netdev: replace napi_reschedule with napi_schedule

Now that napi_schedule return a bool, we can drop napi_reschedule that
does the same exact function. The function comes from a very old commit
bfe13f54f502 ("ibm_emac: Convert to use napi_struct independent of struct
net_device") and the purpose is actually deprecated in favour of
different logic.

Convert every user of napi_reschedule to napi_schedule.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Acked-by: Jeff Johnson <quic_jjohnson@quicinc.com> # ath10k
Acked-by: Nick Child <nnac123@linux.ibm.com> # ibm
Acked-by: Marc Kleine-Budde <mkl@pengutronix.de> # for can/dev/rx-offload.c
Reviewed-by: Eric Dumazet <edumazet@google.com>
Acked-by: Tariq Toukan <tariqt@nvidia.com>
Link: https://lore.kernel.org/r/20231009133754.9834-3-ansuelsmth@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netdev: make napi_schedule return bool on NAPI successful schedule

Change napi_schedule to return a bool on NAPI successful schedule.
This might be useful for some driver to do additional steps after a
NAPI has been scheduled.

Suggested-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20231009133754.9834-2-ansuelsmth@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netdev: replace simple napi_schedule_prep/__napi_schedule to napi_schedule

Replace drivers that still use napi_schedule_prep/__napi_schedule
with napi_schedule helper as it does the same exact check and call.

Signed-off-by: Christian Marangi <ansuelsmth@gmail.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20231009133754.9834-1-ansuelsmth@gmail.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bna: replace deprecated strncpy with strscpy_pad

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

bfa_ioc_get_adapter_manufacturer() simply copies a string literal into
`manufacturer`.

Another implementation of bfa_ioc_get_adapter_manufacturer() from
drivers/scsi/bfa/bfa_ioc.c uses memset + strscpy:
| void
| bfa_ioc_get_adapter_manufacturer(struct bfa_ioc_s *ioc, char *manufacturer)
| {
| memset((void *)manufacturer, 0, BFA_ADAPTER_MFG_NAME_LEN);
| strscpy(manufacturer, BFA_MFG_NAME, BFA_ADAPTER_MFG_NAME_LEN);
| }

Let's use `strscpy_pad` to eliminate some redundant work while still
NUL-terminating and NUL-padding the destination buffer.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-ethernet-brocade-bna-bfa_ioc-c-v2-1-78e0f47985d3@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: dsa: lantiq_gswip: replace deprecated strncpy with ethtool_sprintf

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

ethtool_sprintf() is designed specifically for get_strings() usage.
Let's replace strncpy in favor of this more robust and easier to
understand interface.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-dsa-lantiq_gswip-c-v1-1-d55a986a14cc@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: dsa: mt7530: replace deprecated strncpy with ethtool_sprintf

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

ethtool_sprintf() is designed specifically for get_strings() usage.
Let's replace strncpy in favor of this more robust and easier to
understand interface.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Acked-by: Daniel Golle <daniel@makrotopia.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: AngeloGioacchino Del Regno <angelogioacchino.delregno@collabora.com>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-dsa-mt7530-c-v1-1-ec6677a6436a@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: mvpp2: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

We expect `irqname` to be NUL-terminated based on its use with
of_irq_get_byname() -> of_property_match_string() wherein it is used
with a format string and a `strcmp`:
|       pr_debug("comparing %s with %s\n", string, p);
|       if (strcmp(string, p) == 0)
|               return i; /* Found it; return index */

NUL-padding is not required as is evident by other assignments to
`irqname` which do not NUL-pad:
|       if (port->flags & MVPP2_F_DT_COMPAT)
|               snprintf(irqname, sizeof(irqname), "tx-cpu%d", i);
|       else
|               snprintf(irqname, sizeof(irqname), "hif%d", i);

Considering the above, a suitable replacement is `strscpy` [2] due to
the fact that it guarantees NUL-termination on the destination buffer
without unnecessarily NUL-padding.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231010-strncpy-drivers-net-ethernet-marvell-mvpp2-mvpp2_main-c-v1-1-51be96ad0324@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

octeontx2-af: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

We can see that linfo->lmac_type is expected to be NUL-terminated based
on the `... - 1`'s present in the current code. Presumably making room
for a NUL-byte at the end of the buffer.

Considering the above, a suitable replacement is `strscpy` [2] due to
the fact that it guarantees NUL-termination on the destination buffer
without unnecessarily NUL-padding.

Let's also prefer the more idiomatic strscpy usage of (dest, src,
sizeof(dest)) rather than (dest, src, SOME_LEN).

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231010-strncpy-drivers-net-ethernet-marvell-octeontx2-af-cgx-c-v1-1-a443e18f9de8@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

tools: ynl: use ynl-gen -o instead of stdout in Makefile

Jiri added more careful handling of output of the code generator
to avoid wiping out existing files in
commit f65f305ae008 ("tools: ynl-gen: use temporary file for rendering")
Make use of the -o option in the Makefiles, it is already used
by ynl-regen.sh.

Link: https://lore.kernel.org/r/20231010202714.4045168-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netlink: specs: don't allow version to be specified for genetlink

There is no good reason to specify the version for new protocols.
Forbid it in genetlink schema.

If the future proves me wrong, this restriction could be easily lifted.

Move the version definition in between legacy properties
in genetlink-legacy.

Suggested-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Link: https://lore.kernel.org/r/20231010074810.191177-1-jiri@resnulli.us
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'add-vf-fault-detect-support-for-hns3-ethernet-driver'

Jijie Shao says:

====================
add vf fault detect support for HNS3 ethernet driver
====================

Link: https://lore.kernel.org/r/20231007031215.1067758-1-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: hns3: add vf fault detect support

Currently hns3 driver supports vf fault detect feature. Several ras caused
by VF resources don't need to do PF function reset for recovery. The driver
only needs to reset the specified VF.

So this patch adds process in ras module. New process will get detailed
information about ras and do the most correct measures based on these
accurate information.

Signed-off-by: Jie Wang <wangjie125@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://lore.kernel.org/r/20231007031215.1067758-3-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: hns3: add hns3 vf fault detect cap bit support

Currently hns3 driver is designed to support VF fault detect feature in
new hardwares. For code compatibility, vf fault detect cap bit is added to
the driver.

Signed-off-by: Jie Wang <wangjie125@huawei.com>
Signed-off-by: Jijie Shao <shaojijie@huawei.com>
Link: https://lore.kernel.org/r/20231007031215.1067758-2-shaojijie@huawei.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'skb_segment-testing'

Willem de Bruijn says:

====================
add skb_segment kunit coverage

As discussed at netconf last week. Some kernel code is exercised in
many different ways. skb_segment is a prime example. This ~350 line
function has 49 different patches in git blame with 28 different
authors.

When making a change, e.g., to fix a bug in one specific use case,
it is hard to establish through analysis alone that the change does
not break the many other paths through the code. It is impractical to
exercise all code paths through regression testing from userspace.

Add the minimal infrastructure needed to add KUnit tests to networking,
and add code coverage for this function.

Patch 1 adds the infra and the first simple test case: a linear skb
Patch 2 adds variants with frags[]
Patch 3 adds variants with frag_list skbs
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

net: expand skb_segment unit test with frag_list coverage

Expand the test with these variants that use skb frag_list:

- GSO_TEST_FRAG_LIST:             frag_skb length is gso_size
- GSO_TEST_FRAG_LIST_PURE:        same, data exclusively in frag skbs
- GSO_TEST_FRAG_LIST_NON_UNIFORM: frag_skb length may vary
- GSO_TEST_GSO_BY_FRAGS:          frag_skb length defines gso_size,
                                  i.e., segs may have varying sizes.

Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: parametrize skb_segment unit test to expand coverage

Expand the test with variants

- GSO_TEST_NO_GSO:      payload size less than or equal to gso_size
- GSO_TEST_FRAGS:       payload in both linear and page frags
- GSO_TEST_FRAGS_PURE:  payload exclusively in page frags
- GSO_TEST_GSO_PARTIAL: produce one gso segment of multiple of gso_size,
                        plus optionally one non-gso trailer segment

Define a test struct that encodes the input gso skb and output segs.
Input in terms of linear and fragment lengths. Output as length of
each segment.

Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: add skb_segment kunit test

Add unit testing for skb segment. This function is exercised by many
different code paths, such as GSO_PARTIAL or GSO_BY_FRAGS, linear
(with or without head_frag), frags or frag_list skbs, etc.

It is infeasible to manually run tests that cover all code paths when
making changes. The long and complex function also makes it hard to
establish through analysis alone that a patch has no unintended
side-effects.

Add code coverage through kunit regression testing. Introduce kunit
infrastructure for tests under net/core, and add this first test.

This first skb_segment test exercises a simple case: a linear skb.
Follow-on patches will parametrize the test and add more variants.

Tested: Built and ran the test with

    make ARCH=um mrproper

    ./tools/testing/kunit/kunit.py run \
        --kconfig_add CONFIG_NET=y \
        --kconfig_add CONFIG_DEBUG_KERNEL=y \
        --kconfig_add CONFIG_DEBUG_INFO=y \
        --kconfig_add=CONFIG_DEBUG_INFO_DWARF_TOOLCHAIN_DEFAULT=y \
        net_core_gso

Signed-off-by: Willem de Bruijn <willemb@google.com>
Reviewed-by: Florian Westphal <fw@strlen.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/core: Introduce netdev_core_stats_inc()

Although there is a kfree_skb_reason() helper function that can be used to
find the reason why this skb is dropped, but most callers didn't increase
one of rx_dropped, tx_dropped, rx_nohandler and rx_otherhost_dropped.

For the users, people are more concerned about why the dropped in ip
is increasing.

Introduce netdev_core_stats_inc() for trace the caller of
dev_core_stats_*_inc().

Also, add __code to netdev_core_stats_alloc(), as it's called with small
probability. And add noinline make sure netdev_core_stats_inc was never
inlined.

Signed-off-by: Yajun Deng <yajun.deng@linux.dev>
Suggested-by: Alexander Lobakin <aleksander.lobakin@intel.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'dsa-validate-remove'

Russell King says:

====================
net: dsa: remove validate method

These three patches remove DSA's phylink .validate method which becomes
unnecessary once the last two drivers provide phylink capabilities,
which this patch set adds. Both of these are best guesses.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

net: dsa: remove dsa_port_phylink_validate()

As all drivers now provide phylink capabilities (including MAC), the
if() condition in dsa_port_phylink_validate() will always be true. We
will always use the generic validator, which phylink will call itself
if the .validate method isn't populated. Thus, there is now no need to
implement the .validate method, so this implementation can be removed.

Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: dsa: dsa_loop: add phylink capabilities

Add phylink capabilities for dsa_loop, which I believe being a software
construct means that it supports essentially all interface types and
all speeds.

Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: dsa: vsc73xx: add phylink capabilities

Add phylink capabilities for vsc73xx. Although this switch driver does
populates the .adjust_link method, dsa_slave_phy_setup() will still be
used to create phylink instances for the LAN ports, although phylink
won't be used for shared links.

There are two different classes of switch - 5+1 and 8 port. The 5+1
port switches uses port indicies 0-4 for the user interfaces and 6 for
the CPU port. The 8 port is confusing - some comments in the driver
imply that port index 7 is used, but the driver actually still uses 6,
so that is what we go with. Also, there appear to be no DTs in the
kernel tree that are using the 8 port variety.

It also looks like port 5 is always skipped.

The switch supports 10M, 100M and 1G speeds. It is not clear whether
all these speeds are supported on the CPU interface. It also looks like
symmetric pause is supported, whether asymmetric pause is as well is
unclear. However, it looks like the pause configuration is entirely
static, and doesn't depend on negotiation results.

So, let's do the best effort we can based on the information found in
the driver when creating vsc73xx_phylink_get_caps().

Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

hv_netvsc: fix netvsc_send_completion to avoid multiple message length checks

The switch statement in netvsc_send_completion() is incorrectly validating
the length of incoming network packets by falling through to the next case.
Avoid the fallthrough. Instead break after a case match and then process
the complete() call.
The current code has not caused any known failures. But nonetheless, the
code should be corrected as a different ordering of the switch cases might
cause a length check to fail when it should not.

Signed-off-by: Sonia Sharma <sonia.sharma@linux.microsoft.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'virtio-net-interrupt-moderation'

Heng Qi says:

====================
virtio-net: Fix and update interrupt moderation

The setting of virtio coalescing parameters involves all-queues and
per queue, so we must be careful to synchronize the two.

Regarding napi_tx switching, this patch set is not only
compatible with the previous way of using tx-frames to switch napi_tx,
but also improves the user experience when setting interrupt parameters.

This patch set has been tested and was part of the previous netdim patch
set[1] and is now being split to be rolled out in steps.

[1] https://lore.kernel.org/all/20230811065512.22190-1-hengqi@linux.alibaba.com/

---
v2->v3:
    1. Fix a tiny comment.

v1->v2:
    1. Fix some minor comments and add ack tags.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: a tiny comment update

Update a comment because virtio-net now supports both
VIRTIO_NET_F_NOTF_COAL and VIRTIO_NET_F_VQ_NOTF_COAL.

Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: fix the vq coalescing setting for vq resize

According to the definition of virtqueue coalescing spec[1]:

  Upon disabling and re-enabling a transmit virtqueue, the device MUST set
  the coalescing parameters of the virtqueue to those configured through the
  VIRTIO_NET_CTRL_NOTF_COAL_TX_SET command, or, if the driver did not set
  any TX coalescing parameters, to 0.

  Upon disabling and re-enabling a receive virtqueue, the device MUST set
  the coalescing parameters of the virtqueue to those configured through the
  VIRTIO_NET_CTRL_NOTF_COAL_RX_SET command, or, if the driver did not set
  any RX coalescing parameters, to 0.

We need to add this setting for vq resize (ethtool -G) where vq_reset happens.

[1] https://lists.oasis-open.org/archives/virtio-dev/202303/msg00415.html

Fixes: 394bd87764b6 ("virtio_net: support per queue interrupt coalesce command")
Cc: Gavin Li <gavinl@nvidia.com>
Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: fix per queue coalescing parameter setting

When the user sets a non-zero coalescing parameter to 0 for a specific
virtqueue, it does not work as expected, so let's fix this.

Fixes: 394bd87764b6 ("virtio_net: support per queue interrupt coalesce command")
Reported-by: Xiaoming Zhao <zxm377917@alibaba-inc.com>
Cc: Gavin Li <gavinl@nvidia.com>
Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: consistently save parameters for per-queue

When using .set_coalesce interface to set all queue coalescing
parameters, we need to update both per-queue and global save values.

Fixes: 394bd87764b6 ("virtio_net: support per queue interrupt coalesce command")
Cc: Gavin Li <gavinl@nvidia.com>
Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: fix mismatch of getting tx-frames

Since virtio-net allows switching napi_tx for per txq, we have to
get the specific txq's result now.

Fixes: 394bd87764b6 ("virtio_net: support per queue interrupt coalesce command")
Cc: Gavin Li <gavinl@nvidia.com>
Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

virtio-net: initially change the value of tx-frames

Background:
1. Commit 0c465be183c7 ("virtio_net: ethtool tx napi configuration") uses
   tx-frames to toggle napi_tx (0 off and 1 on) if notification coalescing
   is not supported.
2. Commit 31c03aef9bc2 ("virtio_net: enable napi_tx by default") enables
   napi_tx for all txqs by default.

Status:
When virtio-net supports notification coalescing, after initialization,
tx-frames is 0 and napi_tx is true.

Problem:
When the user only wants to set rx coalescing params using
           ethtool -C eth0 rx-usecs 10, or
   ethtool -Q eth0 queue_mask 0x1 -C rx-usecs 10,
these cmds will carry tx-frames as 0, causing the napi_tx switching condition
is satisfied. Then the user gets:
           netlink error: Device or resource busy.

