platform/kernel/linux-starfive.git
2 years agonet, neigh: Do not trigger immediate probes on NUD_FAILED from neigh_managed_work
Daniel Borkmann [Tue, 1 Feb 2022 19:39:42 +0000 (20:39 +0100)]
net, neigh: Do not trigger immediate probes on NUD_FAILED from neigh_managed_work

syzkaller was able to trigger a deadlock for NTF_MANAGED entries [0]:

  kworker/0:16/14617 is trying to acquire lock:
  ffffffff8d4dd370 (&tbl->lock){++-.}-{2:2}, at: ___neigh_create+0x9e1/0x2990 net/core/neighbour.c:652
  [...]
  but task is already holding lock:
  ffffffff8d4dd370 (&tbl->lock){++-.}-{2:2}, at: neigh_managed_work+0x35/0x250 net/core/neighbour.c:1572

The neighbor entry turned to NUD_FAILED state, where __neigh_event_send()
triggered an immediate probe as per commit cd28ca0a3dd1 ("neigh: reduce
arp latency") via neigh_probe() given table lock was held.

One option to fix this situation is to defer the neigh_probe() back to
the neigh_timer_handler() similarly as pre cd28ca0a3dd1. For the case
of NTF_MANAGED, this deferral is acceptable given this only happens on
actual failure state and regular / expected state is NUD_VALID with the
entry already present.

The fix adds a parameter to __neigh_event_send() in order to communicate
whether immediate probe is allowed or disallowed. Existing call-sites
of neigh_event_send() default as-is to immediate probe. However, the
neigh_managed_work() disables it via use of neigh_event_send_probe().

[0] <TASK>
  __dump_stack lib/dump_stack.c:88 [inline]
  dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
  print_deadlock_bug kernel/locking/lockdep.c:2956 [inline]
  check_deadlock kernel/locking/lockdep.c:2999 [inline]
  validate_chain kernel/locking/lockdep.c:3788 [inline]
  __lock_acquire.cold+0x149/0x3ab kernel/locking/lockdep.c:5027
  lock_acquire kernel/locking/lockdep.c:5639 [inline]
  lock_acquire+0x1ab/0x510 kernel/locking/lockdep.c:5604
  __raw_write_lock_bh include/linux/rwlock_api_smp.h:202 [inline]
  _raw_write_lock_bh+0x2f/0x40 kernel/locking/spinlock.c:334
  ___neigh_create+0x9e1/0x2990 net/core/neighbour.c:652
  ip6_finish_output2+0x1070/0x14f0 net/ipv6/ip6_output.c:123
  __ip6_finish_output net/ipv6/ip6_output.c:191 [inline]
  __ip6_finish_output+0x61e/0xe90 net/ipv6/ip6_output.c:170
  ip6_finish_output+0x32/0x200 net/ipv6/ip6_output.c:201
  NF_HOOK_COND include/linux/netfilter.h:296 [inline]
  ip6_output+0x1e4/0x530 net/ipv6/ip6_output.c:224
  dst_output include/net/dst.h:451 [inline]
  NF_HOOK include/linux/netfilter.h:307 [inline]
  ndisc_send_skb+0xa99/0x17f0 net/ipv6/ndisc.c:508
  ndisc_send_ns+0x3a9/0x840 net/ipv6/ndisc.c:650
  ndisc_solicit+0x2cd/0x4f0 net/ipv6/ndisc.c:742
  neigh_probe+0xc2/0x110 net/core/neighbour.c:1040
  __neigh_event_send+0x37d/0x1570 net/core/neighbour.c:1201
  neigh_event_send include/net/neighbour.h:470 [inline]
  neigh_managed_work+0x162/0x250 net/core/neighbour.c:1574
  process_one_work+0x9ac/0x1650 kernel/workqueue.c:2307
  worker_thread+0x657/0x1110 kernel/workqueue.c:2454
  kthread+0x2e9/0x3a0 kernel/kthread.c:377
  ret_from_fork+0x1f/0x30 arch/x86/entry/entry_64.S:295
  </TASK>

Fixes: 7482e3841d52 ("net, neigh: Add NTF_MANAGED flag for managed neighbor entries")
Reported-by: syzbot+5239d0e1778a500d477a@syzkaller.appspotmail.com
Signed-off-by: Daniel Borkmann <daniel@iogearbox.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Roopa Prabhu <roopa@nvidia.com>
Tested-by: syzbot+5239d0e1778a500d477a@syzkaller.appspotmail.com
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20220201193942.5055-1-daniel@iogearbox.net
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agotcp: add missing tcp_skb_can_collapse() test in tcp_shift_skb_data()
Eric Dumazet [Tue, 1 Feb 2022 18:46:40 +0000 (10:46 -0800)]
tcp: add missing tcp_skb_can_collapse() test in tcp_shift_skb_data()

tcp_shift_skb_data() might collapse three packets into a larger one.

P_A, P_B, P_C  -> P_ABC

Historically, it used a single tcp_skb_can_collapse_to(P_A) call,
because it was enough.

In commit 85712484110d ("tcp: coalesce/collapse must respect MPTCP extensions"),
this call was replaced by a call to tcp_skb_can_collapse(P_A, P_B)

But the now needed test over P_C has been missed.

This probably broke MPTCP.

Then later, commit 9b65b17db723 ("net: avoid double accounting for pure zerocopy skbs")
added an extra condition to tcp_skb_can_collapse(), but the missing call
from tcp_shift_skb_data() is also breaking TCP zerocopy, because P_A and P_C
might have different skb_zcopy_pure() status.

Fixes: 85712484110d ("tcp: coalesce/collapse must respect MPTCP extensions")
Fixes: 9b65b17db723 ("net: avoid double accounting for pure zerocopy skbs")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Mat Martineau <mathew.j.martineau@linux.intel.com>
Cc: Talal Ahmad <talalahmad@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Acked-by: Paolo Abeni <pabeni@redhat.com>
Link: https://lore.kernel.org/r/20220201184640.756716-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: sparx5: do not refer to skb after passing it on
Steen Hegelund [Wed, 2 Feb 2022 08:30:39 +0000 (09:30 +0100)]
net: sparx5: do not refer to skb after passing it on

Do not try to use any SKB fields after the packet has been passed up in the
receive stack.

Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Signed-off-by: Steen Hegelund <steen.hegelund@microchip.com>
Link: https://lore.kernel.org/r/20220202083039.3774851-1-steen.hegelund@microchip.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoPartially revert "net/smc: Add netlink net namespace support"
Dmitry V. Levin [Wed, 2 Feb 2022 03:09:04 +0000 (06:09 +0300)]
Partially revert "net/smc: Add netlink net namespace support"

The change of sizeof(struct smc_diag_linkinfo) by commit 79d39fc503b4
("net/smc: Add netlink net namespace support") introduced an ABI
regression: since struct smc_diag_lgrinfo contains an object of
type "struct smc_diag_linkinfo", offset of all subsequent members
of struct smc_diag_lgrinfo was changed by that change.

As result, applications compiled with the old version
of struct smc_diag_linkinfo will receive garbage in
struct smc_diag_lgrinfo.role if the kernel implements
this new version of struct smc_diag_linkinfo.

Fix this regression by reverting the part of commit 79d39fc503b4 that
changes struct smc_diag_linkinfo.  After all, there is SMC_GEN_NETLINK
interface which is good enough, so there is probably no need to touch
the smc_diag ABI in the first place.

Fixes: 79d39fc503b4 ("net/smc: Add netlink net namespace support")
Signed-off-by: Dmitry V. Levin <ldv@altlinux.org>
Reviewed-by: Karsten Graul <kgraul@linux.ibm.com>
Link: https://lore.kernel.org/r/20220202030904.GA9742@altlinux.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge tag 'mlx5-fixes-2022-02-01' of git://git.kernel.org/pub/scm/linux/kernel/git...
David S. Miller [Wed, 2 Feb 2022 14:19:38 +0000 (14:19 +0000)]
Merge tag 'mlx5-fixes-2022-02-01' of git://git./linux/kernel/git/saeed/linux

Saeed Mahameed says:

====================
mlx5 fixes 2022-02-01

This series provides bug fixes to mlx5 driver.
Please pull and let me know if there is any problem.

Sorry about the long series, but I had to move the top two patches from
net-next to net to help avoiding a build break when kspp branch is merged
into linus-next on next merge window.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Jakub Kicinski [Wed, 2 Feb 2022 05:03:15 +0000 (21:03 -0800)]
Merge branch '1GbE' of git://git./linux/kernel/git/tnguy/net-queue

Tony Nguyen says:

====================
Intel Wired LAN Driver Updates 2022-02-01

This series contains updates to e1000e driver only.

Sasha removes CSME handshake with TGL platform as this is not supported
and is causing hardware unit hangs to be reported.

* '1GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
  e1000e: Handshake with CSME starts from ADL platforms
  e1000e: Separate ADP board type from TGP
====================

Link: https://lore.kernel.org/r/20220201173754.580305-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet/mlx5e: Avoid field-overflowing memcpy()
Kees Cook [Mon, 24 Jan 2022 17:20:28 +0000 (09:20 -0800)]
net/mlx5e: Avoid field-overflowing memcpy()

In preparation for FORTIFY_SOURCE performing compile-time and run-time
field bounds checking for memcpy(), memmove(), and memset(), avoid
intentionally writing across neighboring fields.

Use flexible arrays instead of zero-element arrays (which look like they
are always overflowing) and split the cross-field memcpy() into two halves
that can be appropriately bounds-checked by the compiler.

We were doing:

#define ETH_HLEN  14
#define VLAN_HLEN  4
...
#define MLX5E_XDP_MIN_INLINE (ETH_HLEN + VLAN_HLEN)
...
        struct mlx5e_tx_wqe      *wqe  = mlx5_wq_cyc_get_wqe(wq, pi);
...
        struct mlx5_wqe_eth_seg  *eseg = &wqe->eth;
        struct mlx5_wqe_data_seg *dseg = wqe->data;
...
memcpy(eseg->inline_hdr.start, xdptxd->data, MLX5E_XDP_MIN_INLINE);

target is wqe->eth.inline_hdr.start (which the compiler sees as being
2 bytes in size), but copying 18, intending to write across start
(really vlan_tci, 2 bytes). The remaining 16 bytes get written into
wqe->data[0], covering byte_count (4 bytes), lkey (4 bytes), and addr
(8 bytes).

struct mlx5e_tx_wqe {
        struct mlx5_wqe_ctrl_seg   ctrl;                 /*     0    16 */
        struct mlx5_wqe_eth_seg    eth;                  /*    16    16 */
        struct mlx5_wqe_data_seg   data[];               /*    32     0 */

        /* size: 32, cachelines: 1, members: 3 */
        /* last cacheline: 32 bytes */
};

struct mlx5_wqe_eth_seg {
        u8                         swp_outer_l4_offset;  /*     0     1 */
        u8                         swp_outer_l3_offset;  /*     1     1 */
        u8                         swp_inner_l4_offset;  /*     2     1 */
        u8                         swp_inner_l3_offset;  /*     3     1 */
        u8                         cs_flags;             /*     4     1 */
        u8                         swp_flags;            /*     5     1 */
        __be16                     mss;                  /*     6     2 */
        __be32                     flow_table_metadata;  /*     8     4 */
        union {
                struct {
                        __be16     sz;                   /*    12     2 */
                        u8         start[2];             /*    14     2 */
                } inline_hdr;                            /*    12     4 */
                struct {
                        __be16     type;                 /*    12     2 */
                        __be16     vlan_tci;             /*    14     2 */
                } insert;                                /*    12     4 */
                __be32             trailer;              /*    12     4 */
        };                                               /*    12     4 */

        /* size: 16, cachelines: 1, members: 9 */
        /* last cacheline: 16 bytes */
};

struct mlx5_wqe_data_seg {
        __be32                     byte_count;           /*     0     4 */
        __be32                     lkey;                 /*     4     4 */
        __be64                     addr;                 /*     8     8 */

        /* size: 16, cachelines: 1, members: 3 */
        /* last cacheline: 16 bytes */
};

So, split the memcpy() so the compiler can reason about the buffer
sizes.

"pahole" shows no size nor member offset changes to struct mlx5e_tx_wqe
nor struct mlx5e_umr_wqe. "objdump -d" shows no meaningful object
code changes (i.e. only source line number induced differences and
optimizations).

Fixes: b5503b994ed5 ("net/mlx5e: XDP TX forwarding support")
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Use struct_group() for memcpy() region
Kees Cook [Mon, 24 Jan 2022 17:22:41 +0000 (09:22 -0800)]
net/mlx5e: Use struct_group() for memcpy() region

In preparation for FORTIFY_SOURCE performing compile-time and run-time
field bounds checking for memcpy(), memmove(), and memset(), avoid
intentionally writing across neighboring fields.

Use struct_group() in struct vlan_ethhdr around members h_dest and
h_source, so they can be referenced together. This will allow memcpy()
and sizeof() to more easily reason about sizes, improve readability,
and avoid future warnings about writing beyond the end of h_dest.

"pahole" shows no size nor member offset changes to struct vlan_ethhdr.
"objdump -d" shows no object code changes.

Fixes: 34802a42b352 ("net/mlx5e: Do not modify the TX SKB")
Signed-off-by: Kees Cook <keescook@chromium.org>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Avoid implicit modify hdr for decap drop rule
Roi Dayan [Tue, 1 Feb 2022 13:27:48 +0000 (15:27 +0200)]
net/mlx5e: Avoid implicit modify hdr for decap drop rule

Currently the driver adds implicit modify hdr action for
decap rules on tunnel devices if the port is an ovs port.
This is also done if the action is drop and makes the modify
hdr redundant and also the FW doesn't support it and will generate
a syndrome.

kernel: mlx5_core 0000:08:00.0: mlx5_cmd_check:777:(pid 102063): SET_FLOW_TABLE_ENTRY(0x936) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0x8708c3)

Fix it by adding the implicit modify hdr only for fwd actions.

Fixes: b16eb3c81fe2 ("net/mlx5: Support internal port as decap route device")
Fixes: 077cdda764c7 ("net/mlx5e: TC, Fix memory leak with rules with internal port")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Ariel Levkovich <lariel@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: IPsec: Fix tunnel mode crypto offload for non TCP/UDP traffic
Raed Salem [Thu, 2 Dec 2021 15:49:01 +0000 (17:49 +0200)]
net/mlx5e: IPsec: Fix tunnel mode crypto offload for non TCP/UDP traffic

IPsec Tunnel mode crypto offload software parser (SWP) setting in data
path currently always set the inner L4 offset regardless of the
encapsulated L4 header type and whether it exists in the first place,
this breaks non TCP/UDP traffic as such.

