Jakub Kicinski says:

====================
devlink: remove the wait-for-references on unregister

Move the registration and unregistration of the devlink instances
under their instance locks. Don't perform the netdev-style wait
for all references when unregistering the instance.

Instead the devlink instance refcount will only ensure that
the memory of the instance is not freed. All places which acquire
access to devlink instances via a reference must check that the
instance is still registered under the instance lock.

This fixes the problem of the netdev code accessing devlink
instances before they are registered.

RFC: https://lore.kernel.org/all/20221217011953.152487-1-kuba@kernel.org/
 - rewrite the cover letter
 - rewrite the commit message for patch 1
 - un-export and rename devl_is_alive
 - squash the netdevsim patches
====================
Signed-off-by: David S. Miller <davem@davemloft.net>

To prevent races with netdev code accessing free devlink instances
move the registration under the devlink instance lock.
Core now waits for the instance to be registered before accessing it.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

err_dl_unregister should unregister the devlink instance.
Looks like renaming it was missed in one of the reshufflings.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

It's most natural to register the instance first and then its
subobjects. Now that we can use the instance lock to protect
the atomicity of all init - it should also be safe.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

Requiring devlink_set_features() to be run before devlink is
registered is overzealous. devlink_set_features() itself is
a leftover from old workarounds which were trying to prevent
initiating reload before probe was complete.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

The objective of exposing the devlink instance locks to
drivers was to let them use these locks to prevent user space
from accessing the device before it's fully initialized.
This is difficult because devlink_unregister() waits for all
references to be released, meaning that devlink_unregister()
can't itself be called under the instance lock.

To avoid this issue devlink_register() was moved after subobject
registration a while ago. Unfortunately the netdev paths get
a hold of the devlink instances _before_ they are registered.
Ideally netdev should wait for devlink init to finish (synchronizing
on the instance lock). This can't work because we don't know if the
instance will _ever_ be registered (in case of failures it may not).
The other option of returning an error until devlink_register()
is called is unappealing (user space would get a notification
netdev exist but would have to wait arbitrary amount of time
before accessing some of its attributes).

Weaken the guarantees of the devlink references.

Holding a reference will now only guarantee that the memory
of the object is around. Another way of looking at it is that
the reference now protects the object not its "registered" status.
Use devlink instance lock to synchronize unregistration.

This implies that releasing of the "main" reference of the devlink
instance moves from devlink_unregister() to devlink_free().
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Always check under the instance lock whether the devlink instance
is still / already registered.

This is a no-op for the most part, as the unregistration path currently
waits for all references. On the init path, however, we may temporarily
open up a race with netdev code, if netdevs are registered before the
devlink instance. This is temporary, the next change fixes it, and this
commit has been split out for the ease of review.

Note that in case of iterating over sub-objects which have their
own lock (regions and line cards) we assume an implicit dependency
between those objects existing and devlink unregistration.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

devlink->dev is assumed to be always valid as long as any
outstanding reference to the devlink instance exists.

In prep for weakening of the references take the instance lock.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

devlink_pernet_pre_exit() is the only obvious place which takes
the instance lock without using the devl_ helpers. Update the code
and move the error print after releasing the reference
(having unlock and put together feels slightly idiomatic).
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Signed-off-by: David S. Miller <davem@davemloft.net>

xa_find_after() is designed to handle multi-index entries correctly.
If a xarray has two entries one which spans indexes 0-3 and one at
index 4 xa_find_after(0) will return the entry at index 4.

Having to juggle the two callbacks, however, is unnecessary in case
of the devlink xarray, as there is 1:1 relationship with indexes.

Always use xa_find() and increment the index manually.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

All were not visible to the non-priv users inside netns. However,
with 4ecb9009 ("sysctl: allow override of /proc/sys/net with
CAP_NET_ADMIN"), these vars are protected from getting modified.
A proc with capable(CAP_NET_ADMIN) can change the values so
not having them visible inside netns is just causing nuisance to
process that check certain values (e.g. net.core.somaxconn) and
see different behavior in root-netns vs. other-netns

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: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: Mahesh Bandewar <maheshb@google.com>
Acked-by: Soheil Hassas Yeganeh <soheil@google.com>
Signed-off-by: David S. Miller <davem@davemloft.net>

Jakub Kicinski says:

====================
devlink: code split and structured instance walk

Split devlink.c into a handful of files, trying to keep the "core"
code away from all the command-specific implementations.
The core code has been quite scattered until now. Going forward we can
consider using a source file per-subobject, I think that it's quite
beneficial to newcomers (based on relative ease with which folks
contribute to ethtool vs devlink). But this series doesn't split
everything out, yet - partially due to backporting concerns,
but mostly due to lack of time. Bulk of the netlink command
handling is left in a leftover.c file.