The same happens when trying to set rx-frames, adaptive_rx, adaptive_tx...

How to fix:
When notification coalescing feature is negotiated, initially make the
value of tx-frames to be consistent with napi_tx.

For compatibility with the past, it is still supported to use tx-frames
to toggle napi_tx.

Reported-by: Xiaoming Zhao <zxm377917@alibaba-inc.com>
Signed-off-by: Heng Qi <hengqi@linux.alibaba.com>
Acked-by: Jason Wang <jasowang@redhat.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: dsa: realtek: rtl8365mb: replace deprecated strncpy with ethtool_sprintf

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

ethtool_sprintf() is designed specifically for get_strings() usage.
Let's replace strncpy in favor of this more robust and easier to
understand interface.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Alvin Šipraga <alsi@bang-olufsen.dk>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-dsa-realtek-rtl8365mb-c-v1-1-0537fe9fb08c@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: dsa: realtek: replace deprecated strncpy with ethtool_sprintf

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

ethtool_sprintf() is designed specifically for get_strings() usage.
Let's replace strncpy in favor of this more robust and easier to
understand interface.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Linus Walleij <linus.walleij@linaro.org>
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-dsa-realtek-rtl8366-core-c-v1-1-74e1b5190778@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: dsa: qca8k: replace deprecated strncpy with ethtool_sprintf

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

ethtool_sprintf() is designed specifically for get_strings() usage.
Let's replace strncpy in favor of this more robust and easier to
understand interface.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20231009-strncpy-drivers-net-dsa-qca-qca8k-common-c-v1-1-34c8040e0f32@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

appletalk: remove ipddp driver

After the cops driver is removed, ipddp is now the only
CONFIG_DEV_APPLETALK but as far as I can tell, this also has no users
and can be removed, making appletalk support purely based on ethertalk,
using ethernet hardware.

Link: https://lore.kernel.org/netdev/e490dd0c-a65d-4acf-89c6-c06cb48ec880@app.fastmail.com/
Link: https://lore.kernel.org/netdev/9cac4fbd-9557-b0b8-54fa-93f0290a6fb8@schmorgal.com/
Cc: Doug Brown <doug@schmorgal.com>
Signed-off-by: Arnd Bergmann <arnd@arndb.de>
Acked-by: Greg Kroah-Hartman <gregkh@linuxfoundation.org>
Link: https://lore.kernel.org/r/20231009141139.1766345-1-arnd@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

atm: fore200e: Drop unnecessary of_match_device()

It is not necessary to call of_match_device() in probe. If we made it to
probe, then we've already successfully matched.

Signed-off-by: Rob Herring <robh@kernel.org>
Reviewed-by: Simon Horman <horms@kernel.org>
Link: https://lore.kernel.org/r/20231006214421.339445-1-robh@kernel.org
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Merge branch 'mlxsw-fix-wformat-truncation-warnings'

Petr Machata says:

====================
mlxsw: Fix -Wformat-truncation warnings

Ido Schimmel writes:

Commit 6d4ab2e97dcf ("extrawarn: enable format and stringop overflow
warnings in W=1") enabled format warnings as part of W=1 builds,
resulting in two new warnings in mlxsw. Fix both and target at net-next
as the warnings are not indicative of actual bugs.
====================

Link: https://lore.kernel.org/r/cover.1696600763.git.petrm@nvidia.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

mlxsw: spectrum_ethtool: Fix -Wformat-truncation warning

Ethtool stats strings cannot be longer than 32 characters
('ETH_GSTRING_LEN'), including the terminating null byte. The format
string '%.29s_%.1d' can exceed this limitation if the per-TC counter
name exceeds 28 characters. Together with the underscore, the two digits
of the TC (bounded at 16) and the terminating null byte, more than 32
characters will be used.

Fix this by bounding the counter name at 28 characters which suppresses
the following build warning [1]. This does not affect ethtool output
since the longest counter name does not exceed this limitation.

[1]
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c: In function ‘mlxsw_sp_port_get_strings’:
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:622:58: error: ‘snprintf’ output may be truncated before the last format character [-Werror=format-truncation=]
  622 |                 snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d",
      |                                                          ^
In function ‘mlxsw_sp_port_get_tc_strings’,
    inlined from ‘mlxsw_sp_port_get_strings’ at drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:677:4:
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:622:17: note: ‘snprintf’ output between 3 and 33 bytes into a destination of size 32
  622 |                 snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d",
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  623 |                          mlxsw_sp_port_hw_tc_stats[i].str, tc);
      |                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c: In function ‘mlxsw_sp_port_get_strings’:
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:622:58: error: ‘snprintf’ output may be truncated before the last format character [-Werror=format-truncation=]
  622 |                 snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d",
      |                                                          ^
In function ‘mlxsw_sp_port_get_tc_strings’,
    inlined from ‘mlxsw_sp_port_get_strings’ at drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:677:4:
drivers/net/ethernet/mellanox/mlxsw/spectrum_ethtool.c:622:17: note: ‘snprintf’ output between 3 and 33 bytes into a destination of size 32
  622 |                 snprintf(*p, ETH_GSTRING_LEN, "%.29s_%.1d",
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  623 |                          mlxsw_sp_port_hw_tc_stats[i].str, tc);
      |                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Simon Horman <horms@kernel.org> # build-tested
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

mlxsw: core_thermal: Fix -Wformat-truncation warning

The name of a thermal zone device cannot be longer than 19 characters
('THERMAL_NAME_LENGTH - 1'). The format string 'mlxsw-lc%d-gearbox%d'
can exceed this limitation if the maximum number of line cards and the
maximum number of gearboxes on each line card cannot be represented
using a single digit.

This is not the case with current systems nor future ones. Therefore,
increase the size of the result buffer beyond 'THERMAL_NAME_LENGTH' and
suppress the following build warning [1].

If this limitation is ever exceeded, we will know about it since the
thermal core validates the thermal device's name during registration.

[1]
drivers/net/ethernet/mellanox/mlxsw/core_thermal.c: In function ‘mlxsw_thermal_gearboxes_init.constprop’:
drivers/net/ethernet/mellanox/mlxsw/core_thermal.c:543:71: error: ‘%d’ directive output may be truncated writing between 1 and 3 bytes into a region of size between 1 and 3
[-Werror=format-truncation=]
  543 |                 snprintf(tz_name, sizeof(tz_name), "mlxsw-lc%d-gearbox%d",
      |                                                                       ^~
In function ‘mlxsw_thermal_gearbox_tz_init’,
    inlined from ‘mlxsw_thermal_gearboxes_init.constprop’ at drivers/net/ethernet/mellanox/mlxsw/core_thermal.c:611:9:
drivers/net/ethernet/mellanox/mlxsw/core_thermal.c:543:52: note: directive argument in the range [1, 255]
  543 |                 snprintf(tz_name, sizeof(tz_name), "mlxsw-lc%d-gearbox%d",
      |                                                    ^~~~~~~~~~~~~~~~~~~~~~
drivers/net/ethernet/mellanox/mlxsw/core_thermal.c:543:17: note: ‘snprintf’ output between 19 and 23 bytes into a destination of size 20
  543 |                 snprintf(tz_name, sizeof(tz_name), "mlxsw-lc%d-gearbox%d",
      |                 ^~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~
  544 |                          gearbox_tz->slot_index, gearbox_tz->module + 1);
      |                          ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Signed-off-by: Ido Schimmel <idosch@nvidia.com>
Reviewed-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Reviewed-by: Simon Horman <horms@kernel.org>
Tested-by: Simon Horman <horms@kernel.org> # build-tested
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

net: dsa: microchip: Fix uninitialized var in ksz9477_acl_move_entries()

Address an issue in ksz9477_acl_move_entries() where, in the scenario
(src_idx == dst_idx), ksz9477_validate_and_get_src_count() returns 0,
leading to usage of uninitialized src_count and dst_count variables,
which causes undesired behavior as it attempts to move ACL entries
around.

