Commit 57f230ab ("xen/netfront: raise max number of slots in
xennet_get_responses()") raised the max number of allowed slots by one.
This seems to be problematic in some configurations with netback using
a larger MAX_SKB_FRAGS value (e.g. old Linux kernel with MAX_SKB_FRAGS
defined as 18 instead of nowadays 17).

Instead of BUG_ON() in this case just fall back to retransmission.

Fixes: 57f230ab ("xen/netfront: raise max number of slots in xennet_get_responses()")
Cc: stable@vger.kernel.org
Signed-off-by: NJuergen Gross <jgross@suse.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This fixes a bug where ipv6 tunnels would report that it is
getting offloaded to hardware but would actually be rejected
by hardware.

Fixes: b27d6a95 ("nfp: compile flower vxlan tunnel set actions")
Signed-off-by: NLouis Peens <louis.peens@netronome.com>
Reviewed-by: NJohn Hurley <john.hurley@netronome.com>
Reviewed-by: NSimon Horman <simon.horman@netronome.com>
Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Previously we only checked if the vlan id field is present when trying
to match a vlan tag. The vlan id and vlan pcp field should be treated
independently.

Fixes: 5571e8c9 ("nfp: extend flower matching capabilities")
Signed-off-by: NPieter Jansen van Vuuren <pieter.jansenvanvuuren@netronome.com>
Reviewed-by: NJakub Kicinski <jakub.kicinski@netronome.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

After system suspend, sometimes the r8169 doesn't work when ethernet
cable gets pluggued.

This issue happens because rtl_reset_work() doesn't get called from
rtl8169_runtime_resume(), after system suspend.

In rtl_task(), RTL_FLAG_TASK_* only gets cleared if this condition is
met:
if (!netif_running(dev) ||
    !test_bit(RTL_FLAG_TASK_ENABLED, tp->wk.flags))
    ...

If RTL_FLAG_TASK_ENABLED was cleared during system suspend while
RTL_FLAG_TASK_RESET_PENDING was set, the next rtl_schedule_task() won't
schedule task as the flag is still there.

So in addition to clearing RTL_FLAG_TASK_ENABLED, also clears other
flags.

Cc: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: NKai-Heng Feng <kai.heng.feng@canonical.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

dma_zalloc_coherent() now crashes if no dev pointer is given.
Add a dev pointer to the ltq_dma_channel structure and fill it in the
driver using it.

This fixes a bug introduced in kernel 4.19.
Signed-off-by: NHauke Mehrtens <hauke@hauke-m.de>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Quectel EP06 (and EM06/EG06) supports dynamic configuration of USB
interfaces, without the device changing VID/PID or configuration number.
When the configuration is updated and interfaces are added/removed, the
interface numbers change. This means that the current code for matching
EP06 does not work.

This patch removes the current EP06 interface number match, and replaces
it with a match on class, subclass and protocol. Unfortunately, matching
on those three alone is not enough, as the diag interface exports the
same values as QMI. The other serial interfaces + adb export different
values and do not match.

The diag interface only has two endpoints, while the QMI interface has
three. I have therefore added a check for number of interfaces, and we
ignore the interface if the number of endpoints equals two.
Signed-off-by: NKristian Evensen <kristian.evensen@gmail.com>
Acked-by: NBjørn Mork <bjorn@mork.no>
Acked-by: NDan Williams <dcbw@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch updates license to use SPDX-License-Identifier
instead of verbose license text.
Signed-off-by: NKuninori Morimoto <kuninori.morimoto.gx@renesas.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Added memory barriers where they were missing to support multiple
architectures, and removed redundant ones.

As part of removing the redundant memory barriers and improving
performance, we moved to more relaxed versions of memory barriers,
as well as to the more relaxed version of writel - writel_relaxed,
while maintaining correctness.
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Add READ_ONCE calls where necessary (for example when iterating
over a memory field that gets updated by the hardware).
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

acquire the rtnl_lock during device destruction to avoid
using partially destroyed device.

ena_remove() shares almost the same logic as ena_destroy_device(),
so use ena_destroy_device() and avoid duplications.
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

ena_destroy_device() can potentially be called twice.
To avoid this, check that the device is running and
only then proceed destroying it.
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When ena_destroy_device() is called from ena_suspend(), the device is
still reachable from the driver. Therefore, the driver can send a command
to the device to free all resources.
However, in all other cases of calling ena_destroy_device(), the device is
potentially in an error state and unreachable from the driver. In these
cases the driver must not send commands to the device.

