If phydev->irq is set unconditionally, check
for valid interrupt handler or fall back to polling mode to prevent
nullptr exceptions in interrupt service routine.
Signed-off-by: Andre Werner <andre.werner@systec-electronic.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Link: https://lore.kernel.org/r/20240129135734.18975-2-andre.werner@systec-electronic.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

lan743x_ptp_request_tx_timestamp uses spin_lock_bh, but it is
only called from lan743x_tx_xmit_frame where all IRQs are
already disabled.

This fixes the "IRQs not enabled as expected" warning.
Signed-off-by: Lucas Tanure <tanure@linux.com>
Link: https://lore.kernel.org/r/20240128101849.107298-1-tanure@linux.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

This is cosmetics. Move the call of xa_erase() in dpll_pin_put()
so the order of cleanup calls matches the error path of
dpll_pin_alloc().
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Vadim Fedorenko <vadim.fedorenko@linux.dev>
Link: https://lore.kernel.org/r/20240130155814.268622-1-jiri@resnulli.usSigned-off-by: Jakub Kicinski <kuba@kernel.org>

The test sends packets and compares enqueue, transmit and Ack
timestamps with expected values. It installs netem delays to increase
latency between these points.

The test proves flaky in virtual environment (vng). Increase the
delays to reduce variance. Scale measurement tolerance accordingly.

Time sensitive tests are difficult to calibrate. Increasing delays 10x
also increases runtime 10x, for one. And it may still prove flaky at
some rate.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20240127023212.3746239-1-willemdebruijn.kernel@gmail.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Florian Westphal says:

====================
nf-next pr 2024-01-29

This batch contains updates for your *next* tree.

First three changes, from Phil Sutter, allow userspace to define
a table that is exclusively owned by a daemon (via netlink socket
aliveness) without auto-removing this table when the userspace program
exits.  Such table gets marked as orphaned and a restarting management
daemon may re-attach/reassume ownership.

Next patch, from Pablo, passes already-validated flags variable around
rather than having called code re-fetch it from netlnik message.

Patches 5 and 6 update ipvs and nf_conncount to use the recently
introduced KMEM_CACHE() macro.

Last three patches, from myself, tweak kconfig logic a little to
permit a kernel configuration that can run iptables-over-nftables
but not classic (setsockopt) iptables.

Such builds lack the builtin-filter/mangle/raw/nat/security tables,
the set/getsockopt interface and the "old blob format"
interpreter/traverser.  For now, this is 'oldconfig friendly', users
need to manually deselect existing config options for this.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Stefan Wahren says:

====================
qca_spi: collection of improvements

This series contains a wild collection of improvements for the
qca_spi driver. This is a follow-up series to the recent bugfixes [1].

Patch 1, 2 & 3 in this series is the initially intended rework of
netdev_open/close. Patch 4 & 10 are minor functional improvements and
the rest is clean-up.

Changes in V4:
- keep thread handling in netdev_open, but improve this by two
  separate patches ( 1 & 2 )
- add missing link to bugfix series

Changes in V3:
- rebase on net-next-20240124
- improve commit message in Patch 8
- add Reviewed-by's from Jacob Keller
- add Patch 13 for mailmap as suggested by Jacob
- add Patch 14 for MAINTAINERS entry

Changes in V2:
- rebase on net-next-20231218
- improve cover letter

[1] - https://lore.kernel.org/netdev/20231206141222.52029-1-wahrenst@gmx.net/
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Since upstreaming i contributed a lot to this driver(s),
so add myself as a maintainer.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add a .mailmap entry because my old i2se.com address will be
deactivated in the near future.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

The company I2SE has been acquired a long time ago. Switch to
my private mail address before the I2SE account is deactivated.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

According to MODULE_LICENSE the driver is under a dual license.
So replace the BSD license text with the proper SPDX tag.
Signed-off-by: Stefan Wahren <stefan.wahren@i2se.com>
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

All known SPI registers of the QCA700x are 16 bit long. So adjust
the formater width accordingly.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Most of the users doesn't know the expected signature of the QCA700x.
So provide it within the error message. Btw use lowercase for hex as
in the rest of the driver.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

There are two points with the calculation of RX buffer size which are
not optimal:
1. dev->mtu is a mutual parameter and it's currently initialized with
   QCAFRM_MAX_MTU. But for RX buffer size calculation we always need the
   maximum possible MTU. So better use the define directly.
2. This magic number 4 represent the hardware generated frame length
   which is specific to SPI. We better replace this with the suitable
   define.