Fixes: 002841be134e ("net: dsa: microchip: Add partial ACL support for ksz9477 switches")
Reported-by: Dan Carpenter <dan.carpenter@linaro.org>
Suggested-by: Vladimir Oltean <olteanv@gmail.com>
Signed-off-by: Oleksij Rempel <o.rempel@pengutronix.de>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20231006115822.144152-1-o.rempel@pengutronix.de
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

Merge branch 'tcp-save-flowlabel-and-use-for-receiver-repathing'

David Morley says:

====================
tcp: save flowlabel and use for receiver repathing

This patch series stores the last received ipv6 flowlabel. This last
received flowlabel is then used to help decide whether a packet is
likely an RTO retransmit or the result of a TLP. This new information
is used to better inform the flowlabel change decision for data
receivers.
====================

Link: https://lore.kernel.org/r/20231006011841.3558307-1-morleyd.kernel@gmail.com
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

tcp: change data receiver flowlabel after one dup

This commit changes the data receiver repath behavior to occur after
receiving a single duplicate. This can help recover ACK connectivity
quicker if a TLP was sent along a nonworking path.

For instance, consider the case where we have an initially nonworking
forward path and reverse path and subsequently switch to only working
forward paths. Before this patch we would have the following behavior.

+---------+--------+--------+----------+----------+----------+
| Event   | For FL | Rev FL | FP Works | RP Works | Data Del |
+---------+--------+--------+----------+----------+----------+
| Initial | A      | 1      | N        | N        | 0        |
+---------+--------+--------+----------+----------+----------+
| TLP     | A      | 1      | N        | N        | 0        |
+---------+--------+--------+----------+----------+----------+
| RTO 1   | B      | 1      | Y        | N        | 1        |
+---------+--------+--------+----------+----------+----------+
| RTO 2   | C      | 1      | Y        | N        | 2        |
+---------+--------+--------+----------+----------+----------+
| RTO 3   | D      | 2      | Y        | Y        | 3        |
+---------+--------+--------+----------+----------+----------+

This patch gets rid of at least RTO 3, avoiding additional unnecessary
repaths of a working forward path to a (potentially) nonworking one.

In addition, this commit changes the behavior to avoid repathing upon
rx of duplicate data if the local endpoint is in CA_Loss (in which
case the RTOs will already be changing the outgoing flowlabel).

Signed-off-by: David Morley <morleyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Tested-by: David Morley <morleyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

tcp: record last received ipv6 flowlabel

In order to better estimate whether a data packet has been
retransmitted or is the result of a TLP, we save the last received
ipv6 flowlabel.

To make space for this field we resize the "ato" field in
inet_connection_sock as the current value of TCP_DELACK_MAX can be
fully contained in 8 bits and add a compile_time_assert ensuring this
field is the required size.

v2: addressed kernel bot feedback about dccp_delack_timer()
v3: addressed build error introduced by commit bbf80d713fe7 ("tcp:
derive delack_max from rto_min")

Signed-off-by: David Morley <morleyd@google.com>
Signed-off-by: Neal Cardwell <ncardwell@google.com>
Signed-off-by: Yuchung Cheng <ycheng@google.com>
Tested-by: David Morley <morleyd@google.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Paolo Abeni <pabeni@redhat.com>

tools: ynl-gen: handle do ops with no input attrs

The code supports dumps with no input attributes currently
thru a combination of special-casing and luck.
Clean up the handling of ops with no inputs. Create empty
Structs, and skip printing of empty types.
This makes dos with no inputs work.

Tested-by: Lorenzo Bianconi <lorenzo.bianconi@redhat.com>
Link: https://lore.kernel.org/r/20231006135032.3328523-1-kuba@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'tools-ynl-gen-lift-type-requirement-for-attribute-subsets'

Jiri Pirko says:

====================
tools: ynl-gen: lift type requirement for attribute subsets

Remove the requirement from schema files to specify the "type" for
attribute subsets and adjust existing schema files.
====================

Link: https://lore.kernel.org/r/20231006114436.1725425-1-jiri@resnulli.us
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netlink: specs: remove redundant type keys from attributes in subsets

No longer needed to define type for subset attributes. Remove those.

Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231006114436.1725425-3-jiri@resnulli.us
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

tools: ynl-gen: lift type requirement for attribute subsets

In case an attribute is used in a subset, the type has to be currently
specified. As the attribute is already defined in the original set, this
is a redundant information in yaml file, moreover, may lead to
inconsistencies.

Example:
attribute-sets:
    ...
    name: pin
    enum-name: dpll_a_pin
    attributes:
      ...
      -
        name: parent-id
        type: u32
      ...
  -
    name: pin-parent-device
    subset-of: pin
    attributes:
      -
        name: parent-id
        type: u32             <<<<<<<<<<<<<<<<<<<<<<<<<<<<<<<

Remove the requirement from schema files to specify the "type" for
attribute subsets.

Signed-off-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jakub Kicinski <kuba@kernel.org>
Link: https://lore.kernel.org/r/20231006114436.1725425-2-jiri@resnulli.us
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

cavium/liquidio: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

We expect `app_name` to be NUL-terminated:
dev_info(&oct->pci_dev->dev,
 "Running %s (%llu Hz)\n",
 app_name, CVM_CAST64(cs->corefreq));
... and it seems NUL-padding is not required, let's opt for strscpy().

For `oct->boardinfo.name/serial_number` let's opt for strscpy() as well
since it is expected to be NUL-terminated and does not require
NUL-padding as `oct` is zero-initialized in octeon_device.c +707:
|       buf = vzalloc(size);
|       if (!buf)
|        return NULL;
|
|       oct = (struct octeon_device *)buf;

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-cavium-liquidio-octeon_device-c-v1-1-9a207cef9438@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

liquidio: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

NUL-padding is not required as rep_cfg is memset to 0:
|       memset(&rep_cfg, 0, sizeof(rep_cfg));

A suitable replacement is `strscpy` [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-cavium-liquidio-lio_vf_rep-c-v1-1-92123a747780@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: liquidio: replace deprecated strncpy with strscpy_pad

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

We know `fw_type` must be NUL-terminated based on use here:
|       static bool fw_type_is_auto(void)
|       {
|        return strncmp(fw_type, LIO_FW_NAME_TYPE_AUTO,
|               sizeof(LIO_FW_NAME_TYPE_AUTO)) == 0;
|       }
...and here
|       module_param_string(fw_type, fw_type, sizeof(fw_type), 0444);

Let's opt to NUL-pad the destination buffer as well so that we maintain
the same exact behavior that `strncpy` provided here.

A suitable replacement is `strscpy_pad` due to the fact that it
guarantees both NUL-termination and NUL-padding on the destination
buffer.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-cavium-liquidio-lio_main-c-v1-1-663e3f1d8f99@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

liquidio: replace deprecated strncpy/strcpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

NUL-padding is not required as drvinfo is memset to 0:
| memset(drvinfo, 0, sizeof(struct ethtool_drvinfo));

A suitable replacement is `strscpy` [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-cavium-liquidio-lio_ethtool-c-v1-1-ab565ab4d197@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

bcm63xx_enet: replace deprecated strncpy with strscpy

strncpy() is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

A suitable replacement is strscpy() [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.

bcm_enet_get_drvinfo() already uses strscpy(), let's match it's
implementation:
|       static void bcm_enet_get_drvinfo(struct net_device *netdev,
|         struct ethtool_drvinfo *drvinfo)
|       {
|        strscpy(drvinfo->driver, bcm_enet_driver_name, sizeof(drvinfo->driver));
|        strscpy(drvinfo->bus_info, "bcm63xx", sizeof(drvinfo->bus_info));
|       }

Note that now bcm_enet_get_drvinfo() and bcm_enetsw_get_drvinfo() do the
exact same thing.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-broadcom-bcm63xx_enet-c-v1-1-6823b3c3c443@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: bcmgenet: Remove custom ndo_poll_controller()

The driver gained a .ndo_poll_controller() at a time where the TX
cleaning process was always done from NAPI which makes this unnecessary.
See commit ac3d9dd034e5 ("netpoll: make ndo_poll_controller() optional")
for more background.

