Skip to content

L
linux
Switch branch/tag
  1. 08 Mar, 2022 7 commits
    • Christophe JAILLET's avatar
      ice: Don't use GFP_KERNEL in atomic context · 3d97f1af
      Christophe JAILLET authored 
      ice_misc_intr() is an irq handler. It should not sleep.

Use GFP_ATOMIC instead of GFP_KERNEL when allocating some memory.

Fixes: 348048e7 ("ice: Implement iidc operations")
Signed-off-by: default avatarChristophe JAILLET <christophe.jaillet@wanadoo.fr>
Tested-by: default avatarLeszek Kaliszczuk <leszek.kaliszczuk@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      3d97f1af
    • Dave Ertman's avatar
      ice: Fix error with handling of bonding MTU · 97b01291
      Dave Ertman authored 
      When a bonded interface is destroyed, .ndo_change_mtu can be called
during the tear-down process while the RTNL lock is held.  This is a
problem since the auxiliary driver linked to the LAN driver needs to be
notified of the MTU change, and this requires grabbing a device_lock on
the auxiliary_device's dev.  Currently this is being attempted in the
same execution context as the call to .ndo_change_mtu which is causing a
dead-lock.

Move the notification of the changed MTU to a separate execution context
(watchdog service task) and eliminate the "before" notification.

Fixes: 348048e7 ("ice: Implement iidc operations")
Signed-off-by: default avatarDave Ertman <david.m.ertman@intel.com>
Tested-by: default avatarJonathan Toppins <jtoppins@redhat.com>
Tested-by: Gurucharan G <gurucharanx.g@intel.com> (A Contingent worker at Intel)
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      97b01291
    • Jacob Keller's avatar
      ice: stop disabling VFs due to PF error responses · 79498d5a
      Jacob Keller authored 
      The ice_vc_send_msg_to_vf function has logic to detect "failure"
responses being sent to a VF. If a VF is sent more than
ICE_DFLT_NUM_INVAL_MSGS_ALLOWED then the VF is marked as disabled.
Almost identical logic also existed in the i40e driver.

This logic was added to the ice driver in commit 1071a835 ("ice:
Implement virtchnl commands for AVF support") which itself copied from
the i40e implementation in commit 5c3c48ac ("i40e: implement virtual
device interface").

Neither commit provides a proper explanation or justification of the
check. In fact, later commits to i40e changed the logic to allow
bypassing the check in some specific instances.

The "logic" for this seems to be that error responses somehow indicate a
malicious VF. This is not really true. The PF might be sending an error
for any number of reasons such as lack of resources, etc.

Additionally, this causes the PF to log an info message for every failed
VF response which may confuse users, and can spam the kernel log.

This behavior is not documented as part of any requirement for our
products and other operating system drivers such as the FreeBSD
implementation of our drivers do not include this type of check.

In fact, the change from dev_err to dev_info in i40e commit 18b7af57
("i40e: Lower some message levels") explains that these messages
typically don't actually indicate a real issue. It is quite likely that
a user who hits this in practice will be very confused as the VF will be
disabled without an obvious way to recover.

We already have robust malicious driver detection logic using actual
hardware detection mechanisms that detect and prevent invalid device
usage. Remove the logic since its not a documented requirement and the
behavior is not intuitive.

Fixes: 1071a835 ("ice: Implement virtchnl commands for AVF support")
Signed-off-by: default avatarJacob Keller <jacob.e.keller@intel.com>
Tested-by: default avatarKonrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      79498d5a
    • Jacob Keller's avatar
      i40e: stop disabling VFs due to PF error responses · 5710ab79
      Jacob Keller authored 
      The i40e_vc_send_msg_to_vf_ex (and its wrapper i40e_vc_send_msg_to_vf)
function has logic to detect "failure" responses sent to the VF. If a VF
is sent more than I40E_DEFAULT_NUM_INVALID_MSGS_ALLOWED, then the VF is
marked as disabled. In either case, a dev_info message is printed
stating that a VF opcode failed.

This logic originates from the early implementation of VF support in
commit 5c3c48ac ("i40e: implement virtual device interface").

That commit did not go far enough. The "logic" for this behavior seems
to be that error responses somehow indicate a malicious VF. This is not
really true. The PF might be sending an error for any number of reasons
such as lacking resources, an unsupported operation, etc. This does not
indicate a malicious VF. We already have a separate robust malicious VF
detection which relies on hardware logic to detect and prevent a variety
of behaviors.