The current implementation does not request resource freeing from the
device even when possible. We add the graceful parameter to
ena_destroy_device() to enable resource freeing when possible, and
use it in ena_suspend().
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The buffer length field in the ena rx descriptor is 16 bit, and the
current driver passes a full page in each ena rx descriptor.
When PAGE_SIZE equals 64kB or more, the buffer length field becomes
zero.
To solve this issue, limit the ena Rx descriptor to use 16kB even
when allocating 64kB kernel pages. This change would not impact ena
device functionality, as 16kB is still larger than maximum MTU.
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Starting with driver version 1.5.0, in case of a surprise device
unplug, there is a race caused by invoking ena_destroy_device()
from two different places. As a result, the readless register might
be accessed after it was destroyed.
Signed-off-by: NNetanel Belgazal <netanel@amazon.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 822fb18a ("xen-netfront: wait xenbus state change when load
module manually") added a new wait queue to wait on for a state change
when the module is loaded manually. Unfortunately there is no wakeup
anywhere to stop that waiting.

Instead of introducing a new wait queue rename the existing
module_unload_q to module_wq and use it for both purposes (loading and
unloading).

As any state change of the backend might be intended to stop waiting
do the wake_up_all() in any case when netback_changed() is called.

Fixes: 822fb18a ("xen-netfront: wait xenbus state change when load module manually")
Cc: <stable@vger.kernel.org> #4.18
Signed-off-by: NJuergen Gross <jgross@suse.com>
Reviewed-by: NBoris Ostrovsky <boris.ostrovsky@oracle.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Commit 3559d81e ("r8169: simplify rtl_hw_start_8169") changed order of
two register writes:
1) Caused RxConfig to be written before TX / RX is enabled,
2) Caused TxConfig to be written before TX / RX is enabled.

At least on XIDs 10000000 ("RTL8169sb/8110sb") and
18000000 ("RTL8169sc/8110sc") such writes are ignored by the chip, leaving
values in these registers intact.

Change 1) was reverted by
commit 05212ba8 ("r8169: set RxConfig after tx/rx is enabled for RTL8169sb/8110sb devices"),
however change 2) wasn't.

In practice, this caused TxConfig's "InterFrameGap time" and "Max DMA Burst
Size per Tx DMA Burst" bits to be zero dramatically reducing TX performance
(in my tests it dropped from around 500Mbps to around 50Mbps).

This patch fixes the issue by moving TxConfig register write a bit later in
the code so it happens after TX / RX is already enabled.

Fixes: 05212ba8 ("r8169: set RxConfig after tx/rx is enabled for RTL8169sb/8110sb devices")
Signed-off-by: NMaciej S. Szmigiero <mail@maciej.szmigiero.name>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This is a false positive report due to incorrect nested lock
annotations as we lock multiple fgs with the same subclass.
Instead of locking all fgs only lock the one being used as was
done before.

Fixes: bd71b08e ("net/mlx5: Support multiple updates of steering rules in parallel")
Signed-off-by: NRoi Dayan <roid@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

Copy and paste bug was introduced in the offending patch.
We need to write udp source port value into the headers value and not
headers criteria "mask".

Fixes: 142644f8 ("net/mlx5e: Ethtool steering flow parsing refactoring")
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

Currently, mlx5_attach_interface does not check for error
after calling intf->attach or intf->add. When these two calls
fails, the client is not initialized and will cause issues such as
kernel panic on invalid address in the teardown path (mlx5_detach_interface)

Fixes: 737a234b ("net/mlx5: Introduce attach/detach to interface API")
Signed-off-by: NHuy Nguyen <huyn@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

The PCI BDF is not unique. PCI domain must also be considered when
searching for the next physical device during lag setup. Example below:

mlx5_core 0000:01:00.0: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0000:01:00.1: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0001:01:00.0: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
mlx5_core 0001:01:00.1: MLX5E: StrdRq(1) RqSz(8) StrdSz(128) RxCqeCmprss(0)
Signed-off-by: NDaniel Jurgens <danielj@mellanox.com>
Reviewed-by: NAviv Heller <avivh@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

If building match list fg fails and we never jumped to
search_again_locked label then the function returned without
unlocking the read lock.