There is no functional change.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Currently qca_spi reserves enough space for 4 complete Ethernet over SPI
frames in the receive buffer. Unfortunately this is hidden under a magic
number. So replace it with a more self explaining define.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

All defines in qca_spi.h except of the two ring limit defines have
a QCASPI prefix. Since the name is quite generic add the QCASPI prefix
to avoid possible name conflicts.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This member is never used. So drop it.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

qcafrm_fsm_decode has the almost the same function description in
qca_7k_common.c. So drop the comment here.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The skb spare room needs to be expanded for SPI header, footer
and possible padding within the TX path. So announce the necessary
space in order to avoid expensive skb_copy_expand calls.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The functions qcaspi_netdev_open/close are responsible of request &
free of the SPI interrupt, which wasn't the best choice because
allocation problems are discovered not during probe. So let us split
IRQ allocation & enabling, so we can take advantage of a device
managed IRQ.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Directly storing the result of kthread_run within the private
driver data makes it harder to identify if the pointer has a
running thread or not. So better use a local variable for
the result check and we don't have to care about error pointer
in the rest of the code.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

We better not rely on that spi_thread points to a running
thread. So add an check for this.
Signed-off-by: Stefan Wahren <wahrenst@gmx.net>
Signed-off-by: David S. Miller <davem@davemloft.net>

On lan966x, it is possible to use debugfs to print different information
about the VCAPs. Information like, if it is enabled, how the ports are
configured, print the actual rules. The issue is that when printing how
the ports are configured for IS1 lookups, it was parsing the wrong
register to get this information. The fix consists in reading the
correct register that contains this information.
Signed-off-by: Horatiu Vultur <horatiu.vultur@microchip.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Heiner Kallweit says:

====================
ethtool: switch EEE netlink interface to use EEE linkmode bitmaps

So far only 32bit legacy bitmaps are passed to userspace. This makes
it impossible to manage EEE linkmodes beyond bit 32, e.g. manage EEE
for 2500BaseT and 5000BaseT. This series adds support for passing
full linkmode bitmaps between kernel and userspace.

Fortunately the netlink-based part of ethtool is quite smart and no
changes are needed in ethtool. However this applies to the netlink
interface only, the ioctl interface for now remains restricted to
legacy bitmaps.

Next step will be adding support for the c45 EEE2 standard registers
(3.21, 7.62, 7.63) to the genphy_c45 functions dealing with EEE.
I have a follow-up series for this ready to be submitted.

v2:
- now as RFC
- adopt suggestion from Andrew to start with struct ethtool_keee
  being an identical copy of ethtool_eee, and switch all users
v3:
- switch from RFC to net-next
- add patch 4, and reuse old names in patch 5
- rebase patch 1
v4:
- fix missing replacement in patch 4
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Change genphy_c45_ethtool_[get|set]_eee to use EEE linkmode bitmaps.
This is a prerequisite for adding support for EEE modes beyond bit 31.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add linkmode bitmap members to struct ethtool_keee, but keep the legacy
u32 bitmaps for compatibility with existing drivers.
Use linkmode "supported" not being empty as indicator that a user wants
to use the linkmode bitmap members instead of the legacy bitmaps.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

This is in preparation of using the existing names for linkmode
bitmaps.
Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

This patch changes the following in struct ethtool_keee
- remove member cmd, it's not needed on kernel side
- remove reserved fields
- switch the semantically boolean members to type bool