There is no justification for this behavior in the original
implementation. In fact, a later commit 18b7af57 ("i40e: Lower some
message levels") reduced the opcode failure message from a dev_err to a
dev_info. In addition, recent commit 01cbf508 ("i40e: Fix to not
show opcode msg on unsuccessful VF MAC change") changed the logic to
allow quieting it for expected failures.

That commit prevented this logic from kicking in for specific
circumstances. This change did not go far enough. The behavior is not
documented nor is it part of any requirement for our products. Other
operating systems such as the FreeBSD implementation of our driver do
not include this logic.

It is clear this check does not make sense, and causes problems which
led to ugly workarounds.

Fix this by just removing the entire logic and the need for the
i40e_vc_send_msg_to_vf_ex function.

Fixes: 01cbf508 ("i40e: Fix to not show opcode msg on unsuccessful VF MAC change")
Fixes: 5c3c48ac ("i40e: implement virtual device interface")
Signed-off-by: default avatarJacob Keller <jacob.e.keller@intel.com>
Tested-by: default avatarKonrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      5710ab79
    • Michal Maloszewski's avatar
      iavf: Fix adopting new combined setting · 57d03f56
      Michal Maloszewski authored 
      In some cases overloaded flag IAVF_FLAG_REINIT_ITR_NEEDED
which should indicate that interrupts need to be completely
reinitialized during reset leads to RTNL deadlocks using ethtool -C
while a reset is in progress.
To fix, it was added a new flag IAVF_FLAG_REINIT_MSIX_NEEDED
used to trigger MSI-X reinit.
New combined setting is fixed adopt after VF reset.
This has been implemented by call reinit interrupt scheme
during VF reset.
Without this fix new combined setting has never been adopted.

Fixes: 209f2f9c ("iavf: Add support for VIRTCHNL_VF_OFFLOAD_VLAN_V2 negotiation")
Signed-off-by: default avatarGrzegorz Szczurek <grzegorzx.szczurek@intel.com>
Signed-off-by: default avatarMitch Williams <mitch.a.williams@intel.com>
Signed-off-by: default avatarMichal Maloszewski <michal.maloszewski@intel.com>
Tested-by: default avatarKonrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      57d03f56
    • Michal Maloszewski's avatar
      iavf: Fix handling of vlan strip virtual channel messages · 2cf29e55
      Michal Maloszewski authored 
      Modify netdev->features for vlan stripping based on virtual
channel messages received from the PF. Change is needed
to synchronize vlan strip status between PF sysfs and iavf ethtool.

Fixes: 5951a2b9 ("iavf: Fix VLAN feature flags after VFR")
Signed-off-by: default avatarNorbert Ciosek <norbertx.ciosek@intel.com>
Signed-off-by: default avatarMichal Maloszewski <michal.maloszewski@intel.com>
Tested-by: default avatarKonrad Jankowski <konrad0.jankowski@intel.com>
Signed-off-by: default avatarTony Nguyen <anthony.l.nguyen@intel.com>
      2cf29e55
    • Russell King (Oracle)'s avatar
      net: dsa: mt7530: fix incorrect test in mt753x_phylink_validate() · e5417cbf
      Russell King (Oracle) authored 
      Discussing one of the tests in mt753x_phylink_validate() with Landen
Chao confirms that the "||" should be "&&". Fix this.

Fixes: c288575f ("net: dsa: mt7530: Add the support of MT7531 switch")
Signed-off-by: default avatarRussell King (Oracle) <rmk+kernel@armlinux.org.uk>
Link: https://lore.kernel.org/r/E1nRCF0-00CiXD-7q@rmk-PC.armlinux.org.ukSigned-off-by: default avatarPaolo Abeni <pabeni@redhat.com>
      e5417cbf
  3. 07 Mar, 2022 6 commits
  5. 06 Mar, 2022 1 commit
  7. 05 Mar, 2022 2 commits
  9. 04 Mar, 2022 3 commits
    • Tung Nguyen's avatar
      tipc: fix kernel panic when enabling bearer · be4977b8
      Tung Nguyen authored 
      When enabling a bearer on a node, a kernel panic is observed:

[    4.498085] RIP: 0010:tipc_mon_prep+0x4e/0x130 [tipc]
...
[    4.520030] Call Trace:
[    4.520689]  <IRQ>
[    4.521236]  tipc_link_build_proto_msg+0x375/0x750 [tipc]
[    4.522654]  tipc_link_build_state_msg+0x48/0xc0 [tipc]
[    4.524034]  __tipc_node_link_up+0xd7/0x290 [tipc]
[    4.525292]  tipc_rcv+0x5da/0x730 [tipc]
[    4.526346]  ? __netif_receive_skb_core+0xb7/0xfc0
[    4.527601]  tipc_l2_rcv_msg+0x5e/0x90 [tipc]
[    4.528737]  __netif_receive_skb_list_core+0x20b/0x260
[    4.530068]  netif_receive_skb_list_internal+0x1bf/0x2e0
[    4.531450]  ? dev_gro_receive+0x4c2/0x680
[    4.532512]  napi_complete_done+0x6f/0x180
[    4.533570]  virtnet_poll+0x29c/0x42e [virtio_net]
...

