Eran Ben Elisha [Mon, 31 Aug 2020 12:04:35 +0000 (15:04 +0300)]
net/mlx5: Add retry mechanism to the command entry index allocation
It is possible that new command entry index allocation will temporarily
fail. The new command holds the semaphore, so it means that a free entry
should be ready soon. Add one second retry mechanism before returning an
error.
Patch "net/mlx5: Avoid possible free of command entry while timeout comp
handler" increase the possibility to bump into this temporarily failure
as it delays the entry index release for non-callback commands.
Fixes: 7c120481afcc ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha <eranbe@nvidia.com> Reviewed-by: Moshe Shemesh <moshe@nvidia.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Eran Ben Elisha [Tue, 21 Jul 2020 07:25:52 +0000 (10:25 +0300)]
net/mlx5: poll cmd EQ in case of command timeout
Once driver detects a command interface command timeout, it warns the
user and returns timeout error to the caller. In such case, the entry of
the command is not evacuated (because only real event interrupt is allowed
to clear command interface entry). If the HW event interrupt
of this entry will never arrive, this entry will be left unused forever.
Command interface entries are limited and eventually we can end up without
the ability to post a new command.
In addition, if driver will not consume the EQE of the lost interrupt and
rearm the EQ, no new interrupts will arrive for other commands.
Add a resiliency mechanism for manually polling the command EQ in case of
a command timeout. In case resiliency mechanism will find non-handled EQE,
it will consume it, and the command interface will be fully functional
again. Once the resiliency flow finished, wait another 5 seconds for the
command interface to complete for this command entry.
Define mlx5_cmd_eq_recover() to manage the cmd EQ polling resiliency flow.
Add an async EQ spinlock to avoid races between resiliency flows and real
interrupts that might run simultaneously.
Fixes: 7c120481afcc ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Eran Ben Elisha [Tue, 4 Aug 2020 07:40:21 +0000 (10:40 +0300)]
net/mlx5: Avoid possible free of command entry while timeout comp handler
Upon command completion timeout, driver simulates a forced command
completion. In a rare case where real interrupt for that command arrives
simultaneously, it might release the command entry while the forced
handler might still access it.
Fix that by adding an entry refcount, to track current amount of allowed
handlers. Command entry to be released only when this refcount is
decremented to zero.
Command refcount is always initialized to one. For callback commands,
command completion handler is the symmetric flow to decrement it. For
non-callback commands, it is wait_func().
Before ringing the doorbell, increment the refcount for the real completion
handler. Once the real completion handler is called, it will decrement it.
For callback commands, once the delayed work is scheduled, increment the
refcount. Upon callback command completion handler, we will try to cancel
the timeout callback. In case of success, we need to decrement the callback
refcount as it will never run.
In addition, gather the entry index free and the entry free into a one
flow for all command types release.
Fixes: 7c120481afcc ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Reviewed-by: Moshe Shemesh <moshe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
Eran Ben Elisha [Thu, 13 Aug 2020 13:55:20 +0000 (16:55 +0300)]
net/mlx5: Fix a race when moving command interface to polling mode
As part of driver unload, it destroys the commands EQ (via FW command).
As the commands EQ is destroyed, FW will not generate EQEs for any command
that driver sends afterwards. Driver should poll for later commands status.
Driver commands mode metadata is updated before the commands EQ is
actually destroyed. This can lead for double completion handle by the
driver (polling and interrupt), if a command is executed and completed by
FW after the mode was changed, but before the EQ was destroyed.
Fix that by using the mlx5_cmd_allowed_opcode mechanism to guarantee
that only DESTROY_EQ command can be executed during this time period.
Fixes: 7c120481afcc ("mlx5: Add driver for Mellanox Connect-IB adapters") Signed-off-by: Eran Ben Elisha <eranbe@mellanox.com> Reviewed-by: Moshe Shemesh <moshe@mellanox.com> Signed-off-by: Saeed Mahameed <saeedm@nvidia.com>
====================
Fix bugs in Octeontx2 netdev driver
In existing Octeontx2 network drivers code has issues
like stale entries in broadcast replication list, missing
L3TYPE for IPv6 frames, running tx queues on error and
race condition in mbox reset.
This patch set fixes the above issues.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Mbox implementation in octeontx2 driver has three states
alloc, send and reset in mbox response. VF allocate and
sends message to PF for processing, PF ACKs them back and
reset the mbox memory. In some case we see synchronization
issue where after msgs_acked is incremented and before
mbox_reset API is called, if current execution is scheduled
out and a different thread is scheduled in which checks for
msgs_acked. Since the new thread sees msgs_acked == msgs_sent
it will try to allocate a new message and to send a new mbox
message to PF.Now if mbox_reset is scheduled in, PF will see
'0' in msgs_send.
This patch fixes the issue by calling mbox_reset before
incrementing msgs_acked flag for last processing message and
checks for valid message size.
Fixes: 0bedf57ff ("octeontx2-pf: Enable SRIOV and added VF mbox handling") Signed-off-by: Hariprasad Kelam <hkelam@marvell.com> Signed-off-by: Geetha sowjanya <gakula@marvell.com> Signed-off-by: Sunil Goutham <sgoutham@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Currently in otx2_open on failure of nix_lf_start
transmit queues are not stopped which are already
started in link_event. Since the tx queues are not
stopped network stack still try's to send the packets
leading to driver crash while access the device resources.
Fixes: d674dc650 ("octeontx2-pf: Register and handle link notifications") Signed-off-by: Hariprasad Kelam <hkelam@marvell.com> Signed-off-by: Geetha sowjanya <gakula@marvell.com> Signed-off-by: Sunil Goutham <sgoutham@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
octeontx2-pf: Fix TCP/UDP checksum offload for IPv6 frames
For TCP/UDP checksum offload feature in Octeontx2
expects L3TYPE to be set irrespective of IP header
checksum is being offloaded or not. Currently for
IPv6 frames L3TYPE is not being set resulting in
packet drop with checksum error. This patch fixes
this issue.
Fixes: 471374d48 ("octeontx2-pf: Add packet transmission support") Signed-off-by: Geetha sowjanya <gakula@marvell.com> Signed-off-by: Sunil Goutham <sgoutham@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
octeontx2-af: Fix enable/disable of default NPC entries
Packet replication feature present in Octeontx2
is a hardware linked list of PF and its VF
interfaces so that broadcast packets are sent
to all interfaces present in the list. It is
driver job to add and delete a PF/VF interface
to/from the list when the interface is brought
up and down. This patch fixes the
npc_enadis_default_entries function to handle
broadcast replication properly if packet replication
feature is present.
Fixes: 9a1fa34b33c7 ("octeontx2-af: Support to enable/disable default MCAM entries") Signed-off-by: Subbaraya Sundeep <sbhatta@marvell.com> Signed-off-by: Geetha sowjanya <gakula@marvell.com> Signed-off-by: Sunil Goutham <sgoutham@marvell.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Wed, 30 Sep 2020 22:01:09 +0000 (15:01 -0700)]
Merge branch '100GbE' of https://github.com/anguy11/net-queue
Tony Nguyen says:
====================
Intel Wired LAN Driver Updates 2020-09-30
This series contains updates to ice driver only.
Jake increases the wait time for firmware response as it can take longer
than the current wait time. Preserves the NVM capabilities of the device in
safe mode so the device reports its NVM update capabilities properly
when in this state.
v2: Added cover letter
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Jacob Keller [Wed, 2 Sep 2020 15:53:42 +0000 (08:53 -0700)]
ice: preserve NVM capabilities in safe mode
If the driver initializes in safe mode, it will call
ice_set_safe_mode_caps. This results in clearing the capabilities
structures, in order to set them up for operating in safe mode, ensuring
many features are disabled.