Fixes: bd71b08e ("net/mlx5: Support multiple updates of steering rules in parallel")
Signed-off-by: NRoi Dayan <roid@mellanox.com>
Reviewed-by: NMaor Gottlieb <maorg@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

The memory allocated for the slow path table flow group input structure
was not freed upon successful return, fix that.

Fixes: 1967ce6e ("net/mlx5: E-Switch, Refactor fast path FDB table creation in switchdev mode")
Signed-off-by: NRaed Salem <raeds@mellanox.com>
Reviewed-by: NOr Gerlitz <ogerlitz@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

Minimal stride size is 16.
Hence, the number of strides in a fragment (of PAGE_SIZE)
is <= PAGE_SIZE / 16 <= 4K.

u16 is sufficient to represent this.

Fixes: d7037ad7 ("net/mlx5: Fix QP fragmented buffer allocation")
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Reviewed-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

Minimal stride size is 16.
Hence, the number of strides in a fragment (of PAGE_SIZE)
is <= PAGE_SIZE / 16 <= 4K.

u16 is sufficient to represent this.

Fixes: 388ca8be ("IB/mlx5: Implement fragmented completion queue (CQ)")
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Reviewed-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

When initializing the device (procedure init_one), the driver
calls mlx5_pci_init to perform pci initialization. As part of this
initialization, mlx5_pci_init creates a debugfs directory.
If this creation fails, init_one aborts, returning failure to
the caller (which is the probe method caller).

The main reason for such a failure to occur is if the debugfs
directory already exists. This can happen if the last time
mlx5_pci_close was called, debugfs_remove (silently) failed due
to the debugfs directory not being empty.

Guarantee that such a debugfs_remove failure will not occur by
instead calling debugfs_remove_recursive in procedure mlx5_pci_close.

Fixes: 59211bd3 ("net/mlx5: Split the load/unload flow into hardware and software flows")
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: NDaniel Jurgens <danielj@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

When the mlx5 health mechanism detects a problem while the driver
is in the middle of init_one or remove_one, the driver needs to prevent
the health mechanism from scheduling future work; if future work
is scheduled, there is a problem with use-after-free: the system WQ
tries to run the work item (which has been freed) at the scheduled
future time.

Prevent this by disabling work item scheduling in the health mechanism
when the driver is in the middle of init_one() or remove_one().

Fixes: e126ba97 ("mlx5: Add driver for Mellanox Connect-IB adapters")
Signed-off-by: NJack Morgenstein <jackm@dev.mellanox.co.il>
Reviewed-by: NFeras Daoud <ferasda@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>

With performance optimization the spi transfer and messages of basic
register operations like qcaspi_read_register moved into the private
driver structure. But they weren't protected against mutual access
(e.g. between driver kthread and ethtool). So dumping the QCA7000
registers via ethtool during network traffic could make spi_sync
hang forever, because the completion in spi_message is overwritten.

So revert the optimization completely.

Fixes: 291ab06e ("net: qualcomm: new Ethernet over SPI driver for QCA700")
Signed-off-by: NStefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

DMA allocated memory is lost in be_cmd_get_profile_config() when we
call it with non-NULL port_res parameter.
Signed-off-by: NPetr Oros <poros@redhat.com>
Reviewed-by: NIvan Vecera <ivecera@redhat.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

MC-aware mode was recently enabled by mlxsw on Spectrum switches in
commit 7b819530 ("mlxsw: spectrum: Configure MC-aware mode on mlxsw
ports"). Unfortunately, testing has shown that the fix is incomplete and
in the presented form actually makes the problem even worse, because any
amount of MC traffic causes UC disruption.