We don't have to change any user of the boolean members due to the
implicit casting from/to bool. A small change is needed where a
pointer to bool members is used, in addition remove few now unneeded
double negations.
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to later extend struct ethtool_keee, we have to decouple it
from the userspace format represented by struct ethtool_eee.
Therefore switch back to struct ethtool_eee, representing the userspace
format, and add conversion between ethtool_eee and ethtool_keee.
Struct ethtool_keee will be changed in follow-up patches, therefore
don't do a *keee = *eee here.
Member cmd isn't copied, because it's not used, and we'll remove
it in the next patch of this series. In addition omit setting cmd
to ETHTOOL_GEEE in the ioctl response, userspace ethtool isn't
interested in it.
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to pass EEE link modes beyond bit 32 to userspace we have to
complement the 32 bit bitmaps in struct ethtool_eee with linkmode
bitmaps. Therefore, similar to ethtool_link_settings and
ethtool_link_ksettings, add a struct ethtool_keee. In a first step
it's an identical copy of ethtool_eee. This patch simply does a
s/ethtool_eee/ethtool_keee/g for all users.
No functional change intended.
Suggested-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: Heiner Kallweit <hkallweit1@gmail.com>
Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Signed-off-by: David S. Miller <davem@davemloft.net>

Rohan G says:

====================
net: stmmac: EST conformance support

This patchset enables support for queueMaxSDU and transmission overrun
counters which are required for Qbv conformance.
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

Report taprio offload status. This includes per txq and global
counters of window_drops and tx_overruns.

Window_drops count include count of frames dropped because of
queueMaxSDU setting and HLBF error. Transmission overrun counter
inform the user application whether any packets are currently being
transmitted on a particular queue during a gate-close event.DWMAC IPs
takes care Transmission overrun won't happen hence this is always 0.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Keep per Tx-queue error count on Head-Of-Line Blocking due to frame
size(HLBF) error. The MAC raises HLBF error on one or more queues
when none of the time Intervals of open-gates in the GCL is greater
than or equal to the duration needed for frame transmission and by
default drops those packets that causes HLBF error. EST_FRM_SZ_ERR
register provides the One Hot encoded Queue numbers that have the
Frame Size related error.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Add support for configuring queueMaxSDU. As DWMAC IPs doesn't support
queueMaxSDU table handle this in the SW. The maximum 802.3 frame size
that is allowed to be transmitted by any queue is queueMaxSDU +
16 bytes (i.e. 6 bytes SA + 6 bytes DA + 4 bytes FCS).

Inspired from intel i225 driver.
Signed-off-by: Rohan G Thomas <rohan.g.thomas@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

This test validates per-band packet limits in FQ. Packets are dropped
rather than enqueued if the limit for their band is reached.

This test is timing sensitive. It queues packets in FQ with a future
delivery time to fill the qdisc.

The test failed in a virtual environment (vng). Increase the delays
to make it more tolerant to environments with timing variance.
Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Benjamin Poirier says:

====================
selftests: Add TEST_INCLUDES directive and adjust tests to use it

After commit 25ae948b ("selftests/net: add lib.sh") but before commit
2114e833 ("selftests: forwarding: Avoid failures to source
net/lib.sh"), some net selftests encountered errors when they were being
exported and run. This was because the new net/lib.sh was not exported
along with the tests. The errors were crudely avoided by duplicating some
content between net/lib.sh and net/forwarding/lib.sh in 2114e833.

In order to restore the sourcing of net/lib.sh from net/forwarding/lib.sh
and remove the duplicated content, this series introduces a new selftests
Makefile variable to list extra files to export from other directories and
makes use of it to avoid reintroducing the errors mentioned above.

v2:
* "selftests: Introduce Makefile variable to list shared bash scripts"
Fix rst syntax in Documentation/dev-tools/kselftest.rst (Jakub Kicinski)

v1:
* "selftests: Introduce Makefile variable to list shared bash scripts"
Changed TEST_INCLUDES to take relative paths, like other TEST_* variables.
Paths are adjusted accordingly in the subsequent patches. (Vladimir Oltean)

* selftests: bonding: Change script interpreter
  selftests: forwarding: Remove executable bits from lib.sh
Removed from this series, submitted separately.