This has a side effect of also clearing the capability bits that relate
to NVM update. The result is that the device driver will not indicate
support for unified update, even if the firmware is capable.
Fix this by adding the relevant capability fields to the list of values
we preserve. To simplify the code, use a common_cap structure instead of
a handful of local variables. To reduce some duplication of the
capability name, introduce a couple of macros used to restore the
capabilities values from the cached copy.
Fixes: 001d1efdffe6 ("ice: Add support for unified NVM update flow capability") Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Tested-by: Brijesh Behera <brijeshx.behera@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Jacob Keller [Wed, 26 Aug 2020 18:30:34 +0000 (11:30 -0700)]
ice: increase maximum wait time for flash write commands
The ice driver needs to wait for a firmware response to each command to
write a block of data to the scratch area used to update the device
firmware. The driver currently waits for up to 1 second for this to be
returned.
It turns out that firmware might take longer than 1 second to return
a completion in some cases. If this happens, the flash update will fail
to complete.
Fix this by increasing the maximum time that the driver will wait for
both writing a block of data, and for activating the new NVM bank. The
timeout for an erase command is already several minutes, as the firmware
had to erase the entire bank which was already expected to take a minute
or more in the worst case.
In the case where firmware really won't respond, we will now take longer
to fail. However, this ensures that if the firmware is simply slow to
respond, the flash update can still complete. This new maximum timeout
should not adversely increase the update time, as the implementation for
wait_event_interruptible_timeout, and should wake very soon after we get
a completion event. It is better for a flash update be slow but still
succeed than to fail because we gave up too quickly.
Fixes: 4aa95dd4d31a ("ice: implement device flash update via devlink") Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Tested-by: Brijesh Behera <brijeshx.behera@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
David S. Miller [Wed, 30 Sep 2020 01:16:09 +0000 (18:16 -0700)]
Merge branch 'mptcp-Fix-for-32-bit-DATA_FIN'
Mat Martineau says:
====================
mptcp: Fix for 32-bit DATA_FIN
The main fix is contained in patch 2, and that commit message explains
the issue with not properly converting truncated DATA_FIN sequence
numbers sent by the peer.
With patch 2 adding an unlocked read of msk->ack_seq, patch 1 cleans up
access to that data with READ_ONCE/WRITE_ONCE.
This does introduce two merge conflicts with net-next, but both have
straightforward resolution. Patch 1 modifies a line that got removed in
net-next so the modification can be dropped when merging. Patch 2 will
require a trivial conflict resolution for a modified function
declaration.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Mat Martineau [Tue, 29 Sep 2020 22:08:20 +0000 (15:08 -0700)]
mptcp: Handle incoming 32-bit DATA_FIN values
The peer may send a DATA_FIN mapping with either a 32-bit or 64-bit
sequence number. When a 32-bit sequence number is received for the
DATA_FIN, it must be expanded to 64 bits before comparing it to the
last acked sequence number. This expansion was missing.
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/93 Fixes: 8eefb7836128 ("mptcp: Track received DATA_FIN sequence number and add related helpers") Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
====================
via-rhine: Resume fix and other maintenance work
I use via-rhine based Ethernet regularly, and the Ethernet dying
after resume was really annoying me. I decided to take the
matter into my own hands, and came up with a fix for the Ethernet
disappearing after resume. I will also want to take over the code
maintenance work for via-rhine. The patches apply to the latest
code, but they should be backported to older kernels as well.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Kevin Brace [Tue, 29 Sep 2020 20:09:41 +0000 (13:09 -0700)]
via-rhine: VTunknown1 device is really VT8251 South Bridge
The VIA Technologies VT8251 South Bridge's integrated Rhine-II
Ethernet MAC comes has a PCI revision value of 0x7c. This was
verified on ASUS P5V800-VM mainboard.
Signed-off-by: Kevin Brace <kevinbrace@bracecomputerlab.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Kevin Brace [Tue, 29 Sep 2020 20:09:40 +0000 (13:09 -0700)]
via-rhine: Fix for the hardware having a reset failure after resume
In rhine_resume() and rhine_suspend(), the code calls netif_running()
to see if the network interface is down or not. If it is down (i.e.,
netif_running() returning false), they will skip any housekeeping work
within the function relating to the hardware. This becomes a problem
when the hardware resumes from a standby since it is counting on
rhine_resume() to map its MMIO and power up rest of the hardware.
Not getting its MMIO remapped and rest of the hardware powered
up lead to a soft reset failure and hardware disappearance. The
solution is to map its MMIO and power up rest of the hardware inside
rhine_open() before soft reset is to be performed. This solution was
verified on ASUS P5V800-VM mainboard's integrated Rhine-II Ethernet
MAC inside VIA Technologies VT8251 South Bridge.
Signed-off-by: Kevin Brace <kevinbrace@bracecomputerlab.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Tony Nguyen [Tue, 29 Sep 2020 21:06:18 +0000 (14:06 -0700)]
MAINTAINERS: Update MAINTAINERS for Intel ethernet drivers
Add Jesse Brandeburg and myself; remove Jeff Kirsher.
CC: Jeff Kirsher <jeffrey.t.kirsher@intel.com> CC: Jesse Brandeburg <jesse.brandeburg@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Vladimir Oltean [Tue, 29 Sep 2020 11:20:24 +0000 (14:20 +0300)]
net: dsa: felix: fix incorrect action offsets for VCAP IS2
The port mask width was larger than the actual number of ports, and
therefore, all fields following this one were also shifted by the number
of excess bits. But the driver doesn't use the REW_OP, SMAC_REPLACE_ENA
or ACL_ID bits from the action vector, so the bug was inconsequential.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
There are two chip pins named TXDLY and RXDLY which actually adds the 2ns
delays to TXC and RXC for TXD/RXD latching. These two pins can config via
4.7k-ohm resistor to 3.3V hw setting, but also config via software setting
(extension page 0xa4 register 0x1c bit13 12 and 11).
This table describes how to config these hw pins via external pull-high or pull-
low resistor.
It is a misunderstanding that mapping it as register bits below:
8:6 = PHY Address
5:4 = Auto-Negotiation
3 = Interface Mode Select
2 = RX Delay
1 = TX Delay
0 = SELRGV
So I removed these descriptions above and add related settings as below:
14 = reserved
13 = force Tx RX Delay controlled by bit12 bit11
12 = Tx Delay
11 = Rx Delay
10:0 = Test && debug settings reserved by realtek
Test && debug settings are not recommend to modify by default.
Fixes: 2b3a658d7be8 ("net: phy: realtek: Add rtl8211e rx/tx delays config") Signed-off-by: Willy Liu <willy.liu@realtek.com> Signed-off-by: David S. Miller <davem@davemloft.net>
virtio-net: don't disable guest csum when disable LRO
Open vSwitch and Linux bridge will disable LRO of the interface
when this interface added to them. Now when disable the LRO, the
virtio-net csum is disable too. That drops the forwarding performance.