Set the SWP inner L4 offset only when the IPsec tunnel encapsulated L4
header protocol is TCP/UDP.

While at it fix inner ip protocol read for setting MLX5_ETH_WQE_SWP_INNER_L4_UDP
flag to address the case where the ip header protocol is IPv6.

Fixes: f1267798c980 ("net/mlx5: Fix checksum issue of VXLAN and IPsec crypto offload")
Signed-off-by: Raed Salem <raeds@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: IPsec: Fix crypto offload for non TCP/UDP encapsulated traffic
Raed Salem [Thu, 2 Dec 2021 15:43:50 +0000 (17:43 +0200)]
net/mlx5e: IPsec: Fix crypto offload for non TCP/UDP encapsulated traffic

IPsec crypto offload always set the ethernet segment checksum flags with
the inner L4 header checksum flag enabled for encapsulated IPsec offloaded
packet regardless of the encapsulated L4 header type, and even if it
doesn't exists in the first place, this breaks non TCP/UDP traffic as
such.

Set the inner L4 checksum flag only when the encapsulated L4 header
protocol is TCP/UDP using software parser swp_inner_l4_offset field as
indication.

Fixes: 5cfb540ef27b ("net/mlx5e: Set IPsec WAs only in IP's non checksum partial case.")
Signed-off-by: Raed Salem <raeds@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Don't treat small ceil values as unlimited in HTB offload
Maxim Mikityanskiy [Tue, 18 Jan 2022 11:31:54 +0000 (13:31 +0200)]
net/mlx5e: Don't treat small ceil values as unlimited in HTB offload

The hardware spec defines max_average_bw == 0 as "unlimited bandwidth".
max_average_bw is calculated as `ceil / BYTES_IN_MBIT`, which can become
0 when ceil is small, leading to an undesired effect of having no
bandwidth limit.

This commit fixes it by rounding up small values of ceil to 1 Mbit/s.

Fixes: 214baf22870c ("net/mlx5e: Support HTB offload")
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: E-Switch, Fix uninitialized variable modact
Maor Dickman [Sun, 30 Jan 2022 14:00:41 +0000 (16:00 +0200)]
net/mlx5: E-Switch, Fix uninitialized variable modact

The variable modact is not initialized before used in command
modify header allocation which can cause command to fail.

Fix by initializing modact with zeros.

Addresses-Coverity: ("Uninitialized scalar variable")
Fixes: 8f1e0b97cc70 ("net/mlx5: E-Switch, Mark miss packets with new chain id mapping")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Fix handling of wrong devices during bond netevent
Maor Dickman [Thu, 13 Jan 2022 13:11:42 +0000 (15:11 +0200)]
net/mlx5e: Fix handling of wrong devices during bond netevent

Current implementation of bond netevent handler only check if
the handled netdev is VF representor and it missing a check if
the VF representor is on the same phys device of the bond handling
the netevent.

Fix by adding the missing check and optimizing the check if
the netdev is VF representor so it will not access uninitialized
private data and crashes.

BUG: kernel NULL pointer dereference, address: 000000000000036c
PGD 0 P4D 0
Oops: 0000 [#1] SMP NOPTI
Workqueue: eth3bond0 bond_mii_monitor [bonding]
RIP: 0010:mlx5e_is_uplink_rep+0xc/0x50 [mlx5_core]
RSP: 0018:ffff88812d69fd60 EFLAGS: 00010282
RAX: 0000000000000000 RBX: ffff8881cf800000 RCX: 0000000000000000
RDX: ffff88812d69fe10 RSI: 000000000000001b RDI: ffff8881cf800880
RBP: ffff8881cf800000 R08: 00000445cabccf2b R09: 0000000000000008
R10: 0000000000000004 R11: 0000000000000008 R12: ffff88812d69fe10
R13: 00000000fffffffe R14: ffff88820c0f9000 R15: 0000000000000000
FS:  0000000000000000(0000) GS:ffff88846fb00000(0000) knlGS:0000000000000000
CS:  0010 DS: 0000 ES: 0000 CR0: 0000000080050033
CR2: 000000000000036c CR3: 0000000103d80006 CR4: 0000000000370ea0
DR0: 0000000000000000 DR1: 0000000000000000 DR2: 0000000000000000
DR3: 0000000000000000 DR6: 00000000fffe0ff0 DR7: 0000000000000400
Call Trace:
 mlx5e_eswitch_uplink_rep+0x31/0x40 [mlx5_core]
 mlx5e_rep_is_lag_netdev+0x94/0xc0 [mlx5_core]
 mlx5e_rep_esw_bond_netevent+0xeb/0x3d0 [mlx5_core]
 raw_notifier_call_chain+0x41/0x60
 call_netdevice_notifiers_info+0x34/0x80
 netdev_lower_state_changed+0x4e/0xa0
 bond_mii_monitor+0x56b/0x640 [bonding]
 process_one_work+0x1b9/0x390
 worker_thread+0x4d/0x3d0
 ? rescuer_thread+0x350/0x350
 kthread+0x124/0x150
 ? set_kthread_struct+0x40/0x40
 ret_from_fork+0x1f/0x30

Fixes: 7e51891a237f ("net/mlx5e: Use netdev events to set/del egress acl forward-to-vport rule")
Signed-off-by: Maor Dickman <maord@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Fix broken SKB allocation in HW-GRO
Khalid Manaa [Wed, 26 Jan 2022 12:25:55 +0000 (14:25 +0200)]
net/mlx5e: Fix broken SKB allocation in HW-GRO

In case the HW doesn't perform header-data split, it will write the whole
packet into the data buffer in the WQ, in this case the SHAMPO CQE handler
couldn't use the header entry to build the SKB, instead it should allocate
a new memory to build the SKB using the function:
mlx5e_skb_from_cqe_mpwrq_nonlinear.

Fixes: f97d5c2a453e ("net/mlx5e: Add handle SHAMPO cqe support")
Signed-off-by: Khalid Manaa <khalidm@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Fix wrong calculation of header index in HW_GRO
Khalid Manaa [Wed, 26 Jan 2022 12:14:58 +0000 (14:14 +0200)]
net/mlx5e: Fix wrong calculation of header index in HW_GRO

The HW doesn't wrap the CQE.shampo.header_index field according to the
headers buffer size, instead it always increases it until reaching overflow
of u16 size.

Thus the mlx5e_handle_rx_cqe_mpwrq_shampo handler should mask the
CQE header_index field to find the actual header index in the headers buffer.

Fixes: f97d5c2a453e ("net/mlx5e: Add handle SHAMPO cqe support")
Signed-off-by: Khalid Manaa <khalidm@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Bridge, Fix devlink deadlock on net namespace deletion
Roi Dayan [Mon, 24 Jan 2022 11:56:26 +0000 (13:56 +0200)]
net/mlx5: Bridge, Fix devlink deadlock on net namespace deletion

When changing mode to switchdev, rep bridge init registered to netdevice
notifier holds the devlink lock and then takes pernet_ops_rwsem.
At that time deleting a netns holds pernet_ops_rwsem and then takes
the devlink lock.

Example sequence is:
$ ip netns add foo
$ devlink dev eswitch set pci/0000:00:08.0 mode switchdev &
$ ip netns del foo

deleting netns trace:

[ 1185.365555]  ? devlink_pernet_pre_exit+0x74/0x1c0
[ 1185.368331]  ? mutex_lock_io_nested+0x13f0/0x13f0
[ 1185.370984]  ? xt_find_table+0x40/0x100
[ 1185.373244]  ? __mutex_lock+0x24a/0x15a0
[ 1185.375494]  ? net_generic+0xa0/0x1c0
[ 1185.376844]  ? wait_for_completion_io+0x280/0x280
[ 1185.377767]  ? devlink_pernet_pre_exit+0x74/0x1c0
[ 1185.378686]  devlink_pernet_pre_exit+0x74/0x1c0
[ 1185.379579]  ? devlink_nl_cmd_get_dumpit+0x3a0/0x3a0
[ 1185.380557]  ? xt_find_table+0xda/0x100
[ 1185.381367]  cleanup_net+0x372/0x8e0

changing mode to switchdev trace:

[ 1185.411267]  down_write+0x13a/0x150
[ 1185.412029]  ? down_write_killable+0x180/0x180
[ 1185.413005]  register_netdevice_notifier+0x1e/0x210
[ 1185.414000]  mlx5e_rep_bridge_init+0x181/0x360 [mlx5_core]
[ 1185.415243]  mlx5e_uplink_rep_enable+0x269/0x480 [mlx5_core]
[ 1185.416464]  ? mlx5e_uplink_rep_disable+0x210/0x210 [mlx5_core]
[ 1185.417749]  mlx5e_attach_netdev+0x232/0x400 [mlx5_core]
[ 1185.418906]  mlx5e_netdev_attach_profile+0x15b/0x1e0 [mlx5_core]
[ 1185.420172]  mlx5e_netdev_change_profile+0x15a/0x1d0 [mlx5_core]
[ 1185.421459]  mlx5e_vport_rep_load+0x557/0x780 [mlx5_core]
[ 1185.422624]  ? mlx5e_stats_grp_vport_rep_num_stats+0x10/0x10 [mlx5_core]
[ 1185.424006]  mlx5_esw_offloads_rep_load+0xdb/0x190 [mlx5_core]
[ 1185.425277]  esw_offloads_enable+0xd74/0x14a0 [mlx5_core]

Fix this by registering rep bridges for per net netdev notifier
instead of global one, which operats on the net namespace without holding
the pernet_ops_rwsem.

Fixes: 19e9bfa044f3 ("net/mlx5: Bridge, add offload infrastructure")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Vlad Buslov <vladbu@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Fix offloading with ESWITCH_IPV4_TTL_MODIFY_ENABLE
Dima Chumak [Mon, 17 Jan 2022 13:32:16 +0000 (15:32 +0200)]
net/mlx5: Fix offloading with ESWITCH_IPV4_TTL_MODIFY_ENABLE

Only prio 1 is supported for nic mode when there is no ignore flow level
support in firmware. But for switchdev mode, which supports fixed number
of statically pre-allocated prios, this restriction is not relevant so
it can be relaxed.

Fixes: d671e109bd85 ("net/mlx5: Fix tc max supported prio for nic mode")
Signed-off-by: Dima Chumak <dchumak@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: TC, Reject rules with forward and drop actions
Roi Dayan [Mon, 17 Jan 2022 13:00:30 +0000 (15:00 +0200)]
net/mlx5e: TC, Reject rules with forward and drop actions

Such rules are redundant but allowed and passed to the driver.
The driver does not support offloading such rules so return an error.

Fixes: 03a9d11e6eeb ("net/mlx5e: Add TC drop and mirred/redirect action parsing for SRIOV offloads")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Oz Shlomo <ozsh@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Use del_timer_sync in fw reset flow of halting poll
Maher Sanalla [Thu, 13 Jan 2022 13:48:48 +0000 (15:48 +0200)]
net/mlx5: Use del_timer_sync in fw reset flow of halting poll

Substitute del_timer() with del_timer_sync() in fw reset polling
deactivation flow, in order to prevent a race condition which occurs
when del_timer() is called and timer is deactivated while another
process is handling the timer interrupt. A situation that led to
the following call trace:
RIP: 0010:run_timer_softirq+0x137/0x420
<IRQ>
recalibrate_cpu_khz+0x10/0x10
ktime_get+0x3e/0xa0
? sched_clock_cpu+0xb/0xc0
__do_softirq+0xf5/0x2ea
irq_exit_rcu+0xc1/0xf0
sysvec_apic_timer_interrupt+0x9e/0xc0
asm_sysvec_apic_timer_interrupt+0x12/0x20
</IRQ>

Fixes: 38b9f903f22b ("net/mlx5: Handle sync reset request event")
Signed-off-by: Maher Sanalla <msanalla@nvidia.com>
Reviewed-by: Moshe Shemesh <moshe@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: Fix module EEPROM query
Gal Pressman [Sun, 16 Jan 2022 07:07:22 +0000 (09:07 +0200)]
net/mlx5e: Fix module EEPROM query

When querying the module EEPROM, there was a misusage of the 'offset'
variable vs the 'query.offset' field.
Fix that by always using 'offset' and assigning its value to
'query.offset' right before the mcia register read call.

While at it, the cross-pages read size adjustment was changed to be more
intuitive.

Fixes: e19b0a3474ab ("net/mlx5: Refactor module EEPROM query")
Reported-by: Wang Yugui <wangyugui@e16-tech.com>
Signed-off-by: Gal Pressman <gal@nvidia.com>
Reviewed-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5e: TC, Reject rules with drop and modify hdr action
Roi Dayan [Tue, 4 Jan 2022 08:38:02 +0000 (10:38 +0200)]
net/mlx5e: TC, Reject rules with drop and modify hdr action

This kind of action is not supported by firmware and generates a
syndrome.

kernel: mlx5_core 0000:08:00.0: mlx5_cmd_check:777:(pid 102063): SET_FLOW_TABLE_ENTRY(0x936) op_mod(0x0) failed, status bad parameter(0x3), syndrome (0x8708c3)

Fixes: d7e75a325cb2 ("net/mlx5e: Add offloading of E-Switch TC pedit (header re-write) actions")
Signed-off-by: Roi Dayan <roid@nvidia.com>
Reviewed-by: Oz Shlomo <ozsh@nvidia.com>
Reviewed-by: Maor Dickman <maord@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Bridge, ensure dev_name is null-terminated
Vlad Buslov [Thu, 6 Jan 2022 16:45:26 +0000 (18:45 +0200)]
net/mlx5: Bridge, ensure dev_name is null-terminated

Even though net_device->name is guaranteed to be null-terminated string of
size<=IFNAMSIZ, the test robot complains that return value of netdev_name()
can be larger:

In file included from include/trace/define_trace.h:102,
                    from drivers/net/ethernet/mellanox/mlx5/core/esw/diag/bridge_tracepoint.h:113,
                    from drivers/net/ethernet/mellanox/mlx5/core/esw/bridge.c:12:
   drivers/net/ethernet/mellanox/mlx5/core/esw/diag/bridge_tracepoint.h: In function 'trace_event_raw_event_mlx5_esw_bridge_fdb_template':
>> drivers/net/ethernet/mellanox/mlx5/core/esw/diag/bridge_tracepoint.h:24:29: warning: 'strncpy' output may be truncated copying 16 bytes from a string of length 20 [-Wstringop-truncation]
      24 |                             strncpy(__entry->dev_name,
         |                             ^~~~~~~~~~~~~~~~~~~~~~~~~~
      25 |                                     netdev_name(fdb->dev),
         |                                     ~~~~~~~~~~~~~~~~~~~~~~
      26 |                                     IFNAMSIZ);
         |                                     ~~~~~~~~~

This is caused by the fact that default value of IFNAMSIZ is 16, while
placeholder value that is returned by netdev_name() for unnamed net devices
is larger than that.