Introduce a context structure for dumps, and use it to store
the devlink instance ID of the last dumped devlink instance.
This means we don't have to restart the walk from 0 each time.

Finally - introduce a "structured walk". A centralized dump handler
in devlink/netlink.c which walks the devlink instances, deals with
refcounting/locking, simplifying the per-object implementations quite
a bit. Inspired by the ethtool code.

v1: https://lore.kernel.org/all/20230104041636.226398-1-kuba@kernel.org/
RFC: https://lore.kernel.org/all/20221215020155.1619839-1-kuba@kernel.org/
====================

Link: https://lore.kernel.org/r/20230105040531.353563-1-kuba@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Soon we'll have to check if a devlink instance is alive after
locking it. Convert to the by-instance dumping scheme to make
refactoring easier.

Most of the subobject code no longer has to worry about any devlink
locking / lifetime rules (the only ones that still do are the two subject
types which stubbornly use their own locking). Both dump and do callbacks
are given a devlink instance which is already locked and good-to-access
(do from the .pre_doit handler, dump from the new dump indirection).

Note that we'll now check presence of an op (e.g. for sb_pool_get)
under the devlink instance lock, that will soon be necessary anyway,
because we don't hold refs on the driver modules so the memory
in which ops live may be gone for a dead instance, after upcoming
locking changes.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Most dumpit implementations walk the devlink instances.
This requires careful lock taking and reference dropping.
Factor the loop out and provide just a callback to handle
a single instance dump.

Convert one user as an example, other users converted
in the next change.

Slightly inspired by ethtool netlink code.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Move the lock taking out of devlink_nl_cmd_region_get_devlink_dumpit().
This way all dumps will take the instance lock in the main iteration
loop directly, making refactoring and reading the code easier.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Use xarray id for cases of sub-objects which are iterated in
a function.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Use xarray id for cases of simple sub-object iteration.
We'll now use the state->instance for the devlink instances
and state->idx for subobject index.

Moving the definition of idx into the inner loop makes sense,
so while at it also move other sub-object local variables into
the loop.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

xarray gives each devlink instance an id and allows us to restart
walk based on that id quite neatly. This is nice both from the
perspective of code brevity and from the stability of the dump
(devlink instances disappearing from before the resumption point
will not cause inconsistent dumps).

This patch takes care of simple cases where state->idx counts
devlink instances only.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Walk devlink instances only once. Dump the instance reporters
and port reporters before moving to the next instance.
User space should not depend on ordering of messages.

This will make improving stability of the walk easier.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Looks like devlinks_xa_find_get() was intended to get the mark
from the @filter argument. It doesn't actually use @filter, passing
DEVLINK_REGISTERED to xa_find_fn() directly. Walking marks other
than registered is unlikely so drop @filter argument completely.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The start variables made the code clearer when we had to access
cb->args[0] directly, as the name args doesn't explain much.
Now that we use a structure to hold state this seems no longer
needed.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Create a dump context structure instead of using cb->args
as an unsigned long array. This is a pure conversion which
is intended to be as much of a noop as possible.
Subsequent changes will use this to simplify the code.

The two non-trivial parts are:
 - devlink_nl_cmd_health_reporter_dump_get_dumpit() checks args[0]
   to see if devlink_fmsg_dumpit() has already been called (whether
   this is the first msg), but doesn't use the exact value, so we
   can drop the local variable there already
 - devlink_nl_cmd_region_read_dumpit() uses args[0] for address
   but we'll use args[1] now, shouldn't matter
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

We encourage casting struct netlink_callback::ctx to a local
struct (in a comment above the field). Provide a convenience
macro for checking if the local struct fits into the ctx.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Move out the netlink glue into a separate file.
Leave the ops in the old file because we'd have to export a ton
of functions. Going forward we should switch to split ops which
will let us to put the new ops in the netlink.c file.

Pure code move, no functional changes.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Move core code into a separate file. It's spread around the main
file which makes refactoring and figuring out how devlink works
harder.

