything better than having
an explicit check after __skb_pull?
Cc: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot+f635e86ec3fa0a37e019@syzkaller.appspotmail.com
Signed-off-by: Stanislav Fomichev <sdf@google.com>
Link: https://lore.kernel.org/r/20221027225537.353077-1-sdf@google.com
Signed-off-by: Martin KaFai Lau <martin.lau@kernel.org>
Signed-off-by: Alexei Starovoitov <ast@kernel.org>
Stanislav reported a lockdep warning, which is caused by the
cancel_work_sync() called inside sock_map_close(), as analyzed
below by Jakub:
psock->work.func = sk_psock_backlog()
ACQUIRE psock->work_mutex
sk_psock_handle_skb()
skb_send_sock()
__skb_send_sock()
sendpage_unlocked()
kernel_sendpage()
sock->ops->sendpage = inet_sendpage()
sk->sk_prot->sendpage = tcp_sendpage()
ACQUIRE sk->sk_lock
tcp_sendpage_locked()
RELEASE sk->sk_lock
RELEASE psock->work_mutex
sock_map_close()
ACQUIRE sk->sk_lock
sk_psock_stop()
sk_psock_clear_state(psock, SK_PSOCK_TX_ENABLED)
cancel_work_sync()
__cancel_work_timer()
__flush_work()
// wait for psock->work to finish
RELEASE sk->sk_lock
We can move the cancel_work_sync() out of the sock lock protection,
but still before saved_close() was called.
Fixes: 799aa7f98d ("skmsg: Avoid lock_sock() in sk_psock_backlog()")
Reported-by: Stanislav Fomichev <sdf@google.com>
Signed-off-by: Cong Wang <cong.wang@bytedance.com>
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Tested-by: Jakub Sitnicki <jakub@cloudflare.com>
Acked-by: John Fastabend <john.fastabend@gmail.com>
Acked-by: Jakub Sitnicki <jakub@cloudflare.com>
Link: https://lore.kernel.org/bpf/20221102043417.279409-1-xiyou.wangcong@gmail.com
When IPv6 module gets initialized but hits an error in the middle,
kenel panic with:
KASAN: null-ptr-deref in range [0x0000000000000598-0x000000000000059f]
CPU: 1 PID: 361 Comm: insmod
Hardware name: QEMU Standard PC (i440FX + PIIX, 1996)
RIP: 0010:__neigh_ifdown.isra.0+0x24b/0x370
RSP: 0018:ffff888012677908 EFLAGS: 00000202
...
Call Trace:
<TASK>
neigh_table_clear+0x94/0x2d0
ndisc_cleanup+0x27/0x40 [ipv6]
inet6_init+0x21c/0x2cb [ipv6]
do_one_initcall+0xd3/0x4d0
do_init_module+0x1ae/0x670
...
Kernel panic - not syncing: Fatal exception
When ipv6 initialization fails, it will try to cleanup and calls:
neigh_table_clear()
neigh_ifdown(tbl, NULL)
pneigh_queue_purge(&tbl->proxy_queue, dev_net(dev == NULL))
# dev_net(NULL) triggers null-ptr-deref.
Fix it by passing NULL to pneigh_queue_purge() in neigh_ifdown() if dev
is NULL, to make kernel not panic immediately.
Fixes: 66ba215cb5 ("neigh: fix possible DoS due to net iface start/stop loop")
Signed-off-by: Chen Zhongjin <chenzhongjin@huawei.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Denis V. Lunev <den@openvz.org>
Link: https://lore.kernel.org/r/20221101121552.21890-1-chenzhongjin@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Daniel Borkmann says:
====================
bpf-next 2022-11-02
We've added 70 non-merge commits during the last 14 day(s) which contain
a total of 96 files changed, 3203 insertions(+), 640 deletions(-).
The main changes are:
1) Make cgroup local storage available to non-cgroup attached BPF programs
such as tc BPF ones, from Yonghong Song.
2) Avoid unnecessary deadlock detection and failures wrt BPF task storage
helpers, from Martin KaFai Lau.
3) Add LLVM disassembler as default library for dumping JITed code
in bpftool, from Quentin Monnet.
4) Various kprobe_multi_link fixes related to kernel modules,
from Jiri Olsa.
5) Optimize x86-64 JIT with emitting BMI2-based shift instructions,
from Jie Meng.
6) Improve BPF verifier's memory type compatibility for map key/value
arguments, from Dave Marchevsky.
7) Only create mmap-able data section maps in libbpf when data is exposed
via skeletons, from Andrii Nakryiko.
8) Add an autoattach option for bpftool to load all object assets,
from Wang Yufen.
9) Various memory handling fixes for libbpf and BPF selftests,
from Xu Kuohai.
10) Initial support for BPF selftest's vmtest.sh on arm64,
from Manu Bretelle.
11) Improve libbpf's BTF handling to dedup identical structs,
from Alan Maguire.
12) Add BPF CI and denylist documentation for BPF selftests,
from Daniel Müller.
13) Check BPF cpumap max_entries before doing allocation work,
from Florian Lehner.
* tag 'for-netdev' of https://git.kernel.org/pub/scm/linux/kernel/git/bpf/bpf-next: (70 commits)
samples/bpf: Fix typo in README
bpf: Remove the obsolte u64_stats_fetch_*_irq() users.
bpf: check max_entries before allocating memory
bpf: Fix a typo in comment for DFS algorithm
bpftool: Fix spelling mistake "disasembler" -> "disassembler"
selftests/bpf: Fix bpftool synctypes checking failure
selftests/bpf: Panic on hard/soft lockup
docs/bpf: Add documentation for new cgroup local storage
selftests/bpf: Add test cgrp_local_storage to DENYLIST.s390x
selftests/bpf: Add selftests for new cgroup local storage
selftests/bpf: Fix test test_libbpf_str/bpf_map_type_str
bpftool: Support new cgroup local storage
libbpf: Support new cgroup local storage
bpf: Implement cgroup storage available to non-cgroup-attached bpf progs
bpf: Refactor some inode/task/sk storage functions for reuse
bpf: Make struct cgroup btf id global
selftests/bpf: Tracing prog can still do lookup under busy lock
selftests/bpf: Ensure no task storage failure for bpf_lsm.s prog due to deadlock detection
bpf: Add new bpf_task_storage_delete proto with no deadlock detection
bpf: bpf_task_storage_delete_recur does lookup first before the deadlock check
...
====================
Link: https://lore.kernel.org/r/20221102062120.5724-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
inet[46]_pton check the input length against
a sane length limit (INET[6]_ADDRSTRLEN), but
the strlen value gets truncated due to being stored in an int,
so there's a theoretical potential for a >4G string to pass
the limit test.
Use size_t since that's what strlen actually returns.
I've had a hunt for callers that could hit this, but
I've not managed to find anything that doesn't get checked with
some other limit first; but it's possible that I've missed
something in the depth of the storage target paths.
Signed-off-by: Dr. David Alan Gilbert <linux@treblig.org>
Link: https://lore.kernel.org/r/20221029014604.114024-1-linux@treblig.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