The offending code is in a tracing function that is only called for mlx5
representors, so there is no straightforward way to reproduce the issue but
let's fix it for correctness sake by replacing strncpy() with strscpy() to
ensure that resulting string is always null-terminated.

Fixes: 9724fd5d9c2a ("net/mlx5: Bridge, add tracepoints")
Reported-by: kernel test robot <lkp@intel.com>
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agonet/mlx5: Bridge, take rtnl lock in init error handler
Vlad Buslov [Thu, 6 Jan 2022 14:40:18 +0000 (16:40 +0200)]
net/mlx5: Bridge, take rtnl lock in init error handler

The mlx5_esw_bridge_cleanup() is expected to be called with rtnl lock
taken, which is true for mlx5e_rep_bridge_cleanup() function but not for
error handling code in mlx5e_rep_bridge_init(). Add missing rtnl
lock/unlock calls and extend both mlx5_esw_bridge_cleanup() and its dual
function mlx5_esw_bridge_init() with ASSERT_RTNL() to verify the invariant
from now on.

Fixes: 7cd6a54a8285 ("net/mlx5: Bridge, handle FDB events")
Fixes: 19e9bfa044f3 ("net/mlx5: Bridge, add offload infrastructure")
Signed-off-by: Vlad Buslov <vladbu@nvidia.com>
Reviewed-by: Roi Dayan <roid@nvidia.com>
Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
2 years agoMerge branch '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue
Jakub Kicinski [Wed, 2 Feb 2022 04:39:46 +0000 (20:39 -0800)]
Merge branch '40GbE' of git://git./linux/kernel/git/tnguy/net-queue

Tony Nguyen says:

====================
Intel Wired LAN Driver Updates 2022-01-31

This series contains updates to i40e driver only.

Jedrzej fixes a condition check which would cause an error when
resetting bandwidth when DCB is active with one TC.

Karen resolves a null pointer dereference that could occur when removing
the driver while VSI rings are being disabled.

* '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/net-queue:
  i40e: Fix reset path while removing the driver
  i40e: Fix reset bw limit when DCB enabled with 1 TC
====================

Link: https://lore.kernel.org/r/20220201000522.505909-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: macsec: Verify that send_sci is on when setting Tx sci explicitly
Lior Nahmanson [Sun, 30 Jan 2022 11:37:52 +0000 (13:37 +0200)]
net: macsec: Verify that send_sci is on when setting Tx sci explicitly

When setting Tx sci explicit, the Rx side is expected to use this
sci and not recalculate it from the packet.However, in case of Tx sci
is explicit and send_sci is off, the receiver is wrongly recalculate
the sci from the source MAC address which most likely be different
than the explicit sci.

Fix by preventing such configuration when macsec newlink is established
and return EINVAL error code on such cases.

Fixes: c09440f7dcb3 ("macsec: introduce IEEE 802.1AE driver")
Signed-off-by: Lior Nahmanson <liorna@nvidia.com>
Reviewed-by: Raed Salem <raeds@nvidia.com>
Signed-off-by: Raed Salem <raeds@nvidia.com>
Link: https://lore.kernel.org/r/1643542672-29403-1-git-send-email-raeds@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoipheth: fix EOVERFLOW in ipheth_rcvbulk_callback
Georgi Valkov [Tue, 1 Feb 2022 07:16:18 +0000 (08:16 +0100)]
ipheth: fix EOVERFLOW in ipheth_rcvbulk_callback

When rx_buf is allocated we need to account for IPHETH_IP_ALIGN,
which reduces the usable size by 2 bytes. Otherwise we have 1512
bytes usable instead of 1514, and if we receive more than 1512
bytes, ipheth_rcvbulk_callback is called with status -EOVERFLOW,
after which the driver malfunctiones and all communication stops.

Resolves ipheth 2-1:4.2: ipheth_rcvbulk_callback: urb status: -75

Fixes: f33d9e2b48a3 ("usbnet: ipheth: fix connectivity with iOS 14")
Signed-off-by: Georgi Valkov <gvalkov@abv.bg>
Tested-by: Jan Kiszka <jan.kiszka@siemens.com>
Link: https://lore.kernel.org/all/B60B8A4B-92A0-49B3-805D-809A2433B46C@abv.bg/
Link: https://lore.kernel.org/all/24851bd2769434a5fc24730dce8e8a984c5a4505.1643699778.git.jan.kiszka@siemens.com/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agotcp: fix mem under-charging with zerocopy sendmsg()
Eric Dumazet [Tue, 1 Feb 2022 06:52:54 +0000 (22:52 -0800)]
tcp: fix mem under-charging with zerocopy sendmsg()

We got reports of following warning in inet_sock_destruct()

WARN_ON(sk_forward_alloc_get(sk));

Whenever we add a non zero-copy fragment to a pure zerocopy skb,
we have to anticipate that whole skb->truesize will be uncharged
when skb is finally freed.

skb->data_len is the payload length. But the memory truesize
estimated by __zerocopy_sg_from_iter() is page aligned.

Fixes: 9b65b17db723 ("net: avoid double accounting for pure zerocopy skbs")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Talal Ahmad <talalahmad@google.com>
Cc: Arjun Roy <arjunroy@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Link: https://lore.kernel.org/r/20220201065254.680532-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoaf_packet: fix data-race in packet_setsockopt / packet_setsockopt
Eric Dumazet [Tue, 1 Feb 2022 02:23:58 +0000 (18:23 -0800)]
af_packet: fix data-race in packet_setsockopt / packet_setsockopt

When packet_setsockopt( PACKET_FANOUT_DATA ) reads po->fanout,
no lock is held, meaning that another thread can change po->fanout.

Given that po->fanout can only be set once during the socket lifetime
(it is only cleared from fanout_release()), we can use
READ_ONCE()/WRITE_ONCE() to document the race.

BUG: KCSAN: data-race in packet_setsockopt / packet_setsockopt

write to 0xffff88813ae8e300 of 8 bytes by task 14653 on cpu 0:
 fanout_add net/packet/af_packet.c:1791 [inline]
 packet_setsockopt+0x22fe/0x24a0 net/packet/af_packet.c:3931
 __sys_setsockopt+0x209/0x2a0 net/socket.c:2180
 __do_sys_setsockopt net/socket.c:2191 [inline]
 __se_sys_setsockopt net/socket.c:2188 [inline]
 __x64_sys_setsockopt+0x62/0x70 net/socket.c:2188
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

read to 0xffff88813ae8e300 of 8 bytes by task 14654 on cpu 1:
 packet_setsockopt+0x691/0x24a0 net/packet/af_packet.c:3935
 __sys_setsockopt+0x209/0x2a0 net/socket.c:2180
 __do_sys_setsockopt net/socket.c:2191 [inline]
 __se_sys_setsockopt net/socket.c:2188 [inline]
 __x64_sys_setsockopt+0x62/0x70 net/socket.c:2188
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

value changed: 0x0000000000000000 -> 0xffff888106f8c000

Reported by Kernel Concurrency Sanitizer on:
CPU: 1 PID: 14654 Comm: syz-executor.3 Not tainted 5.16.0-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011

Fixes: 47dceb8ecdc1 ("packet: add classic BPF fanout mode")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Willem de Bruijn <willemb@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Link: https://lore.kernel.org/r/20220201022358.330621-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agortnetlink: make sure to refresh master_dev/m_ops in __rtnl_newlink()
Eric Dumazet [Tue, 1 Feb 2022 01:21:06 +0000 (17:21 -0800)]
rtnetlink: make sure to refresh master_dev/m_ops in __rtnl_newlink()

While looking at one unrelated syzbot bug, I found the replay logic
in __rtnl_newlink() to potentially trigger use-after-free.

It is better to clear master_dev and m_ops inside the loop,
in case we have to replay it.

Fixes: ba7d49b1f0f8 ("rtnetlink: provide api for getting and setting slave info")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Jiri Pirko <jiri@nvidia.com>
Link: https://lore.kernel.org/r/20220201012106.216495-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: sched: fix use-after-free in tc_new_tfilter()
Eric Dumazet [Mon, 31 Jan 2022 17:20:18 +0000 (09:20 -0800)]
net: sched: fix use-after-free in tc_new_tfilter()

Whenever tc_new_tfilter() jumps back to replay: label,
we need to make sure @q and @chain local variables are cleared again,
or risk use-after-free as in [1]

For consistency, apply the same fix in tc_ctl_chain()

BUG: KASAN: use-after-free in mini_qdisc_pair_swap+0x1b9/0x1f0 net/sched/sch_generic.c:1581
Write of size 8 at addr ffff8880985c4b08 by task syz-executor.4/1945

CPU: 0 PID: 1945 Comm: syz-executor.4 Not tainted 5.17.0-rc1-syzkaller-00495-gff58831fa02d #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011
Call Trace:
 <TASK>
 __dump_stack lib/dump_stack.c:88 [inline]
 dump_stack_lvl+0xcd/0x134 lib/dump_stack.c:106
 print_address_description.constprop.0.cold+0x8d/0x336 mm/kasan/report.c:255
 __kasan_report mm/kasan/report.c:442 [inline]
 kasan_report.cold+0x83/0xdf mm/kasan/report.c:459
 mini_qdisc_pair_swap+0x1b9/0x1f0 net/sched/sch_generic.c:1581
 tcf_chain_head_change_item net/sched/cls_api.c:372 [inline]
 tcf_chain0_head_change.isra.0+0xb9/0x120 net/sched/cls_api.c:386
 tcf_chain_tp_insert net/sched/cls_api.c:1657 [inline]
 tcf_chain_tp_insert_unique net/sched/cls_api.c:1707 [inline]
 tc_new_tfilter+0x1e67/0x2350 net/sched/cls_api.c:2086
 rtnetlink_rcv_msg+0x80d/0xb80 net/core/rtnetlink.c:5583
 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494
 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline]
 netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343
 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919
 sock_sendmsg_nosec net/socket.c:705 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:725
 ____sys_sendmsg+0x331/0x810 net/socket.c:2413
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467
 __sys_sendmmsg+0x195/0x470 net/socket.c:2553
 __do_sys_sendmmsg net/socket.c:2582 [inline]
 __se_sys_sendmmsg net/socket.c:2579 [inline]
 __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae
RIP: 0033:0x7f2647172059
Code: ff ff c3 66 2e 0f 1f 84 00 00 00 00 00 0f 1f 40 00 48 89 f8 48 89 f7 48 89 d6 48 89 ca 4d 89 c2 4d 89 c8 4c 8b 4c 24 08 0f 05 <48> 3d 01 f0 ff ff 73 01 c3 48 c7 c1 b8 ff ff ff f7 d8 64 89 01 48
RSP: 002b:00007f2645aa5168 EFLAGS: 00000246 ORIG_RAX: 0000000000000133
RAX: ffffffffffffffda RBX: 00007f2647285100 RCX: 00007f2647172059
RDX: 040000000000009f RSI: 00000000200002c0 RDI: 0000000000000006
RBP: 00007f26471cc08d R08: 0000000000000000 R09: 0000000000000000
R10: 9e00000000000000 R11: 0000000000000246 R12: 0000000000000000
R13: 00007fffb3f7f02f R14: 00007f2645aa5300 R15: 0000000000022000
 </TASK>

Allocated by task 1944:
 kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
 kasan_set_track mm/kasan/common.c:45 [inline]
 set_alloc_info mm/kasan/common.c:436 [inline]
 ____kasan_kmalloc mm/kasan/common.c:515 [inline]
 ____kasan_kmalloc mm/kasan/common.c:474 [inline]
 __kasan_kmalloc+0xa9/0xd0 mm/kasan/common.c:524
 kmalloc_node include/linux/slab.h:604 [inline]
 kzalloc_node include/linux/slab.h:726 [inline]
 qdisc_alloc+0xac/0xa10 net/sched/sch_generic.c:941
 qdisc_create.constprop.0+0xce/0x10f0 net/sched/sch_api.c:1211
 tc_modify_qdisc+0x4c5/0x1980 net/sched/sch_api.c:1660
 rtnetlink_rcv_msg+0x413/0xb80 net/core/rtnetlink.c:5592
 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494
 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline]
 netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343
 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919
 sock_sendmsg_nosec net/socket.c:705 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:725
 ____sys_sendmsg+0x331/0x810 net/socket.c:2413
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467
 __sys_sendmmsg+0x195/0x470 net/socket.c:2553
 __do_sys_sendmmsg net/socket.c:2582 [inline]
 __se_sys_sendmmsg net/socket.c:2579 [inline]
 __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Freed by task 3609:
 kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
 kasan_set_track+0x21/0x30 mm/kasan/common.c:45
 kasan_set_free_info+0x20/0x30 mm/kasan/generic.c:370
 ____kasan_slab_free mm/kasan/common.c:366 [inline]
 ____kasan_slab_free+0x130/0x160 mm/kasan/common.c:328
 kasan_slab_free include/linux/kasan.h:236 [inline]
 slab_free_hook mm/slub.c:1728 [inline]
 slab_free_freelist_hook+0x8b/0x1c0 mm/slub.c:1754
 slab_free mm/slub.c:3509 [inline]
 kfree+0xcb/0x280 mm/slub.c:4562
 rcu_do_batch kernel/rcu/tree.c:2527 [inline]
 rcu_core+0x7b8/0x1540 kernel/rcu/tree.c:2778
 __do_softirq+0x29b/0x9c2 kernel/softirq.c:558

Last potentially related work creation:
 kasan_save_stack+0x1e/0x40 mm/kasan/common.c:38
 __kasan_record_aux_stack+0xbe/0xd0 mm/kasan/generic.c:348
 __call_rcu kernel/rcu/tree.c:3026 [inline]
 call_rcu+0xb1/0x740 kernel/rcu/tree.c:3106
 qdisc_put_unlocked+0x6f/0x90 net/sched/sch_generic.c:1109
 tcf_block_release+0x86/0x90 net/sched/cls_api.c:1238
 tc_new_tfilter+0xc0d/0x2350 net/sched/cls_api.c:2148
 rtnetlink_rcv_msg+0x80d/0xb80 net/core/rtnetlink.c:5583
 netlink_rcv_skb+0x153/0x420 net/netlink/af_netlink.c:2494
 netlink_unicast_kernel net/netlink/af_netlink.c:1317 [inline]
 netlink_unicast+0x539/0x7e0 net/netlink/af_netlink.c:1343
 netlink_sendmsg+0x904/0xe00 net/netlink/af_netlink.c:1919
 sock_sendmsg_nosec net/socket.c:705 [inline]
 sock_sendmsg+0xcf/0x120 net/socket.c:725
 ____sys_sendmsg+0x331/0x810 net/socket.c:2413
 ___sys_sendmsg+0xf3/0x170 net/socket.c:2467
 __sys_sendmmsg+0x195/0x470 net/socket.c:2553
 __do_sys_sendmmsg net/socket.c:2582 [inline]
 __se_sys_sendmmsg net/socket.c:2579 [inline]
 __x64_sys_sendmmsg+0x99/0x100 net/socket.c:2579
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