Signed-off-by: Florian Fainelli <florian.fainelli@broadcom.com>
Reviewed-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

net: sock_dequeue_err_skb() optimization

Exit early if the list is empty.

Some applications using TCP zerocopy are calling
recvmsg( ... MSG_ERRQUEUE) and hit this case quite often,
probably because busy polling only deals with sk_receive_queue.

Signed-off-by: Eric Dumazet <edumazet@google.com>
Link: https://lore.kernel.org/r/20231005114504.642589-1-edumazet@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge tag 'wireless-next-2023-10-06' of git://git./linux/kernel/git/wireless/wireless-next

Kalle Valo says:

====================
wireless-next patches for v6.7

The first pull request for v6.7, with both stack and driver changes.
We have a big change how locking is handled in cfg80211 and mac80211
which removes several locks and hopefully simplifies the locking
overall. In drivers rtw89 got MCC support and smaller features to
other active drivers but nothing out of ordinary.

Major changes:

cfg80211
 - remove wdev mutex, use the wiphy mutex instead
 - annotate iftype_data pointer with sparse
 - first kunit tests, for element defrag
 - remove unused scan_width support

mac80211
 - major locking rework, remove several locks like sta_mtx, key_mtx
   etc. and use the wiphy mutex instead
 - remove unused shifted rate support
 - support antenna control in frame injection (requires driver support)
 - convert RX_DROP_UNUSABLE to more detailed reason codes

rtw89
 - TDMA-based multi-channel concurrency (MCC) support

iwlwifi
 - support set_antenna() operation
 - support frame injection antenna control

ath12k
 - WCN7850: enable 320 MHz channels in 6 GHz band
 - WCN7850: hardware rfkill support
 - WCN7850: enable IEEE80211_HW_SINGLE_SCAN_ON_ALL_BANDS to make scan faster

ath11k
 - add chip id board name while searching board-2.bin

* tag 'wireless-next-2023-10-06' of git://git.kernel.org/pub/scm/linux/kernel/git/wireless/wireless-next: (272 commits)
  wifi: rtlwifi: remove unreachable code in rtl92d_dm_check_edca_turbo()
  wifi: rtw89: debug: txpwr table supports Wi-Fi 7 chips
  wifi: rtw89: debug: show txpwr table according to chip gen
  wifi: rtw89: phy: set TX power RU limit according to chip gen
  wifi: rtw89: phy: set TX power limit according to chip gen
  wifi: rtw89: phy: set TX power offset according to chip gen
  wifi: rtw89: phy: set TX power by rate according to chip gen
  wifi: rtw89: mac: get TX power control register according to chip gen
  wifi: rtlwifi: use unsigned long for rtl_bssid_entry timestamp
  wifi: rtlwifi: fix EDCA limit set by BT coexistence
  wifi: rt2x00: fix MT7620 low RSSI issue
  wifi: rtw89: refine bandwidth 160MHz uplink OFDMA performance
  wifi: rtw89: refine uplink trigger based control mechanism
  wifi: rtw89: 8851b: update TX power tables to R34
  wifi: rtw89: 8852b: update TX power tables to R35
  wifi: rtw89: 8852c: update TX power tables to R67
  wifi: rtw89: regd: configure Thailand in regulation type
  wifi: mac80211: add back SPDX identifier
  wifi: mac80211: fix ieee80211_drop_unencrypted_mgmt return type/value
  wifi: rtlwifi: cleanup few rtlxxxx_set_hw_reg() routines
  ...

====================

Link: https://lore.kernel.org/r/87jzrz6bvw.fsf@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: phy: broadcom: add support for BCM5221 phy

This patch adds the BCM5221 PHY support by reusing brcm_fet_*()
callbacks and adding quirks for BCM5221 when needed.

Cc: Jim Reinhart <jimr@tekvox.com>
Cc: James Autry <jautry@tekvox.com>
Cc: Matthew Maron <matthewm@tekvox.com>
Signed-off-by: Giulio Benetti <giulio.benetti+tekvox@benettiengineering.com>
Reviewed-by: Florian Fainelli <florian.fainelli@broadcom.com>
Link: https://lore.kernel.org/r/20231005182915.153815-1-giulio.benetti@benettiengineering.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch '40GbE' of git://git./linux/kernel/git/tnguy/next-queue

Tony Nguyen says:

====================
i40e: House-keeping and clean-up

Ivan Vecera says:

The series makes some house-keeping tasks on i40e driver:

Patch 1: Removes unnecessary back pointer from i40e_hw
Patch 2: Moves I40E_MASK macro to i40e_register.h where is used
Patch 3: Refactors I40E_MDIO_CLAUSE* to use the common macro
Patch 4: Add header dependencies to <linux/avf/virtchnl.h>
Patch 5: Simplifies memory alloction functions
Patch 6: Moves mem alloc structures to i40e_alloc.h
Patch 7: Splits i40e_osdep.h to i40e_debug.h and i40e_io.h
Patch 8: Removes circular header deps, fixes and cleans headers
Patch 9: Moves DDP specific macros and structs to i40e_ddp.c

* '40GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue:
  i40e: Move DDP specific macros and structures to i40e_ddp.c
  i40e: Remove circular header dependencies and fix headers
  i40e: Split i40e_osdep.h
  i40e: Move memory allocation structures to i40e_alloc.h
  i40e: Simplify memory allocation functions
  virtchnl: Add header dependencies
  i40e: Refactor I40E_MDIO_CLAUSE* macros
  i40e: Move I40E_MASK macro to i40e_register.h
  i40e: Remove back pointer from i40e_hw structure
====================

Link: https://lore.kernel.org/r/20231005162850.3218594-1-anthony.l.nguyen@intel.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: atheros: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

We expect netdev->name to be NUL-terminated based on its use with format
strings and dev_info():
|     dev_info(&adapter->pdev->dev,
|             "%s link is up %d Mbps %s\n",
|             netdev->name, adapter->link_speed,
|             adapter->link_duplex == FULL_DUPLEX ?
|             "full duplex" : "half duplex");

Furthermore, NUL-padding is not required as netdev is already
zero-initialized through alloc_etherdev().

Considering the above, a suitable replacement is `strscpy` [2] due to
the fact that it guarantees NUL-termination on the destination buffer
without unnecessarily NUL-padding.

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-atheros-atlx-atl2-c-v1-1-493f113ebfc7@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ax88796c: replace deprecated strncpy with strscpy

`strncpy` is deprecated for use on NUL-terminated destination strings
[1] and as such we should prefer more robust and less ambiguous string
interfaces.

A suitable replacement is `strscpy` [2] due to the fact that it
guarantees NUL-termination on the destination buffer without
unnecessarily NUL-padding.

It should be noted that there doesn't currently exist a bug here as
DRV_NAME is a small string literal which means no overread bugs are
present.

Also to note, other ethernet drivers are using strscpy in a similar
pattern:
|       dec/tulip/tulip_core.c
|       861:    strscpy(info->driver, DRV_NAME, sizeof(info->driver));
|
|       8390/ax88796.c
|       582:    strscpy(info->driver, DRV_NAME, sizeof(info->driver));
|
|       dec/tulip/dmfe.c
|       1077:   strscpy(info->driver, DRV_NAME, sizeof(info->driver));
|
|       8390/etherh.c
|       558:    strscpy(info->driver, DRV_NAME, sizeof(info->driver));

Link: https://www.kernel.org/doc/html/latest/process/deprecated.html#strncpy-on-nul-terminated-strings
Link: https://manpages.debian.org/testing/linux-manual-4.8/strscpy.9.en.html
Link: https://github.com/KSPP/linux/issues/90
Signed-off-by: Justin Stitt <justinstitt@google.com>
Reviewed-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Lukasz Stelmach <l.stelmach@samsung.com>
Link: https://lore.kernel.org/r/20231005-strncpy-drivers-net-ethernet-asix-ax88796c_ioctl-c-v1-1-6fafdc38b170@google.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: ixp4xx_eth: Support changing the MTU

As we don't specify the MTU in the driver, the framework
will fall back to 1500 bytes and this doesn't work very
well when we try to attach a DSA switch:

  eth1: mtu greater than device maximum
  ixp4xx_eth c800a000.ethernet eth1: error -22 setting
  MTU to 1504 to include DSA overhead

After locating an out-of-tree patch in OpenWrt I found
suitable code to set the MTU on the interface and ported
it and updated it. Now the MTU gets set properly.

Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Linus Walleij <linus.walleij@linaro.org>
Link: https://lore.kernel.org/r/20231005-ixp4xx-eth-mtu-v4-1-08c66ed0bc69@linaro.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

MAINTAINERS: Update LL TEMAC entry to Orphan

Since there's no alternate driver, change this entry from obsolete
to orphan.

Signed-off-by: Harini Katakam <harini.katakam@amd.com>
Link: https://lore.kernel.org/r/20231005131039.25881-1-harini.katakam@amd.com
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge tag 'linux-can-next-for-6.7-20231005' of git://git./linux/kernel/git/mkl/linux-can-next

Marc Kleine-Budde says:

====================
pull-request: can-next 2023-10-05

The first patch is by Miquel Raynal and fixes a comment in the sja1000
driver.

Vincent Mailhol contributes 2 patches that fix W=1 compiler warnings
in the etas_es58x driver.

Jiapeng Chong's patch removes an unneeded NULL pointer check before
dev_put() in the CAN raw protocol.

A patch by Justin Stittreplaces a strncpy() by strscpy() in the
peak_pci sja1000 driver.

The next 5 patches are by me and fix the can_restart() handler and
replace BUG_ON()s in the CAN dev helpers with proper error handling.

The last 27 patches are also by me and target the at91_can driver.
First a new helper function is introduced, the at91_can driver is
cleaned up and updated to use the rx-offload helper.

* tag 'linux-can-next-for-6.7-20231005' of git://git.kernel.org/pub/scm/linux/kernel/git/mkl/linux-can-next: (37 commits)
  can: at91_can: switch to rx-offload implementation
  can: at91_can: at91_alloc_can_err_skb() introduce new function
  can: at91_can: at91_irq_err_line(): send error counters with state change
  can: at91_can: at91_irq_err_line(): make use of can_change_state() and can_bus_off()
  can: at91_can: at91_irq_err_line(): take reg_sr into account for bus off
  can: at91_can: at91_irq_err_line(): make use of can_state_get_by_berr_counter()
  can: at91_can: at91_irq_err(): rename to at91_irq_err_line()
  can: at91_can: at91_irq_err_frame(): move next to at91_irq_err()
  can: at91_can: at91_irq_err_frame(): call directly from IRQ handler
  can: at91_can: at91_poll_err(): increase stats even if no quota left or OOM
  can: at91_can: at91_poll_err(): fold in at91_poll_err_frame()
  can: at91_can: add CAN transceiver support
  can: at91_can: at91_open(): forward request_irq()'s return value in case or an error
  can: at91_can: at91_chip_start(): don't disable IRQs twice
  can: at91_can: at91_set_bittiming(): demote register output to debug level
  can: at91_can: rename struct at91_priv::{tx_next,tx_echo} to {tx_head,tx_tail}
  can: at91_can: at91_setup_mailboxes(): update comments
  can: at91_can: add more register definitions
  can: at91_can: MCR Register: convert to FIELD_PREP()
  can: at91_can: MSR Register: convert to FIELD_PREP()
  ...
====================

Link: https://lore.kernel.org/r/20231005195812.549776-1-mkl@pengutronix.de
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge wireless into wireless-next

Resolve several conflicts, mostly between changes/fixes in
wireless and the locking rework in wireless-next. One of
the conflicts actually shows a bug in wireless that we'll
want to fix separately.

Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Kalle Valo <kvalo@kernel.org>

net: phy: dp83867: Add support for hardware blinking LEDs

This implements the led_hw_* hooks to support hardware blinking LEDs on
the DP83867 phy. The driver supports all LED modes that have a
corresponding TRIGGER_NETDEV_* define. Error and collision do not have
a TRIGGER_NETDEV_* define, so these modes are currently not supported.

Reviewed-by: Andrew Lunn <andrew@lunn.ch>
Tested-by: Alexander Stein <alexander.stein@ew.tq-group.com> #TQMa8MxML/MBa8Mx
Signed-off-by: Sascha Hauer <s.hauer@pengutronix.de>
Signed-off-by: David S. Miller <davem@davemloft.net>

flow_offload: Annotate struct flow_action_entry with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct flow_action_entry.

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: netdev@vger.kernel.org
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

net/packet: Annotate struct packet_fanout with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct packet_fanout.

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Willem de Bruijn <willemb@google.com>
Cc: Anqi Shen <amy.saq@antgroup.com>
Cc: netdev@vger.kernel.org
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

tools: ynl-gen: use uapi header name for the header guard

Chuck points out that we should use the uapi-header property
when generating the guard. Otherwise we may generate the same
guard as another file in the tree.

Tested-by: Chuck Lever <chuck.lever@oracle.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Przemek Kitszel <przemyslaw.kitszel@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'mlxsw-ACL-region'

Petr Machata says:

====================
mlxsw: Control the order of blocks in ACL region

Amit Cohen writes:

For 12 key blocks in the A-TCAM, rules are split into two records, which
constitute two lookups. The two records are linked using a
"large entry key ID".

Due to a Spectrum-4 hardware issue, KVD entries that correspond to key
blocks 0 to 5 of 12 key blocks will be placed in the same KVD pipe if they
only differ in their "large entry key ID", as it is ignored. This results
in a reduced scale, we can insert less than 20k filters and get an error:

    $ tc -b flower.batch
    RTNETLINK answers: Input/output error
    We have an error talking to the kernel

To reduce the probability of this issue, we can place key blocks with
high entropy in blocks 0 to 5. The idea is to place blocks that are often
changed in blocks 0 to 5, for example, key blocks that match on IPv4
addresses or the LSBs of IPv6 addresses. Such placement will reduce the
probability of these blocks to be same.

Mark several blocks with 'high_entropy' flag and place them in blocks 0
to 5. Note that the list of the blocks is just a suggestion, I will verify
it with architects.

Currently, there is a one loop that chooses which blocks should be used
for a given list of elements and fills the blocks - when a block is
chosen, it fills it in the region. To be able to control the order of
the blocks, separate between searching blocks and filling them. Several
pre-changes are required.

Patch set overview:
Patch #1 marks several blocks with 'high_entropy' flag.
Patches #2-#4 prepare the code for filling blocks at the end of the search.
Patch #5 changes the loop to just choose the blocks and fill the blocks at
the end.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

mlxsw: core_acl_flex_keys: Fill blocks with high entropy first

The previous patches prepared the code to allow separating between
choosing blocks and filling blocks.

Do not add blocks as part of the loop that chooses them. When all the
required blocks are set in the bitmap 'chosen_blocks_bm', start filling
blocks. Iterate over the bitmap twice - first add only blocks that are
marked with 'high_entropy' flag. Then, fill the rest of the blocks.

The idea is to place key blocks with high entropy in blocks 0 to 5. See
more details in previous patches.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mlxsw: core_acl_flex_keys: Save chosen elements in all blocks per search

Currently, mlxsw_afk_picker() chooses which blocks will be used for a
given list of elements, and fills the blocks during the searching - when a
key block is found with most hits, it adds it and removes the elements from
the count of hits. This should be changed as we want to be able to choose
which blocks will be placed in blocks 0 to 5.

To separate between choosing blocks and filling blocks, several pre-changes
are required. Currently, the indication of whether all elements were
found in the chosen blocks is by the structure 'key_info->elusage'. This
structure is updated when block is filled as part of
mlxsw_afk_picker_key_info_add(). A following patch will call this
function only after choosing all the blocks. Add a bitmap called
'elusage_chosen' to store which elements were chosen in the chosen blocks.
Change the condition in the loop to check elements that were chosen, not
elements that were already filled in the blocks.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mlxsw: core_acl_flex_keys: Save chosen elements per block

Currently, mlxsw_afk_picker() chooses which blocks will be used for a
given list of elements, and fills the blocks during the searching - when a
key block is found with most hits, it adds it and removes the elements from
the count of hits. This should be changed as we want to be able to choose
which blocks will be placed in blocks 0 to 5.