Fixes: b15706d13e90 ("virtio-net: ethtool configurable LRO") Cc: Michael S. Tsirkin <mst@redhat.com> Cc: Jason Wang <jasowang@redhat.com> Cc: Willem de Bruijn <willemb@google.com> Signed-off-by: Tonghao Zhang <xiangxia.m.yue@gmail.com> Acked-by: Willem de Bruijn <willemb@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
He Zhe [Mon, 28 Sep 2020 09:00:23 +0000 (17:00 +0800)]
bpf, powerpc: Fix misuse of fallthrough in bpf_jit_comp()
The user defined label following "fallthrough" is not considered by GCC
and causes build failure.
kernel-source/include/linux/compiler_attributes.h:208:41: error: attribute
'fallthrough' not preceding a case label or default label [-Werror]
208 define fallthrough _attribute((fallthrough_))
^~~~~~~~~~~~~
Fixes: 79b7a0dd093d ("treewide: Use fallthrough pseudo-keyword") Signed-off-by: He Zhe <zhe.he@windriver.com> Signed-off-by: Daniel Borkmann <daniel@iogearbox.net> Cc: Gustavo A. R. Silva <gustavoars@kernel.org> Link: https://lore.kernel.org/bpf/20200928090023.38117-1-zhe.he@windriver.com
Jakub Kicinski [Tue, 29 Sep 2020 00:58:41 +0000 (17:58 -0700)]
ethtool: mark netlink family as __ro_after_init
Like all genl families ethtool_genl_family needs to not
be a straight up constant, because it's modified/initialized
by genl_register_family(). After init, however, it's only
passed to genlmsg_put() & co. therefore we can mark it
as __ro_after_init.
Since genl_family structure contains function pointers
mark this as a fix.
Fixes: 7dccc0028581 ("ethtool: introduce ethtool netlink interface") Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Jakub Kicinski [Tue, 29 Sep 2020 00:53:29 +0000 (17:53 -0700)]
genetlink: add missing kdoc for validation flags
Validation flags are missing kdoc, add it.
Fixes: 604b526aac84 ("genetlink: optionally validate strictly/dumps") Signed-off-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
net: usb: ax88179_178a: fix missing stop entry in driver_info
Adds the missing .stop entry in the Belkin driver_info structure.
Fixes: c877e151fbb5 ("net: usb: asix88179_178a: Add support for the Belkin B2B128") Signed-off-by: Wilken Gottwalt <wilken.gottwalt@mailbox.org> Signed-off-by: David S. Miller <davem@davemloft.net>
net: qrtr: ns: Protect radix_tree_deref_slot() using rcu read locks
The rcu read locks are needed to avoid potential race condition while
dereferencing radix tree from multiple threads. The issue was identified
by syzbot. Below is the crash report:
Fixes: c2942963826d ("net: qrtr: Migrate nameservice to kernel from userspace") Reported-and-tested-by: syzbot+0f84f6eed90503da72fc@syzkaller.appspotmail.com Signed-off-by: Manivannan Sadhasivam <manivannan.sadhasivam@linaro.org> Signed-off-by: David S. Miller <davem@davemloft.net>
====================
net: core: fix a lockdep splat in the dev_addr_list.
This patchset is to avoid lockdep splat.
When a stacked interface graph is changed, netif_addr_lock() is called
recursively and it internally calls spin_lock_nested().
The parameter of spin_lock_nested() is 'dev->lower_level',
this is called subclass.
The problem of 'dev->lower_level' is that while 'dev->lower_level' is
being used as a subclass of spin_lock_nested(), its value can be changed.
So, spin_lock_nested() would be called recursively with the same
subclass value, the lockdep understands a deadlock.
In order to avoid this, a new variable is needed and it is going to be
used as a parameter of spin_lock_nested().
The first and second patch is a preparation patch for the third patch.
In the third patch, the problem will be fixed.
The first patch is to add __netdev_upper_dev_unlink().
An existed netdev_upper_dev_unlink() is renamed to
__netdev_upper_dev_unlink(). and netdev_upper_dev_unlink()
is added as an wrapper of this function.
The second patch is to add the netdev_nested_priv structure.
netdev_walk_all_{ upper | lower }_dev() pass both private functions
and "data" pointer to handle their own things.
At this point, the data pointer type is void *.
In order to make it easier to expand common variables and functions,
this new netdev_nested_priv structure is added.
The third patch is to add a new variable 'nested_level'
into the net_device structure.
This variable will be used as a parameter of spin_lock_nested() of
dev->addr_list_lock.
Due to this variable, it can avoid lockdep splat.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
net: core: add nested_level variable in net_device
This patch is to add a new variable 'nested_level' into the net_device
structure.
This variable will be used as a parameter of spin_lock_nested() of
dev->addr_list_lock.
netif_addr_lock() can be called recursively so spin_lock_nested() is
used instead of spin_lock() and dev->lower_level is used as a parameter
of spin_lock_nested().
But, dev->lower_level value can be updated while it is being used.
So, lockdep would warn a possible deadlock scenario.
When a stacked interface is deleted, netif_{uc | mc}_sync() is
called recursively.
So, spin_lock_nested() is called recursively too.
At this moment, the dev->lower_level variable is used as a parameter of it.
dev->lower_level value is updated when interfaces are being unlinked/linked
immediately.
Thus, After unlinking, dev->lower_level shouldn't be a parameter of
spin_lock_nested().
A (macvlan)
|
B (vlan)
|
C (bridge)
|
D (macvlan)
|
E (vlan)
|
F (bridge)
When an interface 'A' is removed, it releases resources.
At this moment, netif_addr_lock() would be called.
Then, netdev_upper_dev_unlink() is called recursively.
Then dev->lower_level is updated.
There is no problem.
But, when the bridge module is removed, 'C' and 'F' interfaces
are removed at once.
If 'F' is removed first, a lower_level value is like below.
A->lower_level : 5
B->lower_level : 4
C->lower_level : 3
D->lower_level : 2
E->lower_level : 1
F->lower_level : 1
Then, 'C' is removed. at this moment, netif_addr_lock() is called
recursively.
The ordering is like this.
C(3)->D(2)->E(1)->F(1)
At this moment, the lower_level value of 'E' and 'F' are the same.
So, lockdep warns a possible deadlock scenario.
In order to avoid this problem, a new variable 'nested_level' is added.
This value is the same as dev->lower_level - 1.
But this value is updated in rtnl_unlock().
So, this variable can be used as a parameter of spin_lock_nested() safely
in the rtnl context.
Test commands:
ip link add br0 type bridge vlan_filtering 1
ip link add vlan1 link br0 type vlan id 10
ip link add macvlan2 link vlan1 type macvlan
ip link add br3 type bridge vlan_filtering 1
ip link set macvlan2 master br3
ip link add vlan4 link br3 type vlan id 10
ip link add macvlan5 link vlan4 type macvlan
ip link add br6 type bridge vlan_filtering 1
ip link set macvlan5 master br6
ip link add vlan7 link br6 type vlan id 10
ip link add macvlan8 link vlan7 type macvlan
ip link set br0 up
ip link set vlan1 up
ip link set macvlan2 up
ip link set br3 up
ip link set vlan4 up
ip link set macvlan5 up
ip link set br6 up
ip link set vlan7 up
ip link set macvlan8 up
modprobe -rv bridge
Fixes: f31ab2b7ac1a ("net: change addr_list_lock back to static key") Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net: core: introduce struct netdev_nested_priv for nested interface infrastructure
Functions related to nested interface infrastructure such as
netdev_walk_all_{ upper | lower }_dev() pass both private functions
and "data" pointer to handle their own things.
At this point, the data pointer type is void *.
In order to make it easier to expand common variables and functions,
this new netdev_nested_priv structure is added.
In the following patch, a new member variable will be added into this
struct to fix the lockdep issue.