The buggy address belongs to the object at ffff8880985c4800
 which belongs to the cache kmalloc-1k of size 1024
The buggy address is located 776 bytes inside of
 1024-byte region [ffff8880985c4800ffff8880985c4c00)
The buggy address belongs to the page:
page:ffffea0002617000 refcount:1 mapcount:0 mapping:0000000000000000 index:0x0 pfn:0x985c0
head:ffffea0002617000 order:3 compound_mapcount:0 compound_pincount:0
flags: 0xfff00000010200(slab|head|node=0|zone=1|lastcpupid=0x7ff)
raw: 00fff00000010200 0000000000000000 dead000000000122 ffff888010c41dc0
raw: 0000000000000000 0000000000100010 00000001ffffffff 0000000000000000
page dumped because: kasan: bad access detected
page_owner tracks the page as allocated
page last allocated via order 3, migratetype Unmovable, gfp_mask 0x1d20c0(__GFP_IO|__GFP_FS|__GFP_NOWARN|__GFP_NORETRY|__GFP_COMP|__GFP_NOMEMALLOC|__GFP_HARDWALL), pid 1941, ts 1038999441284, free_ts 1033444432829
 prep_new_page mm/page_alloc.c:2434 [inline]
 get_page_from_freelist+0xa72/0x2f50 mm/page_alloc.c:4165
 __alloc_pages+0x1b2/0x500 mm/page_alloc.c:5389
 alloc_pages+0x1aa/0x310 mm/mempolicy.c:2271
 alloc_slab_page mm/slub.c:1799 [inline]
 allocate_slab mm/slub.c:1944 [inline]
 new_slab+0x28a/0x3b0 mm/slub.c:2004
 ___slab_alloc+0x87c/0xe90 mm/slub.c:3018
 __slab_alloc.constprop.0+0x4d/0xa0 mm/slub.c:3105
 slab_alloc_node mm/slub.c:3196 [inline]
 slab_alloc mm/slub.c:3238 [inline]
 __kmalloc+0x2fb/0x340 mm/slub.c:4420
 kmalloc include/linux/slab.h:586 [inline]
 kzalloc include/linux/slab.h:715 [inline]
 __register_sysctl_table+0x112/0x1090 fs/proc/proc_sysctl.c:1335
 neigh_sysctl_register+0x2c8/0x5e0 net/core/neighbour.c:3787
 devinet_sysctl_register+0xb1/0x230 net/ipv4/devinet.c:2618
 inetdev_init+0x286/0x580 net/ipv4/devinet.c:278
 inetdev_event+0xa8a/0x15d0 net/ipv4/devinet.c:1532
 notifier_call_chain+0xb5/0x200 kernel/notifier.c:84
 call_netdevice_notifiers_info+0xb5/0x130 net/core/dev.c:1919
 call_netdevice_notifiers_extack net/core/dev.c:1931 [inline]
 call_netdevice_notifiers net/core/dev.c:1945 [inline]
 register_netdevice+0x1073/0x1500 net/core/dev.c:9698
 veth_newlink+0x59c/0xa90 drivers/net/veth.c:1722
page last free stack trace:
 reset_page_owner include/linux/page_owner.h:24 [inline]
 free_pages_prepare mm/page_alloc.c:1352 [inline]
 free_pcp_prepare+0x374/0x870 mm/page_alloc.c:1404
 free_unref_page_prepare mm/page_alloc.c:3325 [inline]
 free_unref_page+0x19/0x690 mm/page_alloc.c:3404
 release_pages+0x748/0x1220 mm/swap.c:956
 tlb_batch_pages_flush mm/mmu_gather.c:50 [inline]
 tlb_flush_mmu_free mm/mmu_gather.c:243 [inline]
 tlb_flush_mmu+0xe9/0x6b0 mm/mmu_gather.c:250
 zap_pte_range mm/memory.c:1441 [inline]
 zap_pmd_range mm/memory.c:1490 [inline]
 zap_pud_range mm/memory.c:1519 [inline]
 zap_p4d_range mm/memory.c:1540 [inline]
 unmap_page_range+0x1d1d/0x2a30 mm/memory.c:1561
 unmap_single_vma+0x198/0x310 mm/memory.c:1606
 unmap_vmas+0x16b/0x2f0 mm/memory.c:1638
 exit_mmap+0x201/0x670 mm/mmap.c:3178
 __mmput+0x122/0x4b0 kernel/fork.c:1114
 mmput+0x56/0x60 kernel/fork.c:1135
 exit_mm kernel/exit.c:507 [inline]
 do_exit+0xa3c/0x2a30 kernel/exit.c:793
 do_group_exit+0xd2/0x2f0 kernel/exit.c:935
 __do_sys_exit_group kernel/exit.c:946 [inline]
 __se_sys_exit_group kernel/exit.c:944 [inline]
 __x64_sys_exit_group+0x3a/0x50 kernel/exit.c:944
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x35/0xb0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

Memory state around the buggy address:
 ffff8880985c4a00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
 ffff8880985c4a80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
>ffff8880985c4b00: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
                      ^
 ffff8880985c4b80: fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb fb
 ffff8880985c4c00: fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc fc

Fixes: 470502de5bdb ("net: sched: unlock rules update API")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Cc: Vlad Buslov <vladbu@mellanox.com>
Cc: Jiri Pirko <jiri@mellanox.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Link: https://lore.kernel.org/r/20220131172018.3704490-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoethernet: smc911x: fix indentation in get/set EEPROM
Jakub Kicinski [Mon, 31 Jan 2022 21:17:30 +0000 (13:17 -0800)]
ethernet: smc911x: fix indentation in get/set EEPROM

Build bot produced a smatch indentation warning,
the code looks correct but it mixes spaces and tabs.

Reported-by: kernel test robot <lkp@intel.com>
Link: https://lore.kernel.org/r/20220131211730.3940875-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoe1000e: Handshake with CSME starts from ADL platforms
Sasha Neftin [Tue, 7 Dec 2021 11:23:42 +0000 (13:23 +0200)]
e1000e: Handshake with CSME starts from ADL platforms

Handshake with CSME/AMT on none provisioned platforms during S0ix flow
is not supported on TGL platform and can cause to HW unit hang. Update
the handshake with CSME flow to start from the ADL platform.

Fixes: 3e55d231716e ("e1000e: Add handshake with the CSME to support S0ix")
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Nechama Kraus <nechamax.kraus@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
2 years agoe1000e: Separate ADP board type from TGP
Sasha Neftin [Tue, 7 Dec 2021 11:23:06 +0000 (13:23 +0200)]
e1000e: Separate ADP board type from TGP

We have the same LAN controller on different PCH's. Separate ADP board
type from a TGP which will allow for specific fixes to be applied for
ADP platforms.

Suggested-by: Kai-Heng Feng <kai.heng.feng@canonical.com>
Suggested-by: Dima Ruinskiy <dima.ruinskiy@intel.com>
Signed-off-by: Sasha Neftin <sasha.neftin@intel.com>
Tested-by: Nechama Kraus <nechamax.kraus@linux.intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
2 years agonet: phy: Fix qca8081 with speeds lower than 2.5Gb/s
Jonathan McDowell [Mon, 31 Jan 2022 13:56:41 +0000 (13:56 +0000)]
net: phy: Fix qca8081 with speeds lower than 2.5Gb/s

A typo in qca808x_read_status means we try to set SMII mode on the port
rather than SGMII when the link speed is not 2.5Gb/s. This results in no
traffic due to the mismatch in configuration between the phy and the
mac.

v2:
 Only change interface mode when the link is up

Fixes: 79c7bc0521545 ("net: phy: add qca8081 read_status")
Cc: stable@vger.kernel.org
Signed-off-by: Jonathan McDowell <noodles@earth.li>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: macsec: Fix offload support for NETDEV_UNREGISTER event
Lior Nahmanson [Sun, 30 Jan 2022 11:29:01 +0000 (13:29 +0200)]
net: macsec: Fix offload support for NETDEV_UNREGISTER event

Current macsec netdev notify handler handles NETDEV_UNREGISTER event by
releasing relevant SW resources only, this causes resources leak in case
of macsec HW offload, as the underlay driver was not notified to clean
it's macsec offload resources.

Fix by calling the underlay driver to clean it's relevant resources
by moving offload handling from macsec_dellink() to macsec_common_dellink()
when handling NETDEV_UNREGISTER event.

Fixes: 3cf3227a21d1 ("net: macsec: hardware offloading infrastructure")
Signed-off-by: Lior Nahmanson <liorna@nvidia.com>
Reviewed-by: Raed Salem <raeds@nvidia.com>
Signed-off-by: Raed Salem <raeds@nvidia.com>
Reviewed-by: Antoine Tenart <atenart@kernel.org>
Link: https://lore.kernel.org/r/1643542141-28956-1-git-send-email-raeds@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: stmmac: dump gmac4 DMA registers correctly
Camel Guo [Mon, 31 Jan 2022 08:38:40 +0000 (09:38 +0100)]
net: stmmac: dump gmac4 DMA registers correctly

Unlike gmac100, gmac1000, gmac4 has 27 DMA registers and they are
located at DMA_CHAN_BASE_ADDR (0x1100). In order for ethtool to dump
gmac4 DMA registers correctly, this commit checks if a net_device has
gmac4 and uses different logic to dump its DMA registers.

This fixes the following KASAN warning, which can normally be triggered
by a command similar like "ethtool -d eth0":

BUG: KASAN: vmalloc-out-of-bounds in dwmac4_dump_dma_regs+0x6d4/0xb30
Write of size 4 at addr ffffffc010177100 by task ethtool/1839
 kasan_report+0x200/0x21c
 __asan_report_store4_noabort+0x34/0x60
 dwmac4_dump_dma_regs+0x6d4/0xb30
 stmmac_ethtool_gregs+0x110/0x204
 ethtool_get_regs+0x200/0x4b0
 dev_ethtool+0x1dac/0x3800
 dev_ioctl+0x7c0/0xb50
 sock_ioctl+0x298/0x6c4
 ...

Fixes: fbf68229ffe7 ("net: stmmac: unify registers dumps methods")
Signed-off-by: Camel Guo <camelg@axis.com>
Link: https://lore.kernel.org/r/20220131083841.3346801-1-camel.guo@axis.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: dsa: mt7530: make NET_DSA_MT7530 select MEDIATEK_GE_PHY
Arınç ÜNAL [Sat, 29 Jan 2022 06:27:04 +0000 (09:27 +0300)]
net: dsa: mt7530: make NET_DSA_MT7530 select MEDIATEK_GE_PHY

Make MediaTek MT753x DSA driver enable MediaTek Gigabit PHYs driver to
properly control MT7530 and MT7531 switch PHYs.

A noticeable change is that the behaviour of switchport interfaces going
up-down-up-down is no longer there.

Fixes: b8f126a8d543 ("net-next: dsa: add dsa support for Mediatek MT7530 switch")
Signed-off-by: Arınç ÜNAL <arinc.unal@arinc9.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Reviewed-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/r/20220129062703.595-1-arinc.unal@arinc9.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoi40e: Fix reset path while removing the driver
Karen Sornek [Wed, 12 Jan 2022 09:19:47 +0000 (10:19 +0100)]
i40e: Fix reset path while removing the driver

Fix the crash in kernel while dereferencing the NULL pointer,
when the driver is unloaded and simultaneously the VSI rings
are being stopped.

The hardware requires 50msec in order to finish RX queues
disable. For this purpose the driver spins in mdelay function
for the operation to be completed.

For example changing number of queues which requires reset would
fail in the following call stack:

1) i40e_prep_for_reset
2) i40e_pf_quiesce_all_vsi
3) i40e_quiesce_vsi
4) i40e_vsi_close
5) i40e_down
6) i40e_vsi_stop_rings
7) i40e_vsi_control_rx -> disable requires the delay of 50msecs
8) continue back in i40e_down function where
   i40e_clean_tx_ring(vsi->tx_rings[i]) is going to crash

When the driver was spinning vsi_release called
i40e_vsi_free_arrays where the vsi->tx_rings resources
were freed and the pointer was set to NULL.

Fixes: 5b6d4a7f20b0 ("i40e: Fix crash during removing i40e driver")
Signed-off-by: Slawomir Laba <slawomirx.laba@intel.com>
Signed-off-by: Sylwester Dziedziuch <sylwesterx.dziedziuch@intel.com>
Signed-off-by: Karen Sornek <karen.sornek@intel.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
2 years agoi40e: Fix reset bw limit when DCB enabled with 1 TC
Jedrzej Jagielski [Tue, 14 Dec 2021 10:08:22 +0000 (10:08 +0000)]
i40e: Fix reset bw limit when DCB enabled with 1 TC

There was an AQ error I40E_AQ_RC_EINVAL when trying
to reset bw limit as part of bw allocation setup.
This was caused by trying to reset bw limit with
DCB enabled. Bw limit should not be reset when
DCB is enabled. The code was relying on the pf->flags
to check if DCB is enabled but if only 1 TC is available
this flag will not be set even though DCB is enabled.
Add a check for number of TC and if it is 1
don't try to reset bw limit even if pf->flags shows
DCB as disabled.

Fixes: fa38e30ac73f ("i40e: Fix for Tx timeouts when interface is brought up if DCB is enabled")
Suggested-by: Alexander Lobakin <alexandr.lobakin@intel.com> # Flatten the condition
Signed-off-by: Sylwester Dziedziuch <sylwesterx.dziedziuch@intel.com>
Signed-off-by: Jedrzej Jagielski <jedrzej.jagielski@intel.com>
Reviewed-by: Alexander Lobakin <alexandr.lobakin@intel.com>
Tested-by: Imam Hassan Reza Biswas <imam.hassan.reza.biswas@intel.com>
Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
2 years agonet/smc: Forward wakeup to smc socket waitqueue after fallback
Wen Gu [Wed, 26 Jan 2022 15:33:04 +0000 (23:33 +0800)]
net/smc: Forward wakeup to smc socket waitqueue after fallback

When we replace TCP with SMC and a fallback occurs, there may be
some socket waitqueue entries remaining in smc socket->wq, such
as eppoll_entries inserted by userspace applications.