To separate between choosing blocks and filling blocks, several pre-changes
are required. During the search, the structure 'mlxsw_afk_picker' is
used per block, it contains how many elements from the required list appear
in the block. When a block is chosen and filled, this bitmap of elements is
cleaned. To be able to fill the blocks at the end, add a bitmap called
'chosen_element' as part of picker. When a block is chosen, copy the
'element' bitmap to it. Use the new bitmap as part of
mlxsw_afk_picker_key_info_add(). So later, when filling the block will
be done at the end of the searching, we will use the copied bitmap that
contains the elements that should be used in the block.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mlxsw: core_acl_flex_keys: Add a bitmap to save which blocks are chosen

Currently, mlxsw_afk_picker() chooses which blocks will be used for a
given list of elements, and fills the blocks during the searching - when a
key block is found with most hits, it adds it and removes the elements from
the count of hits. This should be changed as we want to be able to choose
which blocks will be placed in blocks 0 to 5.

To separate between choosing blocks and filling blocks, several pre-changes
are required. The indexes of the chosen blocks should be saved, so then
the relevant blocks will be filled at the end of search.

Allocate a bitmap for chosen blocks, when a block is found with most
hits, set the relevant bit in the bitmap. This bitmap will be used in a
following patch.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

mlxsw: Mark high entropy key blocks

For 12 key blocks in the A-TCAM, rules are split into two records, which
constitute two lookups. The two records are linked using a
"large entry key ID".

Due to a Spectrum-4 hardware issue, KVD entries that correspond to key
blocks 0 to 5 of 12 key blocks A-TCAM entries will be placed in the same
KVD pipe if they only differ in their "large entry key ID", as it is
ignored. This results in a reduced scale. To reduce the probability of this
issue, we can place key blocks with high entropy in blocks 0 to 5. The idea
is to place blocks that are changed often in blocks 0 to 5, for
example, key blocks that match on IPv4 addresses or the LSBs of IPv6
addresses. Such placement will reduce the probability of these blocks to be
same.

Mark several blocks with 'high_entropy' flag, so later we will take into
account this flag and place them in blocks 0 to 5.

Signed-off-by: Amit Cohen <amcohen@nvidia.com>
Reviewed-by: Ido Schimmel <idosch@nvidia.com>
Signed-off-by: Petr Machata <petrm@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Merge branch 'sfc-conntrack-offloads'

Edward Cree says:

====================
sfc: conntrack offload for tunnels

This series adds support for offloading TC flower rules which require
both connection tracking and tunnel decapsulation.  Depending on the
match keys required, the left-hand-side rule may go in either the
Outer Rule table or the Action Rule table.
====================

Signed-off-by: David S. Miller <davem@davemloft.net>

sfc: support TC rules which require OR-AR-CT-AR flow

When a foreign LHS rule (TC rule from a tunnel netdev which requests
 conntrack lookup) matches on inner headers or enc_key_id, these matches
 cannot be performed by the Outer Rule table, as the keys are only
 available after the tunnel type has been identified (by the OR lookup)
 and the rest of the headers parsed accordingly.
Offload such rules with an Action Rule, using the LOOKUP_CONTROL section
 of the AR response to specify the conntrack and/or recirculation actions,
 combined with an Outer Rule which performs only the usual Encap Match
 duties.
This processing flow, as it requires two AR lookups per packet, is less
 performant than OR-CT-AR, so only use it where necessary.

Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-vuuren@amd.com>
Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sfc: ensure an extack msg from efx_tc_flower_replace_foreign EOPNOTSUPPs

There were a few places where no extack error message was set, or the
 extack was not forwarded to callees, potentially resulting in a return
 of -EOPNOTSUPP with no additional information.
Make sure to populate the error message in these cases.  In practice
 this does us no good as TC indirect block callbacks don't come with an
 extack to fill in; but maybe they will someday and when debugging it's
 possible to provide a fake extack and emit its message to the console.

Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-vuuren@amd.com>
Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sfc: offload foreign RHS rules without an encap match

Normally, if a TC filter on a tunnel netdev does not match on any
 encap fields, we decline to offload it, as it cannot meet our
 requirement for a <sip,dip,dport> tuple for the encap match.
However, if the rule has a nonzero chain_index, then for a packet to
 reach the rule, it must already have matched a LHS rule which will
 have included an encap match and determined the tunnel type, so in
 that case we can offload the right-hand-side rule.

Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-vuuren@amd.com>
Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

sfc: support TC left-hand-side rules on foreign netdevs

Allow a tunnel netdevice (such as a vxlan) to offload conntrack lookups,
 in much the same way as efx netdevs.
To ensure this rule does not overlap with other tunnel rules on the same
 sip,dip,dport tuple, register a pseudo encap match of a new type
 (EFX_TC_EM_PSEUDO_OR), which unlike PSEUDO_MASK may only be referenced
 once (because an actual Outer Rule in hardware exists, although its
 fw_id is not recorded in the encap match entry).

Reviewed-by: Pieter Jansen van Vuuren <pieter.jansen-van-vuuren@amd.com>
Signed-off-by: Edward Cree <ecree.xilinx@gmail.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

nexthop: Annotate struct nh_group with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nh_group.

Cc: David Ahern <dsahern@kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: netdev@vger.kernel.org
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

nexthop: Annotate struct nh_notifier_grp_info with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nh_notifier_grp_info.

Cc: David Ahern <dsahern@kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Tom Rix <trix@redhat.com>
Cc: netdev@vger.kernel.org
Cc: llvm@lists.linux.dev
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

netlink: Annotate struct netlink_policy_dump_state with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct netlink_policy_dump_state.

Additionally update the size of the usage array length before accessing
it. This requires remembering the old size for the memset() and later
assignments.

Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Johannes Berg <johannes.berg@intel.com>
Cc: netdev@vger.kernel.org
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

nfp: nsp: Annotate struct nfp_eth_table with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nfp_eth_table.

Cc: Simon Horman <simon.horman@corigine.com>
Cc: Jakub Kicinski <kuba@kernel.org>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: Eric Dumazet <edumazet@google.com>
Cc: Paolo Abeni <pabeni@redhat.com>
Cc: Yinjun Zhang <yinjun.zhang@corigine.com>
Cc: Leon Romanovsky <leon@kernel.org>
Cc: Yu Xiao <yu.xiao@corigine.com>
Cc: Sixiang Chen <sixiang.chen@corigine.com>
Cc: oss-drivers@corigine.com
Cc: netdev@vger.kernel.org
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Acked-by: Louis Peens <louis.peens@corigine.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

nfp: Annotate struct nfp_reprs with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nfp_reprs.

Cc: Simon Horman <simon.horman@corigine.com>
Cc: oss-drivers@corigine.com
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Acked-by: Louis Peens <louis.peens@corigine.com>
Link: https://lore.kernel.org/r/20231003231843.work.811-kees@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

netem: Annotate struct disttable with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct disttable.

Cc: Jamal Hadi Salim <jhs@mojatatu.com>
Cc: Cong Wang <xiyou.wangcong@gmail.com>
Cc: Jiri Pirko <jiri@resnulli.us>
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: Stephen Hemminger <stephen@networkplumber.org>
Link: https://lore.kernel.org/r/20231003231823.work.684-kees@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

nexthop: Annotate struct nh_notifier_res_table_info with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct
nh_notifier_res_table_info.