Signed-off-by: Taehee Yoo <ap420073@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Cong Wang [Mon, 28 Sep 2020 18:31:03 +0000 (11:31 -0700)]
net_sched: remove a redundant goto chain check
All TC actions call tcf_action_check_ctrlact() to validate
goto chain, so this check in tcf_action_init_1() is actually
redundant. Remove it to save troubles of leaking memory.
Fixes: ee6863fdc727 ("net_sched: defer tcf_idr_insert() in tcf_action_init_1()") Reported-by: Vlad Buslov <vladbu@mellanox.com> Suggested-by: Davide Caratti <dcaratti@redhat.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Reviewed-by: Davide Caratti <dcaratti@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
When a user-space software manages fdb entries externally it should
set the ext_learn flag which marks the fdb entry as externally managed
and avoids expiring it (they're treated as static fdbs). Unfortunately
on events where fdb entries are flushed (STP down, netlink fdb flush
etc) these fdbs are also deleted automatically by the bridge. That in turn
causes trouble for the managing user-space software (e.g. in MLAG setups
we lose remote fdb entries on port flaps).
These entries are completely externally managed so we should avoid
automatically deleting them, the only exception are offloaded entries
(i.e. BR_FDB_ADDED_BY_EXT_LEARN + BR_FDB_OFFLOADED). They are flushed as
before.
Fixes: 1bba299424d1 ("net: bridge: allow to add externally learned entries from user-space") Signed-off-by: Nikolay Aleksandrov <nikolay@nvidia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
1) Fix a build warning in ip_vti if CONFIG_IPV6 is not set.
From YueHaibing.
2) Restore IPCB on espintcp before handing the packet to xfrm
as the information there is still needed.
From Sabrina Dubroca.
3) Fix pmtu updating for xfrm interfaces.
From Sabrina Dubroca.
4) Some xfrm state information was not cloned with xfrm_do_migrate.
Fixes to clone the full xfrm state, from Antony Antony.
5) Use the correct address family in xfrm_state_find. The struct
flowi must always be interpreted along with the original
address family. This got lost over the years.
Fix from Herbert Xu.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Mistakenly bit 2 was set instead of bit 3 as in the vendor driver.
Fixes: e39b7735e9e8 ("r8169: sync PCIe PHY init with vendor driver 8.047.01") Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
dt-bindings: net: renesas,ravb: Add support for r8a774e1 SoC
Document RZ/G2H (R8A774E1) SoC bindings.
Signed-off-by: Marian-Cristian Rotariu <marian-cristian.rotariu.rb@bp.renesas.com> Signed-off-by: Lad Prabhakar <prabhakar.mahadev-lad.rj@bp.renesas.com> Reviewed-by: Sergei Shtylyov <sergei.shtylyov@gmail.com> Reviewed-by: Geert Uytterhoeven <geert+renesas@glider.be> Acked-by: Rob Herring <robh@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
If mlxsw_sp_acl_tcam_group_id_get() fails, the mutex initialized earlier
is not destroyed.
Fix this by initializing the mutex after calling the function. This is
symmetric to mlxsw_sp_acl_tcam_group_del().
Fixes: 93a9f0edc4f7 ("mlxsw: spectrum_acl: Introduce a mutex to guard region list updates") Signed-off-by: Ido Schimmel <idosch@nvidia.com> Reviewed-by: Jiri Pirko <jiri@nvidia.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Randy Dunlap [Sun, 27 Sep 2020 04:33:43 +0000 (21:33 -0700)]
mdio: fix mdio-thunder.c dependency & build error
Fix build error by selecting MDIO_DEVRES for MDIO_THUNDER.
Fixes this build error:
ld: drivers/net/phy/mdio-thunder.o: in function `thunder_mdiobus_pci_probe':
drivers/net/phy/mdio-thunder.c:78: undefined reference to `devm_mdiobus_alloc_size'
Fixes: 22908fdf6032 ("phy: mdio-thunder: Add driver for Cavium Thunder SoC MDIO buses.") Reported-by: kernel test robot <lkp@intel.com> Signed-off-by: Randy Dunlap <rdunlap@infradead.org> Cc: Bartosz Golaszewski <bgolaszewski@baylibre.com> Cc: Andrew Lunn <andrew@lunn.ch> Cc: Heiner Kallweit <hkallweit1@gmail.com> Cc: netdev@vger.kernel.org Cc: David Daney <david.daney@cavium.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Signed-off-by: Florian Fainelli <f.fainelli@gmail.com> Acked-by: Vladimir Oltean <olteanv@gmail.com> Acked-by: Andrew Lunn <andrew@lunn.ch> Acked-by: Vivien Didelot <vivien.didelot@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
When adding the support for TBF offload, the improper command version
was added even though the command format is for the V2 of
dpni_set_tx_shaping(). This does not affect the functionality of TBF
since the only change between these two versions is the addition of the
exceeded parameters which are not used in TBF. Still, fix the bug so
that we keep things in sync.
Fixes: 8e692895a861 ("dpaa2-eth: add API for Tx shaping") Signed-off-by: Ioana Ciornei <ioana.ciornei@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Fri, 25 Sep 2020 13:38:08 +0000 (06:38 -0700)]
team: set dev->needed_headroom in team_setup_by_port()
Some devices set needed_headroom. If we ignore it, we might
end up crashing in various skb_push() for example in ipgre_header()
since some layers assume enough headroom has been reserved.
Fixes: 091798117add ("team: add support for non-ethernet devices") Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Eric Dumazet [Fri, 25 Sep 2020 13:38:07 +0000 (06:38 -0700)]
bonding: set dev->needed_headroom in bond_setup_by_slave()
syzbot managed to crash a host by creating a bond
with a GRE device.
For non Ethernet device, bonding calls bond_setup_by_slave()
instead of ether_setup(), and unfortunately dev->needed_headroom
was not copied from the new added member.
Fixes: 5c2e2f1c44b1 ("net: Allow netdevices to specify needed head/tailroom") Signed-off-by: Eric Dumazet <edumazet@google.com> Reported-by: syzbot <syzkaller@googlegroups.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Ivan Khoronzhuk [Fri, 25 Sep 2020 12:44:39 +0000 (15:44 +0300)]
net: ethernet: cavium: octeon_mgmt: use phy_start and phy_stop
To start also "phy state machine", with UP state as it should be,
the phy_start() has to be used, in another case machine even is not
triggered. After this change negotiation is supposed to be triggered
by SM workqueue.
It's not correct usage, but it appears after the following patch,
so add it as a fix.
Fixes: a7ffa9e169cb ("net: phy: add phy_check_link_status") Signed-off-by: Ivan Khoronzhuk <ikhoronz@cisco.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Wong Vee Khee [Fri, 25 Sep 2020 09:54:06 +0000 (17:54 +0800)]
net: stmmac: Fix clock handling on remove path
While unloading the dwmac-intel driver, clk_disable_unprepare() is
being called twice in stmmac_dvr_remove() and
intel_eth_pci_remove(). This causes kernel panic on the second call.
Removing the second call of clk_disable_unprepare() in
intel_eth_pci_remove().
Fixes: 49a4d81c0927 ("stmmac: intel: Fix clock handling on error and remove paths") Cc: Andy Shevchenko <andriy.shevchenko@linux.intel.com> Reviewed-by: Voon Weifeng <weifeng.voon@intel.com> Signed-off-by: Wong Vee Khee <vee.khee.wong@intel.com> Reviewed-by: Andy Shevchenko <andy.shevchenko@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
vmxnet3: fix cksum offload issues for non-udp tunnels
Commit bcef8f5256d6 ("vmxnet3: add geneve and vxlan tunnel offload
support") added support for encapsulation offload. However, the inner
offload capability is to be restrictued to UDP tunnels.