Move the xarray, all the most core devlink instance code out like
locking, ref counting, alloc, register, etc. Leave port stuff in
leftover.c, if we want to move port code it'd probably be to its
own file.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

To make the upcoming change a pure(er?) code move rename
devlink_netdevice_event -> devlink_port_netdevice_event.
This makes it clear that it only touches ports and doesn't
belong cleanly in the core.
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The devlink code is hard to navigate with 13kLoC in one file.
I really like the way Michal split the ethtool into per-command
files and core. It'd probably be too much to split it all up,
but we can at least separate the core parts out of the per-cmd
implementations and put it in a directory so that new commands
can be separate files.

Move the code, subsequent commit will do a partial split.
Reviewed-by: Jacob Keller <jacob.e.keller@intel.com>
Reviewed-by: Jiri Pirko <jiri@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Alex Elder says:

====================
net: ipa: simplify IPA interrupt handling

One of the IPA's two IRQs fires when data on a suspended channel is
available (to request that the channel--or system--be resumed to
recieve the pending data).  This interrupt also handles a few
conditions signaled by the embedded microcontroller.

For this "IPA interrupt", the current code requires a handler to be
dynamically registered for each interrupt condition.  Any condition
that has no registered handler is quietly ignored.  This design is
derived from the downstream IPA driver implementation.

There isn't any need for this complexity.  Even in the downstream
code, only four of the available 30 or so IPA interrupt conditions
are ever handled.  So these handlers can pretty easily just be
called directly in the main IRQ handler function.

This series simplifies the interrupt handling code by having the
small number of IPA interrupt handlers be called directly, rather
than having them be registered dynamically.

Version 2 just adds a missing forward-reference, as suggested by
Caleb.
====================

Link: https://lore.kernel.org/r/20230104175233.2862874-1-elder@linaro.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

We can call the two IPA interrupt handler functions directly;
there's no need to maintain the array of handler function pointers
any more.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The dynamic assignment of IPA interrupt handlers isn't needed; we
only handle three IPA interrupt types, and their handler functions
are now assigned directly.  We can get rid of ipa_interrupt_add()
and ipa_interrupt_remove() now, because they serve no purpose.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Declare the microcontroller IPA interrupt handler publicly, and
assign it directly in ipa_interrupt_config().  Make the SUSPEND IPA
interrupt handler public, and rename it ipa_power_suspend_handler().
Assign it directly in ipa_interrupt_config() as well.

This makes it unnecessary to do this in ipa_interrupt_add().  Make
similar changes for removing IPA interrupt handlers.

The next two patches will finish the cleanup, removing the
add/remove functions and the handler array entirely.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Expose ipa_interrupt_enable() and have functions that register
IPA interrupt handlers enable them directly, rather than having the
registration process do that.  Do the same for disabling IPA
interrupt handlers.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Create new function ipa_interrupt_enable() to encapsulate enabling
one of the IPA interrupt types.  Introduce ipa_interrupt_disable()
to reverse that operation.  Add a helper function to factor out the
common register update used by both.

Use these in ipa_interrupt_add() and ipa_interrupt_remove().
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

The prototype for an IPA interrupt handler supplies the IPA
interrupt ID, so it's possible to use a single function to handle
any type of microcontroller interrupt.

Introduce ipa_uc_interrupt_handler(), which calls the event or the
response handler depending on the IRQ ID provided.  Register the new
function as the handler for both microcontroller IPA interrupt types.

The called functions don't use their "irq_id" arguments, so remove
them.
Signed-off-by: Alex Elder <elder@linaro.org>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Lorenzo Bianconi says:

====================
enetc: unlock XDP_REDIRECT for XDP non-linear buffers

Unlock XDP_REDIRECT for S/G XDP buffer and rely on XDP stack to properly
take care of the frames.
Rely on XDP_FLAGS_HAS_FRAGS flag to check if it really necessary to access
non-linear part of the xdp_buff/xdp_frame.
====================