After the fallback, data flows over TCP/IP and only clcsocket->wq
will be woken up. Applications can't be notified by the entries
which were inserted in smc socket->wq before fallback. So we need
a mechanism to wake up smc socket->wq at the same time if some
entries remaining in it.

The current workaround is to transfer the entries from smc socket->wq
to clcsock->wq during the fallback. But this may cause a crash
like this:

 general protection fault, probably for non-canonical address 0xdead000000000100: 0000 [#1] PREEMPT SMP PTI
 CPU: 3 PID: 0 Comm: swapper/3 Kdump: loaded Tainted: G E     5.16.0+ #107
 RIP: 0010:__wake_up_common+0x65/0x170
 Call Trace:
  <IRQ>
  __wake_up_common_lock+0x7a/0xc0
  sock_def_readable+0x3c/0x70
  tcp_data_queue+0x4a7/0xc40
  tcp_rcv_established+0x32f/0x660
  ? sk_filter_trim_cap+0xcb/0x2e0
  tcp_v4_do_rcv+0x10b/0x260
  tcp_v4_rcv+0xd2a/0xde0
  ip_protocol_deliver_rcu+0x3b/0x1d0
  ip_local_deliver_finish+0x54/0x60
  ip_local_deliver+0x6a/0x110
  ? tcp_v4_early_demux+0xa2/0x140
  ? tcp_v4_early_demux+0x10d/0x140
  ip_sublist_rcv_finish+0x49/0x60
  ip_sublist_rcv+0x19d/0x230
  ip_list_rcv+0x13e/0x170
  __netif_receive_skb_list_core+0x1c2/0x240
  netif_receive_skb_list_internal+0x1e6/0x320
  napi_complete_done+0x11d/0x190
  mlx5e_napi_poll+0x163/0x6b0 [mlx5_core]
  __napi_poll+0x3c/0x1b0
  net_rx_action+0x27c/0x300
  __do_softirq+0x114/0x2d2
  irq_exit_rcu+0xb4/0xe0
  common_interrupt+0xba/0xe0
  </IRQ>
  <TASK>

The crash is caused by privately transferring waitqueue entries from
smc socket->wq to clcsock->wq. The owners of these entries, such as
epoll, have no idea that the entries have been transferred to a
different socket wait queue and still use original waitqueue spinlock
(smc socket->wq.wait.lock) to make the entries operation exclusive,
but it doesn't work. The operations to the entries, such as removing
from the waitqueue (now is clcsock->wq after fallback), may cause a
crash when clcsock waitqueue is being iterated over at the moment.

This patch tries to fix this by no longer transferring wait queue
entries privately, but introducing own implementations of clcsock's
callback functions in fallback situation. The callback functions will
forward the wakeup to smc socket->wq if clcsock->wq is actually woken
up and smc socket->wq has remaining entries.

Fixes: 2153bd1e3d3d ("net/smc: Transfer remaining wait queue entries during fallback")
Suggested-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: Wen Gu <guwen@linux.alibaba.com>
Acked-by: Karsten Graul <kgraul@linux.ibm.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: stmmac: properly handle with runtime pm in stmmac_dvr_remove()
Jisheng Zhang [Fri, 28 Jan 2022 14:15:50 +0000 (22:15 +0800)]
net: stmmac: properly handle with runtime pm in stmmac_dvr_remove()

There are two issues with runtime pm handling in stmmac_dvr_remove():

1. the mac is runtime suspended before stopping dma and rx/tx. We
need to ensure the device is properly resumed back.

2. the stmmaceth clk enable/disable isn't balanced in both exit and
error handling code path. Take the exit code path for example, when we
unbind the driver or rmmod the driver module, the mac is runtime
suspended as said above, so the stmmaceth clk is disabled, but
stmmac_dvr_remove()
  stmmac_remove_config_dt()
    clk_disable_unprepare()
CCF will complain this time. The error handling code path suffers
from the similar situtaion.

Here are kernel warnings in error handling code path on Allwinner D1
platform:

[    1.604695] ------------[ cut here ]------------
[    1.609328] bus-emac already disabled
[    1.613015] WARNING: CPU: 0 PID: 38 at drivers/clk/clk.c:952 clk_core_disable+0xcc/0xec
[    1.621039] CPU: 0 PID: 38 Comm: kworker/u2:1 Not tainted 5.14.0-rc4#1
[    1.627653] Hardware name: Allwinner D1 NeZha (DT)
[    1.632443] Workqueue: events_unbound deferred_probe_work_func
[    1.638286] epc : clk_core_disable+0xcc/0xec
[    1.642561]  ra : clk_core_disable+0xcc/0xec
[    1.646835] epc : ffffffff8023c2ec ra : ffffffff8023c2ec sp : ffffffd00411bb10
[    1.654054]  gp : ffffffff80ec9988 tp : ffffffe00143a800 t0 : ffffffff80ed6a6f
[    1.661272]  t1 : ffffffff80ed6a60 t2 : 0000000000000000 s0 : ffffffe001509e00
[    1.668489]  s1 : 0000000000000001 a0 : 0000000000000019 a1 : ffffffff80e80bd8
[    1.675707]  a2 : 00000000ffffefff a3 : 00000000000000f4 a4 : 0000000000000002
[    1.682924]  a5 : 0000000000000001 a6 : 0000000000000030 a7 : 00000000028f5c29
[    1.690141]  s2 : 0000000000000800 s3 : ffffffe001375000 s4 : ffffffe01fdf7a80
[    1.697358]  s5 : ffffffe001375010 s6 : ffffffff8001fc10 s7 : ffffffffffffffff
[    1.704577]  s8 : 0000000000000001 s9 : ffffffff80ecb248 s10: ffffffe001b80000
[    1.711794]  s11: ffffffe001b80760 t3 : 0000000000000062 t4 : ffffffffffffffff
[    1.719012]  t5 : ffffffff80e0f6d8 t6 : ffffffd00411b8f0
[    1.724321] status: 8000000201800100 badaddr: 0000000000000000 cause: 0000000000000003
[    1.732233] [<ffffffff8023c2ec>] clk_core_disable+0xcc/0xec
[    1.737810] [<ffffffff80240430>] clk_disable+0x38/0x78
[    1.742956] [<ffffffff8001fc0c>] worker_thread+0x1a8/0x4d8
[    1.748451] [<ffffffff8031a500>] stmmac_remove_config_dt+0x1c/0x4c
[    1.754646] [<ffffffff8031c8ec>] sun8i_dwmac_probe+0x378/0x82c
[    1.760484] [<ffffffff8001fc0c>] worker_thread+0x1a8/0x4d8
[    1.765975] [<ffffffff8029a6c8>] platform_probe+0x64/0xf0
[    1.771382] [<ffffffff8029833c>] really_probe.part.0+0x8c/0x30c
[    1.777305] [<ffffffff8029865c>] __driver_probe_device+0xa0/0x148
[    1.783402] [<ffffffff8029873c>] driver_probe_device+0x38/0x138
[    1.789324] [<ffffffff802989cc>] __device_attach_driver+0xd0/0x170
[    1.795508] [<ffffffff802988f8>] __driver_attach_async_helper+0xbc/0xc0
[    1.802125] [<ffffffff802965ac>] bus_for_each_drv+0x68/0xb4
[    1.807701] [<ffffffff80298d1c>] __device_attach+0xd8/0x184
[    1.813277] [<ffffffff802967b0>] bus_probe_device+0x98/0xbc
[    1.818852] [<ffffffff80297904>] deferred_probe_work_func+0x90/0xd4
[    1.825122] [<ffffffff8001f8b8>] process_one_work+0x1e4/0x390
[    1.830872] [<ffffffff8001fd80>] worker_thread+0x31c/0x4d8
[    1.836362] [<ffffffff80026bf4>] kthreadd+0x94/0x188
[    1.841335] [<ffffffff80026bf4>] kthreadd+0x94/0x188
[    1.846304] [<ffffffff8001fa60>] process_one_work+0x38c/0x390
[    1.852054] [<ffffffff80026564>] kthread+0x124/0x160
[    1.857021] [<ffffffff8002643c>] set_kthread_struct+0x5c/0x60
[    1.862770] [<ffffffff80001f08>] ret_from_syscall_rejected+0x8/0xc
[    1.868956] ---[ end trace 8d5c6046255f84a0 ]---
[    1.873675] ------------[ cut here ]------------
[    1.878366] bus-emac already unprepared
[    1.882378] WARNING: CPU: 0 PID: 38 at drivers/clk/clk.c:810 clk_core_unprepare+0xe4/0x168
[    1.890673] CPU: 0 PID: 38 Comm: kworker/u2:1 Tainted: G        W 5.14.0-rc4 #1
[    1.898674] Hardware name: Allwinner D1 NeZha (DT)
[    1.903464] Workqueue: events_unbound deferred_probe_work_func
[    1.909305] epc : clk_core_unprepare+0xe4/0x168
[    1.913840]  ra : clk_core_unprepare+0xe4/0x168
[    1.918375] epc : ffffffff8023d6cc ra : ffffffff8023d6cc sp : ffffffd00411bb10
[    1.925593]  gp : ffffffff80ec9988 tp : ffffffe00143a800 t0 : 0000000000000002
[    1.932811]  t1 : ffffffe01f743be0 t2 : 0000000000000040 s0 : ffffffe001509e00
[    1.940029]  s1 : 0000000000000001 a0 : 000000000000001b a1 : ffffffe00143a800
[    1.947246]  a2 : 0000000000000000 a3 : 00000000000000f4 a4 : 0000000000000001
[    1.954463]  a5 : 0000000000000000 a6 : 0000000005fce2a5 a7 : 0000000000000001
[    1.961680]  s2 : 0000000000000800 s3 : ffffffff80afeb90 s4 : ffffffe01fdf7a80
[    1.968898]  s5 : ffffffe001375010 s6 : ffffffff8001fc10 s7 : ffffffffffffffff
[    1.976115]  s8 : 0000000000000001 s9 : ffffffff80ecb248 s10: ffffffe001b80000
[    1.983333]  s11: ffffffe001b80760 t3 : ffffffff80b39120 t4 : 0000000000000001
[    1.990550]  t5 : 0000000000000000 t6 : ffffffe001600002
[    1.995859] status: 8000000201800120 badaddr: 0000000000000000 cause: 0000000000000003
[    2.003771] [<ffffffff8023d6cc>] clk_core_unprepare+0xe4/0x168
[    2.009609] [<ffffffff802403a0>] clk_unprepare+0x24/0x3c
[    2.014929] [<ffffffff8031a508>] stmmac_remove_config_dt+0x24/0x4c
[    2.021125] [<ffffffff8031c8ec>] sun8i_dwmac_probe+0x378/0x82c
[    2.026965] [<ffffffff8001fc0c>] worker_thread+0x1a8/0x4d8
[    2.032463] [<ffffffff8029a6c8>] platform_probe+0x64/0xf0
[    2.037871] [<ffffffff8029833c>] really_probe.part.0+0x8c/0x30c
[    2.043795] [<ffffffff8029865c>] __driver_probe_device+0xa0/0x148
[    2.049892] [<ffffffff8029873c>] driver_probe_device+0x38/0x138
[    2.055815] [<ffffffff802989cc>] __device_attach_driver+0xd0/0x170
[    2.061999] [<ffffffff802988f8>] __driver_attach_async_helper+0xbc/0xc0
[    2.068616] [<ffffffff802965ac>] bus_for_each_drv+0x68/0xb4
[    2.074193] [<ffffffff80298d1c>] __device_attach+0xd8/0x184
[    2.079769] [<ffffffff802967b0>] bus_probe_device+0x98/0xbc
[    2.085345] [<ffffffff80297904>] deferred_probe_work_func+0x90/0xd4
[    2.091616] [<ffffffff8001f8b8>] process_one_work+0x1e4/0x390
[    2.097367] [<ffffffff8001fd80>] worker_thread+0x31c/0x4d8
[    2.102858] [<ffffffff80026bf4>] kthreadd+0x94/0x188
[    2.107830] [<ffffffff80026bf4>] kthreadd+0x94/0x188
[    2.112800] [<ffffffff8001fa60>] process_one_work+0x38c/0x390
[    2.118551] [<ffffffff80026564>] kthread+0x124/0x160
[    2.123520] [<ffffffff8002643c>] set_kthread_struct+0x5c/0x60
[    2.129268] [<ffffffff80001f08>] ret_from_syscall_rejected+0x8/0xc
[    2.135455] ---[ end trace 8d5c6046255f84a1 ]---

Fixes: 5ec55823438e ("net: stmmac: add clocks management for gmac driver")
Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge tag 'ieee802154-for-net-2022-01-28' of git://git.kernel.org/pub/scm/linux/kerne...
David S. Miller [Fri, 28 Jan 2022 15:10:45 +0000 (15:10 +0000)]
Merge tag 'ieee802154-for-net-2022-01-28' of git://git./linux/kernel/git/sschmidt/wpan

Stefan Schmidt says:

====================
pull-request: ieee802154 for net 2022-01-28

An update from ieee802154 for your *net* tree.

A bunch of fixes in drivers, all from Miquel Raynal.
Clarifying the default channel in hwsim, leak fixes in at86rf230 and ca8210 as
well as a symbol duration fix for mcr20a. Topping up the driver fixes with
better error codes in nl802154 and a cleanup in MAINTAINERS for an orphaned
driver.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agogve: fix the wrong AdminQ buffer queue index check
Haiyue Wang [Fri, 28 Jan 2022 10:47:14 +0000 (18:47 +0800)]
gve: fix the wrong AdminQ buffer queue index check

The 'tail' and 'head' are 'unsigned int' type free-running count, when
'head' is overflow, the 'int i (= tail) < u32 head' will be false:

Only '- loop 0: idx = 63' result is shown, so it needs to use 'int' type
to compare, it can handle the overflow correctly.

typedef uint32_t u32;

int main()
{
        u32 tail, head;
        int stail, shead;
        int i, loop;

        tail = 0xffffffff;
        head = 0x00000000;

        for (i = tail, loop = 0; i < head; i++) {
                unsigned int idx = i & 63;

                printf("+ loop %d: idx = %u\n", loop++, idx);
        }

        stail = tail;
        shead = head;
        for (i = stail, loop = 0; i < shead; i++) {
                unsigned int idx = i & 63;

                printf("- loop %d: idx = %u\n", loop++, idx);
        }

        return 0;
}

Fixes: 5cdad90de62c ("gve: Batch AQ commands for creating and destroying queues.")
Signed-off-by: Haiyue Wang <haiyue.wang@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch 'ax25-fixes'
David S. Miller [Fri, 28 Jan 2022 14:56:47 +0000 (14:56 +0000)]
Merge branch 'ax25-fixes'

Duoming Zhou says:

====================
ax25: fix NPD and UAF bugs when detaching ax25 device

There are NPD and UAF bugs when detaching ax25 device, we
use lock and refcount to mitigate these bugs.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoax25: add refcount in ax25_dev to avoid UAF bugs
Duoming Zhou [Fri, 28 Jan 2022 04:47:16 +0000 (12:47 +0800)]
ax25: add refcount in ax25_dev to avoid UAF bugs

If we dereference ax25_dev after we call kfree(ax25_dev) in
ax25_dev_device_down(), it will lead to concurrency UAF bugs.
There are eight syscall functions suffer from UAF bugs, include
ax25_bind(), ax25_release(), ax25_connect(), ax25_ioctl(),
ax25_getname(), ax25_sendmsg(), ax25_getsockopt() and
ax25_info_show().