This patch fixes the issue for non-udp tunnels by adding features
check capability and filtering appropriate features for non-udp tunnels.
Fixes: bcef8f5256d6 ("vmxnet3: add geneve and vxlan tunnel offload support") Signed-off-by: Ronak Doshi <doshir@vmware.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Jacob Keller [Wed, 2 Sep 2020 15:53:47 +0000 (08:53 -0700)]
ice: fix memory leak in ice_vsi_setup
During ice_vsi_setup, if ice_cfg_vsi_lan fails, it does not properly
release memory associated with the VSI rings. If we had used devres
allocations for the rings, this would be ok. However, we use kzalloc and
kfree_rcu for these ring structures.
Using the correct label to cleanup the rings during ice_vsi_setup
highlights an issue in the ice_vsi_clear_rings function: it can leave
behind stale ring pointers in the q_vectors structure.
When releasing rings, we must also ensure that no q_vector associated
with the VSI will point to this ring again. To resolve this, loop over
all q_vectors and release their ring mapping. Because we are about to
free all rings, no q_vector should remain pointing to any of the rings
in this VSI.
Fixes: 63e81ecf4287 ("ice: Update Tx scheduler tree for VSI multi-Tx queue support") Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Jacob Keller [Wed, 2 Sep 2020 15:53:46 +0000 (08:53 -0700)]
ice: fix memory leak if register_netdev_fails
The ice_setup_pf_sw function can cause a memory leak if register_netdev
fails, due to accidentally failing to free the VSI rings. Fix the memory
leak by using ice_vsi_release, ensuring we actually go through the full
teardown process.
This should be safe even if the netdevice is not registered because we
will have set the netdev pointer to NULL, ensuring ice_vsi_release won't
call unregister_netdev.
An alternative fix would be moving management of the PF VSI netdev into
the main VSI setup code. This is complicated and likely requires
significant refactor in how we manage VSIs
Fixes: 8ed3dc6d887c ("ice: Add support for VSI allocation and deallocation") Signed-off-by: Jacob Keller <jacob.e.keller@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
It appears that the ice_suspend flow is missing a call to pci_save_state
and this is triggering the message "State of device not saved by
ice_suspend" and a call trace. Fix it.
Fixes: 2aafe62a1f22 ("ice: Add advanced power mgmt for WoL") Signed-off-by: Anirudh Venkataramanan <anirudh.venkataramanan@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
When calling iavf_resume there was a crash because wrong
function was used to get iavf_adapter and net_device pointers.
Changed how iavf_resume is getting iavf_adapter and net_device
pointers from pci_dev.
Fixes: 16d6e2c362ef ("i40evf: main driver core") Signed-off-by: Sylwester Dziedziuch <sylwesterx.dziedziuch@intel.com> Reviewed-by: Aleksandr Loktionov <aleksandr.loktionov@intel.com> Tested-by: Aaron Brown <aaron.f.brown@intel.com> Signed-off-by: Tony Nguyen <anthony.l.nguyen@intel.com>
Herbert Xu [Fri, 25 Sep 2020 04:42:56 +0000 (14:42 +1000)]
xfrm: Use correct address family in xfrm_state_find
The struct flowi must never be interpreted by itself as its size
depends on the address family. Therefore it must always be grouped
with its original family value.
In this particular instance, the original family value is lost in
the function xfrm_state_find. Therefore we get a bogus read when
it's coupled with the wrong family which would occur with inter-
family xfrm states.
This patch fixes it by keeping the original family value.
Note that the same bug could potentially occur in LSM through
the xfrm_state_pol_flow_match hook. I checked the current code
there and it seems to be safe for now as only secid is used which
is part of struct flowi_common. But that API should be changed
so that so that we don't get new bugs in the future. We could
do that by replacing fl with just secid or adding a family field.
Reported-by: syzbot+577fbac3145a6eb2e7a5@syzkaller.appspotmail.com Fixes: 9a07d675cb46 ("[XFRM]: State selection update to use inner...") Signed-off-by: Herbert Xu <herbert@gondor.apana.org.au> Signed-off-by: Steffen Klassert <steffen.klassert@secunet.com>
Currently, we use length of DSACKed range to compute number of
delivered packets. And if sequence range in DSACK is corrupted,
we can get bogus dsacked/acked count, and bogus cwnd.
This patch put bounds on DSACKed range to skip update of data
delivery and spurious retransmission information, if the DSACK
is unlikely caused by sender's action:
- DSACKed range shouldn't be greater than maximum advertised rwnd.
- Total no. of DSACKed segments shouldn't be greater than total
no. of retransmitted segs. Unlike spurious retransmits, network
duplicates or corrupted DSACKs shouldn't be counted as delivery.
Signed-off-by: Priyaranjan Jha <priyarjha@google.com> Signed-off-by: Neal Cardwell <ncardwell@google.com> Signed-off-by: Yuchung Cheng <ycheng@google.com> Signed-off-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
MDIO reads can happen during PHY probing, and printing an error with
dev_err can result in a large number of error messages during device
probe. On a platform with a serial console this can result in
excessively long boot times in a way that looks like an infinite loop
when multiple busses are present. Since 72d2c91935f (net/fsl: enable
extended scanning in xgmac_mdio) we perform more scanning so there are
potentially more failures.
Reduce the logging level to dev_dbg which is consistent with the
Freescale enetc driver.
Cc: Jeremy Linton <jeremy.linton@arm.com> Signed-off-by: Jamie Iles <jamie@nuviainc.com> Reviewed-by: Andrew Lunn <andrew@lunn.ch> Signed-off-by: David S. Miller <davem@davemloft.net>
Helmut Grohne [Thu, 24 Sep 2020 08:37:47 +0000 (10:37 +0200)]
net: dsa: microchip: really look for phy-mode in port nodes
The previous implementation failed to account for the "ports" node. The
actual port nodes are not child nodes of the switch node, but a "ports"
node sits in between.
Fixes: 6e7573eb48a2 ("net: dsa: microchip: look for phy-mode in port nodes") Signed-off-by: Helmut Grohne <helmut.grohne@intenta.de> Signed-off-by: David S. Miller <davem@davemloft.net>
tls_sw_release_resources_tx() waits for encrypt_pending, which
can have race, so we need similar changes as in commit 8284b6432e4124ca82b82a60c405171f5c8881d6 here as well.
Fixes: 4e09428b5ad7 ("net/tls: Add support for async encryption of records for performance") Signed-off-by: Rohit Maheshwari <rohitm@chelsio.com> Acked-by: Jakub Kicinski <kuba@kernel.org> Signed-off-by: David S. Miller <davem@davemloft.net>
Xiaoliang Yang [Thu, 24 Sep 2020 02:11:13 +0000 (10:11 +0800)]
net: mscc: ocelot: fix fields offset in SG_CONFIG_REG_3
INIT_IPS and GATE_ENABLE fields have a wrong offset in SG_CONFIG_REG_3.
This register is used by stream gate control of PSFP, and it has not
been used before, because PSFP is not implemented in ocelot driver.
Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xiaoliang Yang [Thu, 24 Sep 2020 01:57:46 +0000 (09:57 +0800)]
net: dsa: felix: convert TAS link speed based on phylink speed
state->speed holds a value of 10, 100, 1000 or 2500, but
QSYS_TAG_CONFIG_LINK_SPEED expects a value of 0, 1, 2, 3. So convert the
speed to a proper value.