Cc: Nathan Chancellor <nathan@kernel.org>
Cc: Nick Desaulniers <ndesaulniers@google.com>
Cc: Tom Rix <trix@redhat.com>
Cc: llvm@lists.linux.dev
Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20231003231818.work.883-kees@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

nexthop: Annotate struct nh_res_table with __counted_by

Prepare for the coming implementation by GCC and Clang of the __counted_by
attribute. Flexible array members annotated with __counted_by can have
their accesses bounds-checked at run-time via CONFIG_UBSAN_BOUNDS (for
array indexing) and CONFIG_FORTIFY_SOURCE (for strcpy/memcpy-family
functions).

As found with Coccinelle[1], add __counted_by for struct nh_res_table.

Link: https://github.com/kees/kernel-tools/blob/trunk/coccinelle/examples/counted_by.cocci
Signed-off-by: Kees Cook <keescook@chromium.org>
Reviewed-by: Gustavo A. R. Silva <gustavoars@kernel.org>
Reviewed-by: David Ahern <dsahern@kernel.org>
Link: https://lore.kernel.org/r/20231003231813.work.042-kees@kernel.org
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge branch 'rework-tx-fault-fixups'

Russell King says:

====================
Rework tx fault fixups

This series reworks the tx-fault fixup and then improves the Nokia GPON
workaround to also ignore the RX LOS signal as well. We do this by
introducing a mask of hardware pin states that should be ignored,
converting the tx-fault fixup to use that, and then augmenting it for
RX LOS.
====================

Link: https://lore.kernel.org/r/ZRwYJXRizvkhm83M@shell.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: sfp: improve Nokia GPON sfp fixup

Improve the Nokia GPON fixup - we need to ignore not only the hardware
LOS signal, but also the software implementation as well. Do this by
using the new state_ignore_mask to indicate that we should ignore not
only the hardware RX_LOS signal, and also clear the LOS bits in the
option field.

Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Tested-by: Christian Marangi <ansuelsmth@gmail.com>
Link: https://lore.kernel.org/r/E1qnfXh-008UDe-F9@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: sfp: re-implement ignoring the hardware TX_FAULT signal

Re-implement how we ignore the hardware TX_FAULT signal. Rather than
having a separate boolean for this, use a bitmask of the hardware
signals that we wish to ignore. This gives more flexibility in the
future to ignore other signals such as RX_LOS.

Signed-off-by: Russell King (Oracle) <rmk+kernel@armlinux.org.uk>
Tested-by: Christian Marangi <ansuelsmth@gmail.com>
Link: https://lore.kernel.org/r/E1qnfXc-008UDY-91@rmk-PC.armlinux.org.uk
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

net: cpmac: remove driver to prepare for platform removal

AR7 is going to be removed from the Kernel, so remove its networking
support in form of the cpmac driver. This allows us to remove the
platform because this driver includes a platform specific header.

Signed-off-by: Wolfram Sang <wsa+renesas@sang-engineering.com>
Acked-by: Florian Fainelli <f.fainelli@gmail.com>
Link: https://lore.kernel.org/all/20230922061530.3121-6-wsa+renesas@sang-engineering.com/
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge git://git./linux/kernel/git/netdev/net

Cross-merge networking fixes after downstream PR.

No conflicts (or adjacent changes of note).

Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Merge patch series "can: at91: add can_state_get_by_berr_counter() helper, cleanup and convert to rx_offload"

Marc Kleine-Budde <mkl@pengutronix.de> says:

This series first introduces the can_state_get_by_berr_counter()
helper function. It returns the current TX and RX state depending on
the provided CAN bit error counters. It will be later used by the
at91_can driver.

The remaining patches of this series first clean up the at91_can
driver, clean up the bus- and line error (including bus-off) handling,
and then convert it use the rx_offload helper. The driver works better
under high system load and the order of received CAN frames is better
maintained.

Due to a hardware limitation the converted driver could trigger a race
condition in the can_restart() CAN bus-off handler. The patch series
[1] fixes the issue.

[1] https://lore.kernel.org/all/20231005-can-dev-fix-can-restart-v2-0-91b5c1fd922c@pengutronix.de

Changes in v2:
- 1/27: can_state_err_to_state(): use symbolic error values instead of
  plain numbers (Thanks Vincent)
- 27/27: fix patch description and typos (Thanks Vincent)
- Link to v1: https://lore.kernel.org/all/20231004-at91_can-rx_offload-v1-0-c32bf99097db@pengutronix.de

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-0-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: switch to rx-offload implementation

The current at91_can driver uses NAPI to handle RX'ed CAN frames, the
RX IRQ is disabled and a NAPI poll is scheduled. Then in
at91_poll_rx() the RX'ed CAN frames are tried to read in order from
the device.

This approach has 2 drawbacks:

- Under high system load it might take too long from the initial RX
  IRQ to the NAPI poll function to run. This causes RX buffer
  overflows.
- The algorithm to read the CAN frames in order is not bullet proof
  and may fail under certain use cases/system loads.

The rx-offload helper fixes these problems by reading the RX'ed CAN
frames in the interrupt handler and adding it to a list sorted by RX
timestamp. This list of RX'ed SKBs is then passed to the networking
stack via NAPI.

Convert the RX path to rx-offload, pass all CAN error frames with
can_rx_offload_queue_timestamp().

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-27-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_alloc_can_err_skb() introduce new function

This is a preparation patch to convert the driver to make use of the
rx-offload helper. With rx-offload the received CAN frames are sorted
by their timestamp. Regular CAN RX'ed and TX'ed CAN frames are
timestamped by the hardware. Error events are not.

Introduce a new function at91_alloc_can_err_skb() the allocates an
error SKB and reads the current timestamp from the controller.

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-26-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_line(): send error counters with state change

Since 3e5c291c7942 ("can: add CAN_ERR_CNT flag to notify availability
of error counter") there is a dedicated flag to inform the user space,
that there are CAN error counters in the CAN error frame.

In case the device is not in bus off mode, send the error counters to
user space and set CAN_ERR_CNT.

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-25-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_line(): make use of can_change_state() and can_bus_off()

The driver implements a hand crafted CAN state handling. Update the
driver to make use of can_change_state(), introduced in ("can: dev:
Consolidate and unify state change handling")

Also switch from hand crafted CAN bus off handling to can_bus_off():
In case of a bus off, abort all pending TX requests, switch off the
device and let can_bus_off() handle the device restart.

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-24-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_line(): take reg_sr into account for bus off

The at91 CAN controller automatically recovers from bus-off after 128
occurrences of 11 consecutive recessive bits.

After an auto-recovered bus-off, the error counters no longer reflect
this fact. On the sam9263 the state bits in the SR register show the
current state (based on the current error counters), while on sam9x5
and newer SoCs these bits are latched.

Take any latched bus-off information from the SR register into account
when calculating the CAN new state, to start the standard CAN bus off
handling.

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-23-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_line(): make use of can_state_get_by_berr_counter()

On the sam9263 the SR bits for bus off, error passive, warning limit,
and error active are not latched and reflect the current status of the
controller. On the sam9x5 and newer SoCs these bits are latched.

To simplify the code, use can_state_get_by_berr_counter() to get the
state of the controller regardless of the SoC version.

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-22-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err(): rename to at91_irq_err_line()

This is a cleanup patch, no functional change intended.

The function at91_irq_err() only handles the CAN line errors, so
rename it accordingly to at91_irq_err_line().

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-21-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_frame(): move next to at91_irq_err()

This is a cleanup patch, no functional change intended. As
at91_irq_err_frame() is called from the IRQ handler move it in front
of the IRQ handler next to at91_irq_err().

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-20-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>

can: at91_can: at91_irq_err_frame(): call directly from IRQ handler

This is a preparation patch to convert the driver to the rx-offload
helper. In rx-offload RX, TX-done and CAN error handling are done in
the IRQ handler, SKB are pushed to the network stack in the NAPI poll
function.

Move the CAN frame error handling from the NAPI function at91_poll()
to the IRQ handler at91_poll(). To reflect this change, rename
at91_poll_err() to at91_irq_err_frame().

Link: https://lore.kernel.org/all/20231005-at91_can-rx_offload-v2-19-9987d53600e0@pengutronix.de
Signed-off-by: Marc Kleine-Budde <mkl@pengutronix.de>