One of the concurrency UAF can be shown as below:

  (USE)                       |    (FREE)
                              |  ax25_device_event
                              |    ax25_dev_device_down
ax25_bind                     |    ...
  ...                         |      kfree(ax25_dev)
  ax25_fillin_cb()            |    ...
    ax25_fillin_cb_from_dev() |
  ...                         |

The root cause of UAF bugs is that kfree(ax25_dev) in
ax25_dev_device_down() is not protected by any locks.
When ax25_dev, which there are still pointers point to,
is released, the concurrency UAF bug will happen.

This patch introduces refcount into ax25_dev in order to
guarantee that there are no pointers point to it when ax25_dev
is released.

Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoax25: improve the incomplete fix to avoid UAF and NPD bugs
Duoming Zhou [Fri, 28 Jan 2022 04:47:15 +0000 (12:47 +0800)]
ax25: improve the incomplete fix to avoid UAF and NPD bugs

The previous commit 1ade48d0c27d ("ax25: NPD bug when detaching
AX25 device") introduce lock_sock() into ax25_kill_by_device to
prevent NPD bug. But the concurrency NPD or UAF bug will occur,
when lock_sock() or release_sock() dereferences the ax25_cb->sock.

The NULL pointer dereference bug can be shown as below:

ax25_kill_by_device()        | ax25_release()
                             |   ax25_destroy_socket()
                             |     ax25_cb_del()
  ...                        |     ...
                             |     ax25->sk=NULL;
  lock_sock(s->sk); //(1)    |
  s->ax25_dev = NULL;        |     ...
  release_sock(s->sk); //(2) |
  ...                        |

The root cause is that the sock is set to null before dereference
site (1) or (2). Therefore, this patch extracts the ax25_cb->sock
in advance, and uses ax25_list_lock to protect it, which can synchronize
with ax25_cb_del() and ensure the value of sock is not null before
dereference sites.

The concurrency UAF bug can be shown as below:

ax25_kill_by_device()        | ax25_release()
                             |   ax25_destroy_socket()
  ...                        |   ...
                             |   sock_put(sk); //FREE
  lock_sock(s->sk); //(1)    |
  s->ax25_dev = NULL;        |   ...
  release_sock(s->sk); //(2) |
  ...                        |

The root cause is that the sock is released before dereference
site (1) or (2). Therefore, this patch uses sock_hold() to increase
the refcount of sock and uses ax25_list_lock to protect it, which
can synchronize with ax25_cb_del() in ax25_destroy_socket() and
ensure the sock wil not be released before dereference sites.

Signed-off-by: Duoming Zhou <duoming@zju.edu.cn>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: stmmac: dwmac-visconti: No change to ETHER_CLOCK_SEL for unexpected speed request.
Yuji Ishikawa [Thu, 27 Jan 2022 12:17:14 +0000 (21:17 +0900)]
net: stmmac: dwmac-visconti: No change to ETHER_CLOCK_SEL for unexpected speed request.

Variable clk_sel_val is not initialized in the default case of the first switch statement.
In that case, the function should return immediately without any changes to the hardware.

Reported-by: kernel test robot <lkp@intel.com>
Reported-by: Dan Carpenter <dan.carpenter@oracle.com>
Fixes: b38dd98ff8d0 ("net: stmmac: Add Toshiba Visconti SoCs glue driver")
Signed-off-by: Yuji Ishikawa <yuji2.ishikawa@toshiba.co.jp>
Reviewed-by: Nobuhiro Iwamatsu <nobuhiro1.iwamatsu@toshiba.co.jp>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: amd-xgbe: ensure to reset the tx_timer_active flag
Raju Rangoju [Thu, 27 Jan 2022 06:02:22 +0000 (11:32 +0530)]
net: amd-xgbe: ensure to reset the tx_timer_active flag

Ensure to reset the tx_timer_active flag in xgbe_stop(),
otherwise a port restart may result in tx timeout due to
uncleared flag.

Fixes: c635eaacbf77 ("amd-xgbe: Remove Tx coalescing")
Co-developed-by: Sudheesh Mavila <sudheesh.mavila@amd.com>
Signed-off-by: Sudheesh Mavila <sudheesh.mavila@amd.com>
Signed-off-by: Raju Rangoju <Raju.Rangoju@amd.com>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lore.kernel.org/r/20220127060222.453371-1-Raju.Rangoju@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf
Jakub Kicinski [Fri, 28 Jan 2022 02:53:01 +0000 (18:53 -0800)]
Merge git://git./pub/scm/linux/kernel/git/pablo/nf

Pablo Neira Ayuso says:

====================
Netfilter fixes for net

1) Remove leftovers from flowtable modules, from Geert Uytterhoeven.

2) Missing refcount increment of conntrack template in nft_ct,
   from Florian Westphal.

3) Reduce nft_zone selftest time, also from Florian.

4) Add selftest to cover stateless NAT on fragments, from Florian Westphal.

5) Do not set net_device when for reject packets from the bridge path,
   from Phil Sutter.

6) Cancel register tracking info on nft_byteorder operations.

7) Extend nft_concat_range selftest to cover set reload with no elements,
   from Florian Westphal.

8) Remove useless update of pointer in chain blob builder, reported
   by kbuild test robot.

* git://git.kernel.org/pub/scm/linux/kernel/git/pablo/nf:
  netfilter: nf_tables: remove assignment with no effect in chain blob builder
  selftests: nft_concat_range: add test for reload with no element add/del
  netfilter: nft_byteorder: track register operations
  netfilter: nft_reject_bridge: Fix for missing reply from prerouting
  selftests: netfilter: check stateless nat udp checksum fixup
  selftests: netfilter: reduce zone stress test running time
  netfilter: nft_ct: fix use after free when attaching zone template
  netfilter: Remove flowtable relics
====================

Link: https://lore.kernel.org/r/20220127235235.656931-1-pablo@netfilter.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonet: amd-xgbe: Fix skb data length underflow
Shyam Sundar S K [Thu, 27 Jan 2022 09:20:03 +0000 (14:50 +0530)]
net: amd-xgbe: Fix skb data length underflow

There will be BUG_ON() triggered in include/linux/skbuff.h leading to
intermittent kernel panic, when the skb length underflow is detected.

Fix this by dropping the packet if such length underflows are seen
because of inconsistencies in the hardware descriptors.

Fixes: 622c36f143fc ("amd-xgbe: Fix jumbo MTU processing on newer hardware")
Suggested-by: Tom Lendacky <thomas.lendacky@amd.com>
Signed-off-by: Shyam Sundar S K <Shyam-sundar.S-k@amd.com>
Acked-by: Tom Lendacky <thomas.lendacky@amd.com>
Link: https://lore.kernel.org/r/20220127092003.2812745-1-Shyam-sundar.S-k@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge tag 'net-5.17-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net
Linus Torvalds [Thu, 27 Jan 2022 18:58:39 +0000 (20:58 +0200)]
Merge tag 'net-5.17-rc2' of git://git./linux/kernel/git/netdev/net

Pull networking fixes from Jakub Kicinski:
 "Including fixes from netfilter and can.

  Current release - new code bugs:

   - tcp: add a missing sk_defer_free_flush() in tcp_splice_read()

   - tcp: add a stub for sk_defer_free_flush(), fix CONFIG_INET=n

   - nf_tables: set last expression in register tracking area

   - nft_connlimit: fix memleak if nf_ct_netns_get() fails

   - mptcp: fix removing ids bitmap setting

   - bonding: use rcu_dereference_rtnl when getting active slave

   - fix three cases of sleep in atomic context in drivers: lan966x, gve

   - handful of build fixes for esoteric drivers after netdev->dev_addr
     was made const

  Previous releases - regressions:

   - revert "ipv6: Honor all IPv6 PIO Valid Lifetime values", it broke
     Linux compatibility with USGv6 tests

   - procfs: show net device bound packet types

   - ipv4: fix ip option filtering for locally generated fragments

   - phy: broadcom: hook up soft_reset for BCM54616S

  Previous releases - always broken:

   - ipv4: raw: lock the socket in raw_bind()

   - ipv4: decrease the use of shared IPID generator to decrease the
     chance of attackers guessing the values

   - procfs: fix cross-netns information leakage in /proc/net/ptype

   - ethtool: fix link extended state for big endian

   - bridge: vlan: fix single net device option dumping

   - ping: fix the sk_bound_dev_if match in ping_lookup"

* tag 'net-5.17-rc2' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (86 commits)
  net: bridge: vlan: fix memory leak in __allowed_ingress
  net: socket: rename SKB_DROP_REASON_SOCKET_FILTER
  ipv4: remove sparse error in ip_neigh_gw4()
  ipv4: avoid using shared IP generator for connected sockets
  ipv4: tcp: send zero IPID in SYNACK messages
  ipv4: raw: lock the socket in raw_bind()
  MAINTAINERS: add missing IPv4/IPv6 header paths
  MAINTAINERS: add more files to eth PHY
  net: stmmac: dwmac-sun8i: use return val of readl_poll_timeout()
  net: bridge: vlan: fix single net device option dumping
  net: stmmac: skip only stmmac_ptp_register when resume from suspend
  net: stmmac: configure PTP clock source prior to PTP initialization
  Revert "ipv6: Honor all IPv6 PIO Valid Lifetime values"
  connector/cn_proc: Use task_is_in_init_pid_ns()
  pid: Introduce helper task_is_in_init_pid_ns()
  gve: Fix GFP flags when allocing pages
  net: lan966x: Fix sleep in atomic context when updating MAC table
  net: lan966x: Fix sleep in atomic context when injecting frames
  ethernet: seeq/ether3: don't write directly to netdev->dev_addr
  ethernet: 8390/etherh: don't write directly to netdev->dev_addr
  ...

2 years agonet: bridge: vlan: fix memory leak in __allowed_ingress
Tim Yi [Thu, 27 Jan 2022 07:49:53 +0000 (15:49 +0800)]
net: bridge: vlan: fix memory leak in __allowed_ingress

When using per-vlan state, if vlan snooping and stats are disabled,
untagged or priority-tagged ingress frame will go to check pvid state.
If the port state is forwarding and the pvid state is not
learning/forwarding, untagged or priority-tagged frame will be dropped
but skb memory is not freed.
Should free skb when __allowed_ingress returns false.

Fixes: a580c76d534c ("net: bridge: vlan: add per-vlan state")
Signed-off-by: Tim Yi <tim.yi@pica8.com>
Acked-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Link: https://lore.kernel.org/r/20220127074953.12632-1-tim.yi@pica8.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agonetfilter: nf_tables: remove assignment with no effect in chain blob builder
Pablo Neira Ayuso [Thu, 27 Jan 2022 12:40:38 +0000 (13:40 +0100)]
netfilter: nf_tables: remove assignment with no effect in chain blob builder

cppcheck possible warnings:

>> net/netfilter/nf_tables_api.c:2014:2: warning: Assignment of function parameter has no effect outside the function. Did you forget dereferencing it? [uselessAssignmentPtrArg]
    ptr += offsetof(struct nft_rule_dp, data);
    ^

Reported-by: kernel test robot <yujie.liu@intel.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agonet: socket: rename SKB_DROP_REASON_SOCKET_FILTER
Menglong Dong [Thu, 27 Jan 2022 09:13:01 +0000 (17:13 +0800)]
net: socket: rename SKB_DROP_REASON_SOCKET_FILTER

Rename SKB_DROP_REASON_SOCKET_FILTER, which is used
as the reason of skb drop out of socket filter before
it's part of a released kernel. It will be used for
more protocols than just TCP in future series.

Signed-off-by: Menglong Dong <imagedong@tencent.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/all/20220127091308.91401-2-imagedong@tencent.com/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoipv4: remove sparse error in ip_neigh_gw4()
Eric Dumazet [Thu, 27 Jan 2022 01:34:04 +0000 (17:34 -0800)]
ipv4: remove sparse error in ip_neigh_gw4()

./include/net/route.h:373:48: warning: incorrect type in argument 2 (different base types)
./include/net/route.h:373:48:    expected unsigned int [usertype] key
./include/net/route.h:373:48:    got restricted __be32 [usertype] daddr

Fixes: 5c9f7c1dfc2e ("ipv4: Add helpers for neigh lookup for nexthop")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20220127013404.1279313-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge branch 'ipv4-less-uses-of-shared-ip-generator'
Jakub Kicinski [Thu, 27 Jan 2022 16:37:05 +0000 (08:37 -0800)]
Merge branch 'ipv4-less-uses-of-shared-ip-generator'

Eric Dumazet says:

====================
ipv4: less uses of shared IP generator

From: Eric Dumazet <edumazet@google.com>

We keep receiving research reports based on linux IPID generation.

Before breaking part of the Internet by switching to pure
random generator, this series reduces the need for the
shared IP generator for TCP sockets.
====================

Link: https://lore.kernel.org/r/20220127011022.1274803-1-eric.dumazet@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoipv4: avoid using shared IP generator for connected sockets
Eric Dumazet [Thu, 27 Jan 2022 01:10:22 +0000 (17:10 -0800)]
ipv4: avoid using shared IP generator for connected sockets

ip_select_ident_segs() has been very conservative about using
the connected socket private generator only for packets with IP_DF
set, claiming it was needed for some VJ compression implementations.

As mentioned in this referenced document, this can be abused.
(Ref: Off-Path TCP Exploits of the Mixed IPID Assignment)

Before switching to pure random IPID generation and possibly hurt
some workloads, lets use the private inet socket generator.

Not only this will remove one vulnerability, this will also
improve performance of TCP flows using pmtudisc==IP_PMTUDISC_DONT

Fixes: 73f156a6e8c1 ("inetpeer: get rid of ip_id_count")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reported-by: Ray Che <xijiache@gmail.com>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoipv4: tcp: send zero IPID in SYNACK messages
Eric Dumazet [Thu, 27 Jan 2022 01:10:21 +0000 (17:10 -0800)]
ipv4: tcp: send zero IPID in SYNACK messages

In commit 431280eebed9 ("ipv4: tcp: send zero IPID for RST and
ACK sent in SYN-RECV and TIME-WAIT state") we took care of some
ctl packets sent by TCP.