Fixes: 66fc7ae56755 ("net: dsa: felix: Configure Time-Aware Scheduler via taprio offload") Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com> Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Luo bin [Thu, 24 Sep 2020 01:31:51 +0000 (09:31 +0800)]
hinic: fix wrong return value of mac-set cmd
It should also be regarded as an error when hw return status=4 for PF's
setting mac cmd. Only if PF return status=4 to VF should this cmd be
taken special treatment.
Fixes: fb7d7b5fb3de ("hinic: add sriov feature support") Signed-off-by: Luo bin <luobin9@huawei.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xie He [Wed, 23 Sep 2020 18:18:18 +0000 (11:18 -0700)]
drivers/net/wan/x25_asy: Correct the ndo_open and ndo_stop functions
1.
Move the lapb_register/lapb_unregister calls into the ndo_open/ndo_stop
functions.
This makes the LAPB protocol start/stop when the network interface
starts/stops. When the network interface is down, the LAPB protocol
shouldn't be running and the LAPB module shoudn't be generating control
frames.
2.
Move netif_start_queue/netif_stop_queue into the ndo_open/ndo_stop
functions.
This makes the TX queue start/stop when the network interface
starts/stops.
(netif_stop_queue was originally in the ndo_stop function. But to make
the code look better, I created a new function to use as ndo_stop, and
made it call the original ndo_stop function. I moved netif_stop_queue
from the original ndo_stop function to the new ndo_stop function.)
Cc: Martin Schiller <ms@dev.tdt.de> Signed-off-by: Xie He <xie.he.0141@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net/ipv4: always honour route mtu during forwarding
Documentation/networking/ip-sysctl.txt:46 says:
ip_forward_use_pmtu - BOOLEAN
By default we don't trust protocol path MTUs while forwarding
because they could be easily forged and can lead to unwanted
fragmentation by the router.
You only need to enable this if you have user-space software
which tries to discover path mtus by itself and depends on the
kernel honoring this information. This is normally not the case.
Default: 0 (disabled)
Possible values:
0 - disabled
1 - enabled
Which makes it pretty clear that setting it to 1 is a potential
security/safety/DoS issue, and yet it is entirely reasonable to want
forwarded traffic to honour explicitly administrator configured
route mtus (instead of defaulting to device mtu).
Indeed, I can't think of a single reason why you wouldn't want to.
Since you configured a route mtu you probably know better...
It is pretty common to have a higher device mtu to allow receiving
large (jumbo) frames, while having some routes via that interface
(potentially including the default route to the internet) specify
a lower mtu.
Note that ipv6 forwarding uses device mtu unless the route is locked
(in which case it will use the route mtu).
This approach is not usable for IPv4 where an 'mtu lock' on a route
also has the side effect of disabling TCP path mtu discovery via
disabling the IPv4 DF (don't frag) bit on all outgoing frames.
I'm not aware of a way to lock a route from an IPv6 RA, so that also
potentially seems wrong.
Signed-off-by: Maciej Żenczykowski <maze@google.com> Cc: Eric Dumazet <edumazet@google.com> Cc: Willem de Bruijn <willemb@google.com> Cc: Lorenzo Colitti <lorenzo@google.com> Cc: Sunmeet Gill (Sunny) <sgill@quicinc.com> Cc: Vinay Paradkar <vparadka@qti.qualcomm.com> Cc: Tyler Wear <twear@quicinc.com> Cc: David Ahern <dsahern@kernel.org> Reviewed-by: Eric Dumazet <edumazet@google.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Cong Wang [Wed, 23 Sep 2020 03:56:24 +0000 (20:56 -0700)]
net_sched: commit action insertions together
syzbot is able to trigger a failure case inside the loop in
tcf_action_init(), and when this happens we clean up with
tcf_action_destroy(). But, as these actions are already inserted
into the global IDR, other parallel process could free them
before tcf_action_destroy(), then we will trigger a use-after-free.
Fix this by deferring the insertions even later, after the loop,
and committing all the insertions in a separate loop, so we will
never fail in the middle of the insertions any more.
One side effect is that the window between alloction and final
insertion becomes larger, now it is more likely that the loop in
tcf_del_walker() sees the placeholder -EBUSY pointer. So we have
to check for error pointer in tcf_del_walker().
Reported-and-tested-by: syzbot+2287853d392e4b42374a@syzkaller.appspotmail.com Fixes: 514093d812c1 ("net: sched: atomically check-allocate action") Cc: Vlad Buslov <vladbu@mellanox.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Cong Wang [Wed, 23 Sep 2020 03:56:23 +0000 (20:56 -0700)]
net_sched: defer tcf_idr_insert() in tcf_action_init_1()
All TC actions call tcf_idr_insert() for new action at the end
of their ->init(), so we can actually move it to a central place
in tcf_action_init_1().
And once the action is inserted into the global IDR, other parallel
process could free it immediately as its refcnt is still 1, so we can
not fail after this, we need to move it after the goto action
validation to avoid handling the failure case after insertion.
This is found during code review, is not directly triggered by syzbot.
And this prepares for the next patch.
Cc: Vlad Buslov <vladbu@mellanox.com> Cc: Jamal Hadi Salim <jhs@mojatatu.com> Cc: Jiri Pirko <jiri@resnulli.us> Signed-off-by: Cong Wang <xiyou.wangcong@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
libbpf: Fix XDP program load regression for old kernels
Fix regression in libbpf, introduced by XDP link change, which causes XDP
programs to fail to be loaded into kernel due to specified BPF_XDP
expected_attach_type. While kernel doesn't enforce expected_attach_type for
BPF_PROG_TYPE_XDP, some old kernels already support XDP program, but they
don't yet recognize expected_attach_type field in bpf_attr, so setting it to
non-zero value causes program load to fail.
Luckily, libbpf already has a mechanism to deal with such cases, so just make
expected_attach_type optional for XDP programs.
Felix Fietkau [Wed, 23 Sep 2020 05:24:42 +0000 (07:24 +0200)]
mt76: mt7615: reduce maximum VHT MPDU length to 7991
After fixing mac80211 to allow larger A-MSDUs in some cases, there have been
reports of performance regressions and packet loss with some clients.
It appears that the issue occurs when the hardware is transmitting A-MSDUs
bigger than 8k. Limit the local VHT MPDU size capability to 7991, matching
the value used for MT7915 as well.
net: stmmac: removed enabling eee in EEE set callback
EEE should be only be enabled during stmmac_mac_link_up() when the
link are up and being set up properly. set_eee should only do settings
configuration and disabling the eee.
Without this fix, turning on EEE using ethtool will return
"Operation not supported". This is due to the driver is in a dead loop
waiting for eee to be advertised in the for eee to be activated but the
driver will only configure the EEE advertisement after the eee is
activated.
Ethtool should only return "Operation not supported" if there is no EEE
capbility in the MAC controller.
Fixes: c500a60e6d13 ("net: stmmac: Fix a race in EEE enable callback") Signed-off-by: Voon Weifeng <weifeng.voon@intel.com> Acked-by: Mark Gross <mgross@linux.intel.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The TX DMA channel data is accessed by the xrx200_start_xmit() and the
xrx200_tx_housekeeping() function from different threads. Make sure the
accesses are synchronized by acquiring the netif_tx_lock() in the
xrx200_tx_housekeeping() function too. This lock is acquired by the
kernel before calling xrx200_start_xmit().