Since commit 2114e833 ("selftests: forwarding: Avoid failures to source
net/lib.sh") resolved the test errors, this version of the series is
focused on removing the duplication that was added in that commit. Directly
rebasing the series would reintroduce the problems that 2114e833
avoided before fixing them again. In order to prevent such breakage partway
through the series, patches are reordered and content changed slightly but
there is no diff at the end compared with the simple rebasing approach. I
have dropped most review tags on account of this reordering.

RFC:
https://lore.kernel.org/netdev/20231222135836.992841-1-bpoirier@nvidia.com/

Link: https://lore.kernel.org/netdev/ZXu7dGj7F9Ng8iIX@Laptop-X1/
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

commit 25ae948b ("selftests/net: add lib.sh") added net/lib.sh to
contain code shared by tests under net/ and net/forwarding/. However, this
caused issues with selftests from directories other than net/forwarding/,
in particular those under drivers/net/. Those issues were avoided in a
simple way by duplicating some content in commit 2114e833 ("selftests:
forwarding: Avoid failures to source net/lib.sh").

In order to remove the duplicated content, restore the inclusion of
net/lib.sh from net/forwarding/lib.sh but with the following changes:

* net/lib.sh is imported through the net_forwarding_dir path
The original expression "source ../lib.sh" would look for lib.sh in the
directory above the script file's one, which did not work for tests under
drivers/net/.

* net/lib.sh is added to TEST_INCLUDES
Since net/forwarding/lib.sh now sources net/lib.sh, both of those files
must be exported along with tests which source net/forwarding/lib.sh.
Suggested-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The following code which is part of lib.sh:
relative_path="${BASH_SOURCE%/*}"
if [[ "$relative_path" == "${BASH_SOURCE}" ]]; then
       relative_path="."
fi

reimplements functionality that is part of `dirname`:
$ dirname ""
.

To avoid this duplication, replace "relative_path" by "net_forwarding_dir",
a new variable defined using dirname.

Furthermore, to avoid the potential confusion about what "relative_path" is
about (cwd, test script directory or test library directory), define
"net_forwarding_dir" as the absolute path to net/forwarding/.
Tested-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The dsa tests which are symlinks of tests from net/forwarding/ (like
tc_actions.sh) become regular files after export (because `rsync
--copy-unsafe-links` is used) and expect to source lib.sh
(net/forwarding/lib.sh) from the same directory.

In the last patch of this series, net/forwarding/lib.sh will source lib.sh
from its parent directory (ie. net/lib.sh). This would not work for dsa
tests because net/lib.sh is not present under drivers/net/.

Since the tests in net/forwarding/ are not meant to be copied and run from
another directory, as a preparation for that last patch, replace the test
symlinks by a wrapper script which runs the original tests under
net/forwarding/. Following from that, the links to shared library scripts
in dsa/ are no longer used so remove them and add all the original files
needed from parent directories to TEST_INCLUDES.
Suggested-by: Hangbin Liu <liuhangbin@gmail.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Tested-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to avoid duplicated files when both the team and bonding tests are
exported together, add lag_lib.sh to TEST_INCLUDES.

Do likewise for net/forwarding/lib.sh regarding team and forwarding tests.
Reviewed-by: Petr Machata <petrm@nvidia.com>
Tested-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

In order to avoid duplicated files when both the bonding and forwarding
tests are exported together, add net/forwarding/lib.sh to TEST_INCLUDES and
include it via its relative path.
Reviewed-by: Petr Machata <petrm@nvidia.com>
Tested-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Hangbin Liu <liuhangbin@gmail.com>
Signed-off-by: Benjamin Poirier <bpoirier@nvidia.com>
Reviewed-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: David S. Miller <davem@davemloft.net>