The reason for this is that currently, mlxsw configures the MC-specific
TCs (8..15) to map to pool 0. It also configures a maximum buffer size
of 0, but for MC traffic that maximum is disregarded and not part of the
quota. Therefore MC traffic is always admitted to the egress buffer.

Fix the configuration by directing the MC TCs into pool 15, which is
dedicated to MC traffic and recognized as such by the silicon.

Fixes: 7b819530 ("mlxsw: spectrum: Configure MC-aware mode on mlxsw ports")
Signed-off-by: NPetr Machata <petrm@mellanox.com>
Acked-by: NJiri Pirko <jiri@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Not all SFPs implement the registers containing sensor limits and
alarms. Luckily, there is a bit indicating if they are implemented or
not. Add checking for this bit, when deciding if the hwmon attributes
should be visible.

Fixes: 1323061a ("net: phy: sfp: Add HWMON support for module sensors")
Signed-off-by: NAndrew Lunn <andrew@lunn.ch>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Correct the formula for calculating the RQ page remainder,
which should be in byte granularity.  The result will be
non-zero only for RQs smaller than PAGE_SIZE, as an RQ size
is a power of 2.

Divide this by the SQ stride (MLX5_SEND_WQE_BB) to get the
SQ offset in strides granularity.

Fixes: d7037ad7 ("net/mlx5: Fix QP fragmented buffer allocation")
Signed-off-by: NTariq Toukan <tariqt@mellanox.com>
Reviewed-by: NEran Ben Elisha <eranbe@mellanox.com>
Signed-off-by: NSaeed Mahameed <saeedm@mellanox.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Currently, the driver adjusts the bp->hw_resc.max_cp_rings by the number
of MSIX vectors used by RDMA.  There is one code path in open that needs
to check the true max_cp_rings including any used by RDMA.  This code
is now checking for the reduced max_cp_rings which will fail when the
number of cp rings is very small.

To fix this in a clean way, we don't adjust max_cp_rings anymore.
Instead, we add a helper bnxt_get_max_func_cp_rings_for_en() to get the
reduced max_cp_rings when appropriate.

Fixes: ec86f14e ("bnxt_en: Add ULP calls to stop and restart IRQs.")
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Remove unused bnxt_subtract_ulp_resources().  Change
bnxt_get_max_func_irqs() to static since it is only locally used.
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

When the driver detects that resources have changed during open, it
should reset the rx and tx rings to 0.  This will properly setup the
init sequence to initialize the default rings again.  We also need
to signal the RDMA driver to stop and clear its interrupts.  We then
call the RoCE driver to restart if a new set of default rings is
successfully reserved.

Fixes: 25e1acd6 ("bnxt_en: Notify firmware about IF state changes.")
Signed-off-by: NMichael Chan <michael.chan@broadcom.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

__emac_calc_base_mr1 was used instead of __emac4_calc_base_mr1
by copy-paste mistake for emac4syn.

Fixes: 45d6e545 ("net/ibm/emac: add 8192 rx/tx fifo size")
Signed-off-by: NIvan Mikhaylov <ivan@de.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This card identifies itself as:
  Ethernet controller [0200]: NCube Device [10ff:8168] (rev 06)
  Subsystem: TP-LINK Technologies Co., Ltd. Device [7470:3468]

Adding a new entry to rtl8169_pci_tbl makes the card work.

Link: http://launchpad.net/bugs/1788730Signed-off-by: NAnthony Wong <anthony.wong@ubuntu.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The kernel module may sleep with holding a spinlock.

The function call paths (from bottom to top) in Linux-4.16 are:

[FUNC] usleep_range
drivers/net/ethernet/cadence/macb_main.c, 648:
	usleep_range in macb_halt_tx
drivers/net/ethernet/cadence/macb_main.c, 730:
	macb_halt_tx in macb_tx_error_task
drivers/net/ethernet/cadence/macb_main.c, 721:
	_raw_spin_lock_irqsave in macb_tx_error_task

To fix this bug, usleep_range() is replaced with udelay().

This bug is found by my static analysis tool DSAC.
Signed-off-by: NJia-Ju Bai <baijiaju1990@gmail.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

The cpsw-phy-sel device is not a child of the cpsw interconnect target
module. It lives in the system control module.