Signed-off-by: Hauke Mehrtens <hauke@hauke-m.de> Signed-off-by: David S. Miller <davem@davemloft.net>
Tian Tao [Tue, 22 Sep 2020 13:32:19 +0000 (21:32 +0800)]
net: switchdev: Fixed kerneldoc warning
Update kernel-doc line comments to fix warnings reported by make W=1.
net/switchdev/switchdev.c:413: warning: Function parameter or
member 'extack' not described in 'call_switchdev_notifiers'
Signed-off-by: Tian Tao <tiantao6@hisilicon.com> Acked-by: Ivan Vecera <ivecera@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
This commit moved the ravb_mdio_init() call (and thus the
of_mdiobus_register() call) from the ravb_probe() to the ravb_open()
call. This causes a regression during system resume (s2idle/s2ram), as
new PHY devices cannot be bound while suspended.
During boot, the Micrel PHY is detected like this:
Micrel KSZ9031 Gigabit PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Micrel KSZ9031 Gigabit PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=228)
ravb e6800000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off
During system suspend, (A) defer_all_probes is set to true, and (B)
usermodehelper_disabled is set to UMH_DISABLED, to avoid drivers being
probed while suspended.
A. If CONFIG_MODULES=n, phy_device_register() calling device_add()
merely adds the device, but does not probe it yet, as
really_probe() returns early due to defer_all_probes being set:
Later, phy_attach_direct() notices no PHY driver has been bound,
and falls back to the Generic PHY, leading to degraded operation:
Generic PHY e6800000.ethernet-ffffffff:00: attached PHY driver [Generic PHY] (mii_bus:phy_addr=e6800000.ethernet-ffffffff:00, irq=POLL)
ravb e6800000.ethernet eth0: Link is Up - 1Gbps/Full - flow control off
B. If CONFIG_MODULES=y, request_module() returns early with -EBUSY due
to UMH_DISABLED, and MDIO initialization fails completely:
mdio_bus e6800000.ethernet-ffffffff:00: error -16 loading PHY driver module for ID 0x00221622
ravb e6800000.ethernet eth0: failed to initialize MDIO
PM: dpm_run_callback(): ravb_resume+0x0/0x1b8 returns -16
PM: Device e6800000.ethernet failed to resume: error -16
Ignoring -EBUSY in phy_request_driver_module(), like was done for
-ENOENT in commit cfa35c391a18a346 ("net: phy: fix issue with loading
PHY driver w/o initramfs"), would makes it fall back to the Generic
PHY, like in the CONFIG_MODULES=n case.
Signed-off-by: Geert Uytterhoeven <geert+renesas@glider.be> Cc: stable@vger.kernel.org Reviewed-by: Sergei Shtylyov <sergei.shtylyov@gmail.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Mat Martineau [Mon, 21 Sep 2020 14:57:58 +0000 (16:57 +0200)]
mptcp: Wake up MPTCP worker when DATA_FIN found on a TCP FIN packet
When receiving a DATA_FIN MPTCP option on a TCP FIN packet, the DATA_FIN
information would be stored but the MPTCP worker did not get
scheduled. In turn, the MPTCP socket state would remain in
TCP_ESTABLISHED and no blocked operations would be awakened.
TCP FIN packets are seen by the MPTCP socket when moving skbs out of the
subflow receive queues, so schedule the MPTCP worker when a skb with
DATA_FIN but no data payload is moved from a subflow queue. Other cases
(DATA_FIN on a bare TCP ACK or on a packet with data payload) are
already handled.
Closes: https://github.com/multipath-tcp/mptcp_net-next/issues/84 Fixes: 2e59714ed2c9 ("mptcp: Use full MPTCP-level disconnect state machine") Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: Mat Martineau <mathew.j.martineau@linux.intel.com> Signed-off-by: Matthieu Baerts <matthieu.baerts@tessares.net> Signed-off-by: David S. Miller <davem@davemloft.net>
Merge branch 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs
Pull vfs fixes from Al Viro:
"No common topic, just assorted fixes"
* 'fixes' of git://git.kernel.org/pub/scm/linux/kernel/git/viro/vfs:
fuse: fix the ->direct_IO() treatment of iov_iter
fs: fix cast in fsparam_u32hex() macro
vboxsf: Fix the check for the old binary mount-arguments struct
- fix failure to add bond interfaces to a bridge, the offload-handling
code was too defensive there and recent refactoring unearthed that.
Users complained (Ido)
- fix unnecessarily reflecting ECN bits within TOS values / QoS marking
in TCP ACK and reset packets (Wei)
- fix a deadlock with bpf iterator. Hopefully we're in the clear on
this front now... (Yonghong)
- BPF fix for clobbering r2 in bpf_gen_ld_abs (Daniel)
- fix AQL on mt76 devices with FW rate control and add a couple of AQL
issues in mac80211 code (Felix)
- fix authentication issue with mwifiex (Maximilian)
- WiFi connectivity fix: revert IGTK support in ti/wlcore (Mauro)
- fix exception handling for multipath routes via same device (David
Ahern)
- revert back to a BH spin lock flavor for nsid_lock: there are paths
which do require the BH context protection (Taehee)
- fix interrupt / queue / NAPI handling in the lantiq driver (Hauke)
- fix ife module load deadlock (Cong)
- make an adjustment to netlink reply message type for code added in
this release (the sole change touching uAPI here) (Michal)
- a number of fixes for small NXP and Microchip switches (Vladimir)
[ Pull request acked by David: "you can expect more of this in the
future as I try to delegate more things to Jakub" ]
* git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (167 commits)
net: mscc: ocelot: fix some key offsets for IP4_TCP_UDP VCAP IS2 entries
net: dsa: seville: fix some key offsets for IP4_TCP_UDP VCAP IS2 entries
net: dsa: felix: fix some key offsets for IP4_TCP_UDP VCAP IS2 entries
inet_diag: validate INET_DIAG_REQ_PROTOCOL attribute
net: bridge: br_vlan_get_pvid_rcu() should dereference the VLAN group under RCU
net: Update MAINTAINERS for MediaTek switch driver
net/mlx5e: mlx5e_fec_in_caps() returns a boolean
net/mlx5e: kTLS, Avoid kzalloc(GFP_KERNEL) under spinlock
net/mlx5e: kTLS, Fix leak on resync error flow
net/mlx5e: kTLS, Add missing dma_unmap in RX resync
net/mlx5e: kTLS, Fix napi sync and possible use-after-free
net/mlx5e: TLS, Do not expose FPGA TLS counter if not supported
net/mlx5e: Fix using wrong stats_grps in mlx5e_update_ndo_stats()
net/mlx5e: Fix multicast counter not up-to-date in "ip -s"
net/mlx5e: Fix endianness when calculating pedit mask first bit
net/mlx5e: Enable adding peer miss rules only if merged eswitch is supported
net/mlx5e: CT: Fix freeing ct_label mapping
net/mlx5e: Fix memory leak of tunnel info when rule under multipath not ready
net/mlx5e: Use synchronize_rcu to sync with NAPI
net/mlx5e: Use RCU to protect rq->xdp_prog
...