The node in question is receiving activate messages in another
thread after changing bearer status to allow message sending/
receiving in current thread:

         thread 1           |              thread 2
         --------           |              --------
                            |
tipc_enable_bearer()        |
  test_and_set_bit_lock()   |
    tipc_bearer_xmit_skb()  |
                            | tipc_l2_rcv_msg()
                            |   tipc_rcv()
                            |     __tipc_node_link_up()
                            |       tipc_link_build_state_msg()
                            |         tipc_link_build_proto_msg()
                            |           tipc_mon_prep()
                            |           {
                            |             ...
                            |             // null-pointer dereference
                            |             u16 gen = mon->dom_gen;
                            |             ...
                            |           }
  // Not being executed yet |
  tipc_mon_create()         |
  {                         |
    ...                     |
    // allocate             |
    mon = kzalloc();        |
    ...                     |
  }                         |

Monitoring pointer in thread 2 is dereferenced before monitoring data
is allocated in thread 1. This causes kernel panic.

This commit fixes it by allocating the monitoring data before enabling
the bearer to receive messages.

Fixes: 35c55c98 ("tipc: add neighbor monitoring framework")
Reported-by: default avatarShuang Li <shuali@redhat.com>
Acked-by: default avatarJon Maloy <jmaloy@redhat.com>
Signed-off-by: default avatarTung Nguyen <tung.q.nguyen@dektech.com.au>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      be4977b8
    • Robert Hancock's avatar
      net: macb: Fix lost RX packet wakeup race in NAPI receive · 0bf476fc
      Robert Hancock authored 
      There is an oddity in the way the RSR register flags propagate to the
ISR register (and the actual interrupt output) on this hardware: it
appears that RSR register bits only result in ISR being asserted if the
interrupt was actually enabled at the time, so enabling interrupts with
RSR bits already set doesn't trigger an interrupt to be raised. There
was already a partial fix for this race in the macb_poll function where
it checked for RSR bits being set and re-triggered NAPI receive.
However, there was a still a race window between checking RSR and
actually enabling interrupts, where a lost wakeup could happen. It's
necessary to check again after enabling interrupts to see if RSR was set
just prior to the interrupt being enabled, and re-trigger receive in that
case.

This issue was noticed in a point-to-point UDP request-response protocol
which periodically saw timeouts or abnormally high response times due to
received packets not being processed in a timely fashion. In many
applications, more packets arriving, including TCP retransmissions, would
cause the original packet to be processed, thus masking the issue.

Fixes: 02f7a34f ("net: macb: Re-enable RX interrupt only when RX is done")
Cc: stable@vger.kernel.org
Co-developed-by: default avatarScott McNutt <scott.mcnutt@siriusxm.com>
Signed-off-by: default avatarScott McNutt <scott.mcnutt@siriusxm.com>
Signed-off-by: default avatarRobert Hancock <robert.hancock@calian.com>
Tested-by: default avatarClaudiu Beznea <claudiu.beznea@microchip.com>
Signed-off-by: default avatarDavid S. Miller <davem@davemloft.net>
      0bf476fc
    • Jakub Kicinski's avatar
      Merge tag 'for-net-2022-03-03' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth · 9f3956d6
      Jakub Kicinski authored 
      Luiz Augusto von Dentz says:

====================
bluetooth pull request for net:

 - Fix regression with processing of MGMT commands
 - Fix unbalanced unlock in Set Device Flags

* tag 'for-net-2022-03-03' of git://git.kernel.org/pub/scm/linux/kernel/git/bluetooth/bluetooth:
  Bluetooth: hci_sync: Fix not processing all entries on cmd_sync_work
  Bluetooth: hci_core: Fix unbalanced unlock in set_device_flags()
====================

Link: https://lore.kernel.org/r/20220303210743.314679-1-luiz.dentz@gmail.comSigned-off-by: default avatarJakub Kicinski <kuba@kernel.org>
      9f3956d6
  11. 03 Mar, 2022 21 commits
GitLab Nexedi Edition | About GitLab | About Nexedi | 沪ICP备2021021310号-2 | 沪ICP备2021021310号-7