Let's fix this issue by trying to use cpsw-phy-sel phandle first if it
exists and if not fall back to current usage of trying to find the
cpsw-phy-sel child. That way the phy sel driver can be a child of the
system control module where it belongs in the device tree.

Without this fix, we cannot have a proper interconnect target module
hierarchy in device tree for things like genpd.

Note that deferred probe is mostly not supported by cpsw and this patch
does not attempt to fix that. In case deferred probe support is needed,
this could be added to cpsw_slave_open() and phy_connect() so they start
handling and returning errors.

For documenting it, looks like the cpsw-phy-sel is used for all cpsw device
tree nodes. It's missing the related binding documentation, so let's also
update the binding documentation accordingly.

Cc: devicetree@vger.kernel.org
Cc: Andrew Lunn <andrew@lunn.ch>
Cc: Grygorii Strashko <grygorii.strashko@ti.com>
Cc: Ivan Khoronzhuk <ivan.khoronzhuk@linaro.org>
Cc: Mark Rutland <mark.rutland@arm.com>
Cc: Murali Karicheri <m-karicheri2@ti.com>
Cc: Rob Herring <robh+dt@kernel.org>
Signed-off-by: NTony Lindgren <tony@atomide.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

Check the return codes of these functions and halt reset
in case of failure. The driver will remain in a dormant state
until the next reset event, when device initialization will be
re-attempted.
Signed-off-by: NThomas Falcon <tlfalcon@linux.vnet.ibm.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>

This patch fixes the race between netvsc_probe() and
rndis_set_subchannel(), which can cause a deadlock.

These are the related 3 paths which show the deadlock:

path #1:
    Workqueue: hv_vmbus_con vmbus_onmessage_work [hv_vmbus]
    Call Trace:
     schedule
     schedule_preempt_disabled
     __mutex_lock
     __device_attach
     bus_probe_device
     device_add
     vmbus_device_register
     vmbus_onoffer
     vmbus_onmessage_work
     process_one_work
     worker_thread
     kthread
     ret_from_fork

path #2:
    schedule
     schedule_preempt_disabled
     __mutex_lock
     netvsc_probe
     vmbus_probe
     really_probe
     __driver_attach
     bus_for_each_dev
     driver_attach_async
     async_run_entry_fn
     process_one_work
     worker_thread
     kthread
     ret_from_fork

path #3:
    Workqueue: events netvsc_subchan_work [hv_netvsc]
    Call Trace:
     schedule
     rndis_set_subchannel
     netvsc_subchan_work
     process_one_work
     worker_thread
     kthread
     ret_from_fork

Before path #1 finishes, path #2 can start to run, because just before
the "bus_probe_device(dev);" in device_add() in path #1, there is a line
"object_uevent(&dev->kobj, KOBJ_ADD);", so systemd-udevd can
immediately try to load hv_netvsc and hence path #2 can start to run.

Next, path #2 offloads the subchannal's initialization to a workqueue,
i.e. path #3, so we can end up in a deadlock situation like this:

Path #2 gets the device lock, and is trying to get the rtnl lock;
Path #3 gets the rtnl lock and is waiting for all the subchannel messages
to be processed;
Path #1 is trying to get the device lock, but since #2 is not releasing
the device lock, path #1 has to sleep; since the VMBus messages are
processed one by one, this means the sub-channel messages can't be
procedded, so #3 has to sleep with the rtnl lock held, and finally #2
has to sleep... Now all the 3 paths are sleeping and we hit the deadlock.

With the patch, we can make sure #2 gets both the device lock and the
rtnl lock together, gets its job done, and releases the locks, so #1
and #3 will not be blocked for ever.

Fixes: 8195b139 ("hv_netvsc: fix deadlock on hotplug")
Signed-off-by: NDexuan Cui <decui@microsoft.com>
Cc: Stephen Hemminger <sthemmin@microsoft.com>
Cc: K. Y. Srinivasan <kys@microsoft.com>
Cc: Haiyang Zhang <haiyangz@microsoft.com>
Signed-off-by: NDavid S. Miller <davem@davemloft.net>