Merge tag 'io_uring-5.9-2020-09-22' of git://git.kernel.dk/linux-block
Pull io_uring fixes from Jens Axboe:
"A few fixes - most of them regression fixes from this cycle, but also
a few stable heading fixes, and a build fix for the included demo tool
since some systems now actually have gettid() available"
* tag 'io_uring-5.9-2020-09-22' of git://git.kernel.dk/linux-block:
io_uring: fix openat/openat2 unified prep handling
io_uring: mark statx/files_update/epoll_ctl as non-SQPOLL
tools/io_uring: fix compile breakage
io_uring: don't use retry based buffered reads for non-async bdev
io_uring: don't re-setup vecs/iter in io_resumit_prep() is already there
io_uring: don't run task work on an exiting task
io_uring: drop 'ctx' ref on task work cancelation
io_uring: grab any needed state during defer prep
Merge tag 'block-5.9-2020-09-22' of git://git.kernel.dk/linux-block
Pull block fixes from Jens Axboe:
"A few NVMe fixes, and a dasd write zero fix"
* tag 'block-5.9-2020-09-22' of git://git.kernel.dk/linux-block:
nvmet: get transport reference for passthru ctrl
nvme-core: get/put ctrl and transport module in nvme_dev_open/release()
nvme-tcp: fix kconfig dependency warning when !CRYPTO
nvme-pci: disable the write zeros command for Intel 600P/P3100
s390/dasd: Fix zero write for FBA devices
Merge tag 'trace-v5.9-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace
Pull tracing fixes from Steven Rostedt:
- Check kprobe is enabled before unregistering from ftrace as it isn't
registered when disabled.
- Remove kprobes enabled via command-line that is on init text when
freed.
- Add missing RCU synchronization for ftrace trampoline symbols removed
from kallsyms.
- Free trampoline on error path if ftrace_startup() fails.
- Give more space for the longer PID numbers in trace output.
- Fix a possible double free in the histogram code.
- A couple of fixes that were discovered by sparse.
* tag 'trace-v5.9-rc5' of git://git.kernel.org/pub/scm/linux/kernel/git/rostedt/linux-trace:
bootconfig: init: make xbc_namebuf static
kprobes: tracing/kprobes: Fix to kill kprobes on initmem after boot
tracing: fix double free
ftrace: Let ftrace_enable_sysctl take a kernel pointer buffer
tracing: Make the space reserved for the pid wider
ftrace: Fix missing synchronize_rcu() removing trampoline from kallsyms
ftrace: Free the trampoline when ftrace_startup() fails
kprobes: Fix to check probe enabled before disarm_kprobe_ftrace()
====================
Fix broken tc-flower rules for mscc_ocelot switches
All 3 switch drivers from the Ocelot family have the same bug in the
VCAP IS2 key offsets, which is that some keys are in the incorrect
order.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Vladimir Oltean [Mon, 21 Sep 2020 22:56:38 +0000 (01:56 +0300)]
net: mscc: ocelot: fix some key offsets for IP4_TCP_UDP VCAP IS2 entries
The IS2 IP4_TCP_UDP key offsets do not correspond to the VSC7514
datasheet. Whether they work or not is unknown to me. On VSC9959 and
VSC9953, with the same mistake and same discrepancy from the
documentation, tc-flower src_port and dst_port rules did not work, so I
am assuming the same is true here.
Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Xiaoliang Yang [Mon, 21 Sep 2020 22:56:36 +0000 (01:56 +0300)]
net: dsa: felix: fix some key offsets for IP4_TCP_UDP VCAP IS2 entries
Some of the IS2 IP4_TCP_UDP keys are not correct, like L4_DPORT,
L4_SPORT and other L4 keys. This prevents offloaded tc-flower rules from
matching on src_port and dst_port for TCP and UDP packets.
Signed-off-by: Xiaoliang Yang <xiaoliang.yang_1@nxp.com> Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
Fixes: 85d75fd89e14 ("inet_diag: support for wider protocol numbers") Signed-off-by: Eric Dumazet <edumazet@google.com> Cc: Paolo Abeni <pabeni@redhat.com> Cc: Christoph Paasch <cpaasch@apple.com> Cc: Mat Martineau <mathew.j.martineau@linux.intel.com> Acked-by: Paolo Abeni <pabeni@redhat.com> Signed-off-by: David S. Miller <davem@davemloft.net>
The warning is because br_vlan_get_pvid_rcu() calls nbp_vlan_group()
which calls rtnl_dereference() instead of rcu_dereference(). In turn,
rtnl_dereference() calls rcu_dereference_protected() which assumes
operation under an RCU write-side critical section, which obviously is
not the case here. So, when the incorrect primitive is used to access
the RCU-protected VLAN group pointer, READ_ONCE() is not used, which may
cause various unexpected problems.
I'm sad to say that br_vlan_get_pvid() and br_vlan_get_pvid_rcu() cannot
share the same implementation. So fix the bug by splitting the 2
functions, and making br_vlan_get_pvid_rcu() retrieve the VLAN groups
under proper locking annotations.
Fixes: 76487dd779cc ("bridge: add br_vlan_get_pvid_rcu()") Signed-off-by: Vladimir Oltean <vladimir.oltean@nxp.com> Signed-off-by: David S. Miller <davem@davemloft.net>
David S. Miller [Tue, 22 Sep 2020 00:32:42 +0000 (17:32 -0700)]
Merge tag 'mlx5-fixes-2020-09-18' of git://git.kernel.org/pub/scm/linux/kernel/git/saeed/linux
Saeed Mahameed says:
====================
mlx5 fixes-2020-09-18
This series introduces some fixes to mlx5 driver.
Please pull and let me know if there is any problem.
v1->v2:
Remove missing patch from -stable list.
For -stable v5.1
('net/mlx5: Fix FTE cleanup')
For -stable v5.3
('net/mlx5e: TLS, Do not expose FPGA TLS counter if not supported')
('net/mlx5e: Enable adding peer miss rules only if merged eswitch is supported')
For -stable v5.7
('net/mlx5e: Fix memory leak of tunnel info when rule under multipath not ready')
For -stable v5.8
('net/mlx5e: Use RCU to protect rq->xdp_prog')
('net/mlx5e: Fix endianness when calculating pedit mask first bit')
('net/mlx5e: Use synchronize_rcu to sync with NAPI')
====================
Signed-off-by: David S. Miller <davem@davemloft.net>
Sean Wang [Mon, 21 Sep 2020 23:09:23 +0000 (07:09 +0800)]
net: Update MAINTAINERS for MediaTek switch driver
Update maintainers for MediaTek switch driver with Landen Chao who is
familiar with MediaTek MT753x switch devices and will help maintenance
from the vendor side.
Cc: Steven Liu <steven.liu@mediatek.com> Signed-off-by: Sean Wang <sean.wang@mediatek.com> Signed-off-by: Landen Chao <Landen.Chao@mediatek.com> Signed-off-by: David S. Miller <davem@davemloft.net>
net/mlx5e: kTLS, Avoid kzalloc(GFP_KERNEL) under spinlock
The spinlock only needed when accessing the channel's icosq, grab the lock
after the buf allocation in resync_post_get_progress_params() to avoid
kzalloc(GFP_KERNEL) in atomic context.
Tariq Toukan [Sun, 28 Jun 2020 10:06:06 +0000 (13:06 +0300)]
net/mlx5e: TLS, Do not expose FPGA TLS counter if not supported
The set of TLS TX global SW counters in mlx5e_tls_sw_stats_desc
is updated from all rings by using atomic ops.
This set of stats is used only in the FPGA TLS use case, not in
the Connect-X TLS one, where regular per-ring counters are used.
Do not expose them in the Connect-X use case, as this would cause
counter duplication. For example, tx_tls_drop_no_sync_data would
appear twice in the ethtool stats.