It turns out we need to use a similar strategy for SYNACK packets.

By default, they carry IP_DF and IPID==0, but there are ways
to ask them to use the hashed IP ident generator and thus
be used to build off-path attacks.
(Ref: Off-Path TCP Exploits of the Mixed IPID Assignment)

One of this way is to force (before listener is started)
echo 1 >/proc/sys/net/ipv4/ip_no_pmtu_disc

Another way is using forged ICMP ICMP_FRAG_NEEDED
with a very small MTU (like 68) to force a false return from
ip_dont_fragment()

In this patch, ip_build_and_send_pkt() uses the following
heuristics.

1) Most SYNACK packets are smaller than IPV4_MIN_MTU and therefore
can use IP_DF regardless of the listener or route pmtu setting.

2) In case the SYNACK packet is bigger than IPV4_MIN_MTU,
we use prandom_u32() generator instead of the IPv4 hashed ident one.

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: Ray Che <xijiache@gmail.com>
Reviewed-by: David Ahern <dsahern@kernel.org>
Cc: Geoff Alexander <alexandg@cs.unm.edu>
Cc: Willy Tarreau <w@1wt.eu>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agodrm/vmwgfx: Fix stale file descriptors on failed usercopy
Mathias Krause [Thu, 27 Jan 2022 08:34:19 +0000 (18:34 +1000)]
drm/vmwgfx: Fix stale file descriptors on failed usercopy

A failing usercopy of the fence_rep object will lead to a stale entry in
the file descriptor table as put_unused_fd() won't release it. This
enables userland to refer to a dangling 'file' object through that still
valid file descriptor, leading to all kinds of use-after-free
exploitation scenarios.

Fix this by deferring the call to fd_install() until after the usercopy
has succeeded.

Fixes: c906965dee22 ("drm/vmwgfx: Add export fence to file descriptor support")
Signed-off-by: Mathias Krause <minipli@grsecurity.net>
Signed-off-by: Zack Rusin <zackr@vmware.com>
Signed-off-by: Dave Airlie <airlied@redhat.com>
Signed-off-by: Linus Torvalds <torvalds@linux-foundation.org>
2 years agoipv4: raw: lock the socket in raw_bind()
Eric Dumazet [Thu, 27 Jan 2022 00:51:16 +0000 (16:51 -0800)]
ipv4: raw: lock the socket in raw_bind()

For some reason, raw_bind() forgot to lock the socket.

BUG: KCSAN: data-race in __ip4_datagram_connect / raw_bind

write to 0xffff8881170d4308 of 4 bytes by task 5466 on cpu 0:
 raw_bind+0x1b0/0x250 net/ipv4/raw.c:739
 inet_bind+0x56/0xa0 net/ipv4/af_inet.c:443
 __sys_bind+0x14b/0x1b0 net/socket.c:1697
 __do_sys_bind net/socket.c:1708 [inline]
 __se_sys_bind net/socket.c:1706 [inline]
 __x64_sys_bind+0x3d/0x50 net/socket.c:1706
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

read to 0xffff8881170d4308 of 4 bytes by task 5468 on cpu 1:
 __ip4_datagram_connect+0xb7/0x7b0 net/ipv4/datagram.c:39
 ip4_datagram_connect+0x2a/0x40 net/ipv4/datagram.c:89
 inet_dgram_connect+0x107/0x190 net/ipv4/af_inet.c:576
 __sys_connect_file net/socket.c:1900 [inline]
 __sys_connect+0x197/0x1b0 net/socket.c:1917
 __do_sys_connect net/socket.c:1927 [inline]
 __se_sys_connect net/socket.c:1924 [inline]
 __x64_sys_connect+0x3d/0x50 net/socket.c:1924
 do_syscall_x64 arch/x86/entry/common.c:50 [inline]
 do_syscall_64+0x44/0xd0 arch/x86/entry/common.c:80
 entry_SYSCALL_64_after_hwframe+0x44/0xae

value changed: 0x00000000 -> 0x0003007f

Reported by Kernel Concurrency Sanitizer on:
CPU: 1 PID: 5468 Comm: syz-executor.5 Not tainted 5.17.0-rc1-syzkaller #0
Hardware name: Google Google Compute Engine/Google Compute Engine, BIOS Google 01/01/2011

Fixes: 1da177e4c3f4 ("Linux-2.6.12-rc2")
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reported-by: syzbot <syzkaller@googlegroups.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMAINTAINERS: add missing IPv4/IPv6 header paths
Jakub Kicinski [Wed, 26 Jan 2022 22:55:35 +0000 (14:55 -0800)]
MAINTAINERS: add missing IPv4/IPv6 header paths

Add missing headers to the IP entry.

Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMAINTAINERS: add more files to eth PHY
Jakub Kicinski [Wed, 26 Jan 2022 20:24:24 +0000 (12:24 -0800)]
MAINTAINERS: add more files to eth PHY

include/linux/linkmode.h and include/linux/mii.h
do not match anything in MAINTAINERS. Looks like
they should be under Ethernet PHY.

Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: stmmac: dwmac-sun8i: use return val of readl_poll_timeout()
Jisheng Zhang [Wed, 26 Jan 2022 16:52:15 +0000 (00:52 +0800)]
net: stmmac: dwmac-sun8i: use return val of readl_poll_timeout()

When readl_poll_timeout() timeout, we'd better directly use its return
value.

Before this patch:
[    2.145528] dwmac-sun8i: probe of 4500000.ethernet failed with error -14

After this patch:
[    2.138520] dwmac-sun8i: probe of 4500000.ethernet failed with error -110

Signed-off-by: Jisheng Zhang <jszhang@kernel.org>
Acked-by: Jernej Skrabec <jernej.skrabec@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: bridge: vlan: fix single net device option dumping
Nikolay Aleksandrov [Wed, 26 Jan 2022 13:10:25 +0000 (15:10 +0200)]
net: bridge: vlan: fix single net device option dumping

When dumping vlan options for a single net device we send the same
entries infinitely because user-space expects a 0 return at the end but
we keep returning skb->len and restarting the dump on retry. Fix it by
returning the value from br_vlan_dump_dev() if it completed or there was
an error. The only case that must return skb->len is when the dump was
incomplete and needs to continue (-EMSGSIZE).

Reported-by: Benjamin Poirier <bpoirier@nvidia.com>
Fixes: 8dcea187088b ("net: bridge: vlan: add rtm definitions and dump support")
Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch 'stmmac-ptp-fix'
David S. Miller [Thu, 27 Jan 2022 13:46:18 +0000 (13:46 +0000)]
Merge branch 'stmmac-ptp-fix'

Mohammad Athari Bin Ismail says:

====================
Fix PTP issue in stmmac

This patch series to fix PTP issue in stmmac related to:
1/ PTP clock source configuration during initialization.
2/ PTP initialization during resume from suspend.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: stmmac: skip only stmmac_ptp_register when resume from suspend
Mohammad Athari Bin Ismail [Wed, 26 Jan 2022 09:47:23 +0000 (17:47 +0800)]
net: stmmac: skip only stmmac_ptp_register when resume from suspend

When resume from suspend, besides skipping PTP registration, it also
skipping PTP HW initialization. This could cause PTP clock not able to
operate properly when resume from suspend.

To fix this, only stmmac_ptp_register() is skipped when resume from
suspend.

Fixes: fe1319291150 ("stmmac: Don't init ptp again when resume from suspend/hibernation")
Cc: <stable@vger.kernel.org> # 5.15.x
Signed-off-by: Mohammad Athari Bin Ismail <mohammad.athari.ismail@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: stmmac: configure PTP clock source prior to PTP initialization
Mohammad Athari Bin Ismail [Wed, 26 Jan 2022 09:47:22 +0000 (17:47 +0800)]
net: stmmac: configure PTP clock source prior to PTP initialization

For Intel platform, it is required to configure PTP clock source prior PTP
initialization in MAC. So, need to move ptp_clk_freq_config execution from
stmmac_ptp_register() to stmmac_init_ptp().

Fixes: 76da35dc99af ("stmmac: intel: Add PSE and PCH PTP clock source selection")
Cc: <stable@vger.kernel.org> # 5.15.x
Signed-off-by: Mohammad Athari Bin Ismail <mohammad.athari.ismail@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoRevert "ipv6: Honor all IPv6 PIO Valid Lifetime values"
Guillaume Nault [Wed, 26 Jan 2022 15:38:52 +0000 (16:38 +0100)]
Revert "ipv6: Honor all IPv6 PIO Valid Lifetime values"

This reverts commit b75326c201242de9495ff98e5d5cff41d7fc0d9d.

This commit breaks Linux compatibility with USGv6 tests. The RFC this
commit was based on is actually an expired draft: no published RFC
currently allows the new behaviour it introduced.

Without full IETF endorsement, the flash renumbering scenario this
patch was supposed to enable is never going to work, as other IPv6
equipements on the same LAN will keep the 2 hours limit.

Fixes: b75326c20124 ("ipv6: Honor all IPv6 PIO Valid Lifetime values")
Signed-off-by: Guillaume Nault <gnault@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge tag 'rpmsg-v5.17-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/remote...
Linus Torvalds [Thu, 27 Jan 2022 09:23:26 +0000 (11:23 +0200)]
Merge tag 'rpmsg-v5.17-fixes' of git://git./linux/kernel/git/remoteproc/linux

Pull rpmsg fixes from Bjorn Andersson:
 "The cdev cleanup in the rpmsg_char driver was not performed properly,
  resulting in unpredicable behaviour when the parent remote processor
  is stopped with any of the cdevs open by a client.

  Two patches transitions the implementation to use cdev_device_add()
  and cdev_del_device(), to capture the relationship between the two
  objects, and relocates the incorrectly placed cdev_del()"

* tag 'rpmsg-v5.17-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/remoteproc/linux:
  rpmsg: char: Fix race between the release of rpmsg_eptdev and cdev
  rpmsg: char: Fix race between the release of rpmsg_ctrldev and cdev

2 years agoMerge tag 'rproc-v5.17-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/remote...
Linus Torvalds [Thu, 27 Jan 2022 09:19:20 +0000 (11:19 +0200)]
Merge tag 'rproc-v5.17-fixes' of git://git./linux/kernel/git/remoteproc/linux

Pull remoteproc fix from Bjorn Andersson:
 "The interaction between the various Qualcomm remoteproc drivers and
  the Qualcomm 'QMP' driver (used to communicate with the
  power-management hardware) was reworked in v5.17-rc1, but failed to
  account for the new Kconfig dependency"

* tag 'rproc-v5.17-fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/remoteproc/linux:
  remoteproc: qcom: q6v5: fix service routines build errors

2 years agoMAINTAINERS: Remove Harry Morris bouncing address
Miquel Raynal [Tue, 25 Jan 2022 12:14:26 +0000 (13:14 +0100)]
MAINTAINERS: Remove Harry Morris bouncing address

Harry's e-mail address from Cascoda bounces, I have not found any
contributions from him since 2018 so let's drop the Maintainer entry
from the CA8210 driver and mark it Orphan.

Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-7-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agonet: ieee802154: Return meaningful error codes from the netlink helpers
Miquel Raynal [Tue, 25 Jan 2022 12:14:25 +0000 (13:14 +0100)]
net: ieee802154: Return meaningful error codes from the netlink helpers

Returning -1 does not indicate anything useful.

Use a standard and meaningful error code instead.

Fixes: a26c5fd7622d ("nl802154: add support for security layer")
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-6-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agonet: ieee802154: ca8210: Stop leaking skb's
Miquel Raynal [Tue, 25 Jan 2022 12:14:24 +0000 (13:14 +0100)]
net: ieee802154: ca8210: Stop leaking skb's

Upon error the ieee802154_xmit_complete() helper is not called. Only
ieee802154_wake_queue() is called manually. We then leak the skb
structure.

Free the skb structure upon error before returning.

Fixes: ded845a781a5 ("ieee802154: Add CA8210 IEEE 802.15.4 device driver")
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-5-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agonet: ieee802154: at86rf230: Stop leaking skb's
Miquel Raynal [Tue, 25 Jan 2022 12:14:23 +0000 (13:14 +0100)]
net: ieee802154: at86rf230: Stop leaking skb's

Upon error the ieee802154_xmit_complete() helper is not called. Only
ieee802154_wake_queue() is called manually. In the Tx case we then leak
the skb structure.

Free the skb structure upon error before returning when appropriate.

As the 'is_tx = 0' cannot be moved in the complete handler because of a
possible race between the delay in switching to STATE_RX_AACK_ON and a
new interrupt, we introduce an intermediate 'was_tx' boolean just for
this purpose.

There is no Fixes tag applying here, many changes have been made on this
area and the issue kind of always existed.

Suggested-by: Alexander Aring <alex.aring@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-4-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agonet: ieee802154: mcr20a: Fix lifs/sifs periods
Miquel Raynal [Tue, 25 Jan 2022 12:14:22 +0000 (13:14 +0100)]
net: ieee802154: mcr20a: Fix lifs/sifs periods

These periods are expressed in time units (microseconds) while 40 and 12
are the number of symbol durations these periods will last. We need to
multiply them both with phy->symbol_duration in order to get these
values in microseconds.

Fixes: 8c6ad9cc5157 ("ieee802154: Add NXP MCR20A IEEE 802.15.4 transceiver driver")
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-3-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agonet: ieee802154: hwsim: Ensure proper channel selection at probe time
Miquel Raynal [Tue, 25 Jan 2022 12:14:21 +0000 (13:14 +0100)]
net: ieee802154: hwsim: Ensure proper channel selection at probe time

Drivers are expected to set the PHY current_channel and current_page
according to their default state. The hwsim driver is advertising being
configured on channel 13 by default but that is not reflected in its own
internal pib structure. In order to ensure that this driver consider the
current channel as being 13 internally, we at least need to set the
pib->channel field to 13.

Fixes: f25da51fdc38 ("ieee802154: hwsim: add replacement for fakelb")
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
[stefan@datenfreihafen.org: fixed assigment from page to channel]
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220125121426.848337-2-miquel.raynal@bootlin.com
Signed-off-by: Stefan Schmidt <stefan@datenfreihafen.org>
2 years agoMerge branch 'pid-introduce-helper-task_is_in_root_ns'
Jakub Kicinski [Thu, 27 Jan 2022 02:57:12 +0000 (18:57 -0800)]
Merge branch 'pid-introduce-helper-task_is_in_root_ns'

Leo Yan says:

====================
pid: Introduce helper task_is_in_root_ns()

This patch series introduces a helper function task_is_in_init_pid_ns()
to replace open code.  The two patches are extracted from the original
series [1] for network subsystem.

As a plan, we can firstly land this patch set into kernel 5.18; there
have 5 patches are left out from original series [1], as a next step,
I will resend them for appropriate linux-next merging.

[1] https://lore.kernel.org/lkml/20211208083320.472503-1-leo.yan@linaro.org/
====================

Link: https://lore.kernel.org/r/20220126050427.605628-1-leo.yan@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoconnector/cn_proc: Use task_is_in_init_pid_ns()
Leo Yan [Wed, 26 Jan 2022 05:04:27 +0000 (13:04 +0800)]
connector/cn_proc: Use task_is_in_init_pid_ns()

This patch replaces open code with task_is_in_init_pid_ns() to check if
a task is in root PID namespace.

Signed-off-by: Leo Yan <leo.yan@linaro.org>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agopid: Introduce helper task_is_in_init_pid_ns()
Leo Yan [Wed, 26 Jan 2022 05:04:26 +0000 (13:04 +0800)]
pid: Introduce helper task_is_in_init_pid_ns()

Currently the kernel uses open code in multiple places to check if a
task is in the root PID namespace with the kind of format:

  if (task_active_pid_ns(current) == &init_pid_ns)
      do_something();

This patch creates a new helper function, task_is_in_init_pid_ns(), it
returns true if a passed task is in the root PID namespace, otherwise
returns false.  So it will be used to replace open codes.

Suggested-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Signed-off-by: Leo Yan <leo.yan@linaro.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Acked-by: Suzuki K Poulose <suzuki.poulose@arm.com>
Acked-by: Balbir Singh <bsingharora@gmail.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agogve: Fix GFP flags when allocing pages
Catherine Sullivan [Wed, 26 Jan 2022 00:38:43 +0000 (16:38 -0800)]
gve: Fix GFP flags when allocing pages

Use GFP_ATOMIC when allocating pages out of the hotpath,
continue to use GFP_KERNEL when allocating pages during setup.

GFP_KERNEL will allow blocking which allows it to succeed
more often in a low memory enviornment but in the hotpath we do
not want to allow the allocation to block.

Fixes: f5cedc84a30d2 ("gve: Add transmit and receive support")
Signed-off-by: Catherine Sullivan <csully@google.com>
Signed-off-by: David Awogbemila <awogbemila@google.com>
Link: https://lore.kernel.org/r/20220126003843.3584521-1-awogbemila@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoselftests: nft_concat_range: add test for reload with no element add/del
Florian Westphal [Wed, 26 Jan 2022 11:54:54 +0000 (12:54 +0100)]
selftests: nft_concat_range: add test for reload with no element add/del

Add a specific test for the reload issue fixed with
commit 23c54263efd7cb ("netfilter: nft_set_pipapo: allocate pcpu scratch maps on clone").

Add to set, then flush set content + restore without other add/remove in
the transaction.

On kernels before the fix, this test case fails:
  net,mac with reload    [FAIL]

Signed-off-by: Florian Westphal <fw@strlen.de>
Reviewed-by: Stefano Brivio <sbrivio@redhat.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agonetfilter: nft_byteorder: track register operations
Pablo Neira Ayuso [Wed, 26 Jan 2022 00:46:58 +0000 (01:46 +0100)]
netfilter: nft_byteorder: track register operations

Cancel tracking for byteorder operation, otherwise selector + byteorder
operation is incorrectly reduced if source and destination registers are
the same.

Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agonetfilter: nft_reject_bridge: Fix for missing reply from prerouting
Phil Sutter [Tue, 25 Jan 2022 19:06:03 +0000 (20:06 +0100)]
netfilter: nft_reject_bridge: Fix for missing reply from prerouting

Prior to commit fa538f7cf05aa ("netfilter: nf_reject: add reject skbuff
creation helpers"), nft_reject_bridge did not assign to nskb->dev before
passing nskb on to br_forward(). The shared skbuff creation helpers
introduced in above commit do which seems to confuse br_forward() as
reject statements in prerouting hook won't emit a packet anymore.

Fix this by simply passing NULL instead of 'dev' to the helpers - they
use the pointer for just that assignment, nothing else.

Fixes: fa538f7cf05aa ("netfilter: nf_reject: add reject skbuff creation helpers")
Signed-off-by: Phil Sutter <phil@nwl.cc>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agoselftests: netfilter: check stateless nat udp checksum fixup
Florian Westphal [Mon, 24 Jan 2022 21:09:15 +0000 (22:09 +0100)]
selftests: netfilter: check stateless nat udp checksum fixup

Add a test that sends large udp packet (which is fragmented)
via a stateless nft nat rule, i.e. 'ip saddr set 10.2.3.4'
and check that the datagram is received by peer.

On kernels without
commit 4e1860a38637 ("netfilter: nft_payload: do not update layer 4 checksum when mangling fragments")',
this will fail with:

cmp: EOF on /tmp/tmp.V1q0iXJyQF which is empty
-rw------- 1 root root 4096 Jan 24 22:03 /tmp/tmp.Aaqnq4rBKS
-rw------- 1 root root    0 Jan 24 22:03 /tmp/tmp.V1q0iXJyQF
ERROR: in and output file mismatch when checking udp with stateless nat
FAIL: nftables v1.0.0 (Fearless Fosdick #2)

On patched kernels, this will show:
PASS: IP statless for ns2-PFp89amx

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agoselftests: netfilter: reduce zone stress test running time
Florian Westphal [Sun, 23 Jan 2022 14:45:54 +0000 (15:45 +0100)]
selftests: netfilter: reduce zone stress test running time

This selftests needs almost 3 minutes to complete, reduce the
insertes zones to 1000.  Test now completes in about 20 seconds.

Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agonetfilter: nft_ct: fix use after free when attaching zone template
Florian Westphal [Sun, 23 Jan 2022 14:24:00 +0000 (15:24 +0100)]
netfilter: nft_ct: fix use after free when attaching zone template

The conversion erroneously removed the refcount increment.
In case we can use the percpu template, we need to increment
the refcount, else it will be released when the skb gets freed.

In case the slowpath is taken, the new template already has a
refcount of 1.

Fixes: 719774377622 ("netfilter: conntrack: convert to refcount_t api")
Reported-by: kernel test robot <oliver.sang@intel.com>
Signed-off-by: Florian Westphal <fw@strlen.de>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agonetfilter: Remove flowtable relics
Geert Uytterhoeven [Sun, 23 Jan 2022 12:57:17 +0000 (13:57 +0100)]
netfilter: Remove flowtable relics

NF_FLOW_TABLE_IPV4 and NF_FLOW_TABLE_IPV6 are invisble, selected by
nothing (so they can no longer be enabled), and their last real users
have been removed (nf_flow_table_ipv6.c is empty).

Clean up the leftovers.

Fixes: c42ba4290b2147aa ("netfilter: flowtable: remove ipv4/ipv6 modules")
Signed-off-by: Geert Uytterhoeven <geert@linux-m68k.org>
Signed-off-by: Pablo Neira Ayuso <pablo@netfilter.org>
2 years agoMerge branch 'lan966x-fixes'
David S. Miller [Wed, 26 Jan 2022 15:45:49 +0000 (15:45 +0000)]
Merge branch 'lan966x-fixes'

Horatiu Vultur says:

====================
net: lan966x: Fixes for sleep in atomic context

This patch series contains 2 fixes for lan966x that is sleeping in atomic
context. The first patch fixes the injection of the frames while the second
one fixes the updating of the MAC table.

v1->v2:
 - correct the fix tag in the second patch, it was using the wrong sha.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan966x: Fix sleep in atomic context when updating MAC table
Horatiu Vultur [Tue, 25 Jan 2022 11:48:16 +0000 (12:48 +0100)]
net: lan966x: Fix sleep in atomic context when updating MAC table

The function lan966x_mac_wait_for_completion is used to poll the status
of the MAC table using the function readx_poll_timeout. The problem with
this function is that is called also from atomic context. Therefore
update the function to use readx_poll_timeout_atomic.

Fixes: e18aba8941b40b ("net: lan966x: add mactable support")
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agonet: lan966x: Fix sleep in atomic context when injecting frames
Horatiu Vultur [Tue, 25 Jan 2022 11:48:15 +0000 (12:48 +0100)]
net: lan966x: Fix sleep in atomic context when injecting frames

On lan966x, when injecting a frame it was polling the register
QS_INJ_STATUS to see if it can continue with the injection of the frame.
The problem was that it was using readx_poll_timeout which could sleep
in atomic context.
This patch fixes this issue by using readx_poll_timeout_atomic.

Fixes: d28d6d2e37d10d ("net: lan966x: add port module support")
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoMerge branch 'dev_addr-const-fixes'
David S. Miller [Wed, 26 Jan 2022 15:40:02 +0000 (15:40 +0000)]
Merge branch 'dev_addr-const-fixes'

Jakub Kicinski says:

====================
ethernet: fix some esoteric drivers after netdev->dev_addr constification

Looking at recent fixes for drivers which don't get included with
allmodconfig builds I thought it's worth grepping for more instances of:

  dev->dev_addr\[.*\] =

This set contains the fixes.

v2: add last 3 patches which fix drivers for the RiscPC ARM platform.
Thanks to Arnd Bergmann for explaining how to build test that.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: seeq/ether3: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:38:01 +0000 (16:38 -0800)]
ethernet: seeq/ether3: don't write directly to netdev->dev_addr

netdev->dev_addr is const now.

Compile tested rpc_defconfig w/ GCC 8.5.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: 8390/etherh: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:38:00 +0000 (16:38 -0800)]
ethernet: 8390/etherh: don't write directly to netdev->dev_addr

netdev->dev_addr is const now.

Compile tested rpc_defconfig w/ GCC 8.5.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: i825xx: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:37:59 +0000 (16:37 -0800)]
ethernet: i825xx: don't write directly to netdev->dev_addr

netdev->dev_addr is const now.

Compile tested rpc_defconfig w/ GCC 8.5.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: broadcom/sb1250-mac: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:37:58 +0000 (16:37 -0800)]
ethernet: broadcom/sb1250-mac: don't write directly to netdev->dev_addr

netdev->dev_addr is const now.

Compile tested bigsur_defconfig and sb1250_swarm_defconfig.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: tundra: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:37:57 +0000 (16:37 -0800)]
ethernet: tundra: don't write directly to netdev->dev_addr

netdev->dev_addr is const now.

Maintain the questionable offsetting in ndo_set_mac_address.

Compile tested holly_defconfig and mpc7448_hpc2_defconfig.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agoethernet: 3com/typhoon: don't write directly to netdev->dev_addr
Jakub Kicinski [Wed, 26 Jan 2022 00:37:56 +0000 (16:37 -0800)]
ethernet: 3com/typhoon: don't write directly to netdev->dev_addr

This driver casts off the const and writes directly to netdev->dev_addr.
This will result in a MAC address tree corruption and a warning.

Compile tested ppc6xx_defconfig.

Fixes: adeef3e32146 ("net: constify netdev->dev_addr")
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>
2 years agosch_htb: Fail on unsupported parameters when offload is requested
Maxim Mikityanskiy [Tue, 25 Jan 2022 10:06:54 +0000 (12:06 +0200)]
sch_htb: Fail on unsupported parameters when offload is requested

The current implementation of HTB offload doesn't support some
parameters. Instead of ignoring them, actively return the EINVAL error
when they are set to non-defaults.

As this patch goes to stable, the driver API is not changed here. If
future drivers support more offload parameters, the checks can be moved
to the driver side.

Note that the buffer and cbuffer parameters are also not supported, but
the tc userspace tool assigns some default values derived from rate and
ceil, and identifying these defaults in sch_htb would be unreliable, so
they are still ignored.

Fixes: d03b195b5aa0 ("sch_htb: Hierarchical QoS hardware offload")
Reported-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Maxim Mikityanskiy <maximmi@nvidia.com>
Reviewed-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20220125100654.424570-1-maximmi@nvidia.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
2 years agoMerge tag 'nfs-for-5.17-1' of git://git.linux-nfs.org/projects/anna/linux-nfs
Linus Torvalds [Tue, 25 Jan 2022 18:16:03 +0000 (20:16 +0200)]
Merge tag 'nfs-for-5.17-1' of git://git.linux-nfs.org/projects/anna/linux-nfs

Pull NFS client updates from Anna Schumaker:
 "New Features:

   - Basic handling for case insensitive filesystems

   - Initial support for fs_locations and server trunking

  Bugfixes and Cleanups:

   - Cleanups to how the "struct cred *" is handled for the
     nfs_access_entry

   - Ensure the server has an up to date ctimes before hardlinking or
     renaming

   - Update 'blocks used' after writeback, fallocate, and clone

   - nfs_atomic_open() fixes

   - Improvements to sunrpc tracing

   - Various null check & indenting related cleanups

   - Some improvements to the sunrpc sysfs code:
      - Use default_groups in kobj_type
      - Fix some potential races and reference leaks

   - A few tracepoint cleanups in xprtrdma"

[ This should have gone in during the merge window, but didn't. The
  original pull request - sent during the merge window - had gotten
  marked as spam and discarded due missing DKIM headers in the email
  from Anna.   - Linus ]

* tag 'nfs-for-5.17-1' of git://git.linux-nfs.org/projects/anna/linux-nfs: (35 commits)
  SUNRPC: Don't dereference xprt->snd_task if it's a cookie
  xprtrdma: Remove definitions of RPCDBG_FACILITY
  xprtrdma: Remove final dprintk call sites from xprtrdma
  sunrpc: Fix potential race conditions in rpc_sysfs_xprt_state_change()
  net/sunrpc: fix reference count leaks in rpc_sysfs_xprt_state_change
  NFSv4.1 test and add 4.1 trunking transport
  SUNRPC allow for unspecified transport time in rpc_clnt_add_xprt
  NFSv4 handle port presence in fs_location server string
  NFSv4 expose nfs_parse_server_name function
  NFSv4.1 query for fs_location attr on a new file system
  NFSv4 store server support for fs_location attribute
  NFSv4 remove zero number of fs_locations entries error check
  NFSv4: nfs_atomic_open() can race when looking up a non-regular file
  NFSv4: Handle case where the lookup of a directory fails
  NFSv42: Fallocate and clone should also request 'blocks used'
  NFSv4: Allow writebacks to request 'blocks used'
  SUNRPC: use default_groups in kobj_type
  NFS: use default_groups in kobj_type
  NFS: Fix the verifier for case sensitive filesystem in nfs_atomic_open()
  NFS: Add a helper to remove case-insensitive aliases
  ...