Link: https://lore.kernel.org/r/cover.1672840490.git.lorenzo@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Move XDP skb_shared_info structure initialization in from
enetc_map_rx_buff_to_xdp() to enetc_add_rx_buff_to_xdp() and do not always
access skb_shared_info in the xdp_buff/xdp_frame since it is located in a
different cacheline with respect to hard_start and data xdp pointers.
Rely on XDP_FLAGS_HAS_FRAGS flag to check if it really necessary to access
non-linear part of the xdp_buff/xdp_frame.
Reviewed-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Remove xdp_redirect_sg counter and the related ethtool entry since it is
no longer used.
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Even if full XDP_REDIRECT is not supported yet for non-linear XDP buffers
since we allow redirecting just into CPUMAPs, unlock XDP_REDIRECT for
S/G XDP buffer and rely on XDP stack to properly take care of the
frames.
Tested-by: Vladimir Oltean <vladimir.oltean@nxp.com>
Signed-off-by: Lorenzo Bianconi <lorenzo@kernel.org>
Reviewed-by: Leon Romanovsky <leonro@nvidia.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

No conflicts.
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Pull networking fixes from Jakub Kicinski:
 "Including fixes from bpf, wifi, and netfilter.

  Current release - regressions:

   - bpf: fix nullness propagation for reg to reg comparisons, avoid
     null-deref

   - inet: control sockets should not use current thread task_frag

   - bpf: always use maximal size for copy_array()

   - eth: bnxt_en: don't link netdev to a devlink port for VFs

  Current release - new code bugs:

   - rxrpc: fix a couple of potential use-after-frees

   - netfilter: conntrack: fix IPv6 exthdr error check

   - wifi: iwlwifi: fw: skip PPAG for JF, avoid FW crashes

   - eth: dsa: qca8k: various fixes for the in-band register access

   - eth: nfp: fix schedule in atomic context when sync mc address

   - eth: renesas: rswitch: fix getting mac address from device tree

   - mobile: ipa: use proper endpoint mask for suspend

  Previous releases - regressions:

   - tcp: add TIME_WAIT sockets in bhash2, fix regression caught by
     Jiri / python tests

   - net: tc: don't intepret cls results when asked to drop, fix
     oob-access

   - vrf: determine the dst using the original ifindex for multicast

   - eth: bnxt_en:
      - fix XDP RX path if BPF adjusted packet length
      - fix HDS (header placement) and jumbo thresholds for RX packets

   - eth: ice: xsk: do not use xdp_return_frame() on tx_buf->raw_buf,
     avoid memory corruptions

  Previous releases - always broken:

   - ulp: prevent ULP without clone op from entering the LISTEN status

   - veth: fix race with AF_XDP exposing old or uninitialized
     descriptors

   - bpf:
      - pull before calling skb_postpull_rcsum() (fix checksum support
        and avoid a WARN())
      - fix panic due to wrong pageattr of im->image (when livepatch and
        kretfunc coexist)
      - keep a reference to the mm, in case the task is dead

   - mptcp: fix deadlock in fastopen error path

   - netfilter:
      - nf_tables: perform type checking for existing sets
      - nf_tables: honor set timeout and garbage collection updates
      - ipset: fix hash:net,port,net hang with /0 subnet
      - ipset: avoid hung task warning when adding/deleting entries

   - selftests: net:
      - fix cmsg_so_mark.sh test hang on non-x86 systems
      - fix the arp_ndisc_evict_nocarrier test for IPv6

   - usb: rndis_host: secure rndis_query check against int overflow

   - eth: r8169: fix dmar pte write access during suspend/resume with
     WOL

   - eth: lan966x: fix configuration of the PCS

   - eth: sparx5: fix reading of the MAC address

   - eth: qed: allow sleep in qed_mcp_trace_dump()

   - eth: hns3:
      - fix interrupts re-initialization after VF FLR
      - fix handling of promisc when MAC addr table gets full
      - refine the handling for VF heartbeat

   - eth: mlx5:
      - properly handle ingress QinQ-tagged packets on VST
      - fix io_eq_size and event_eq_size params validation on big endian
      - fix RoCE setting at HCA level if not supported at all
      - don't turn CQE compression on by default for IPoIB

   - eth: ena:
      - fix toeplitz initial hash key value
      - account for the number of XDP-processed bytes in interface stats
      - fix rx_copybreak value update

  Misc:

   - ethtool: harden phy stat handling against buggy drivers

   - docs: netdev: convert maintainer's doc from FAQ to a normal
     document"

* tag 'net-6.2-rc3' of git://git.kernel.org/pub/scm/linux/kernel/git/netdev/net: (112 commits)
  caif: fix memory leak in cfctrl_linkup_request()
  inet: control sockets should not use current thread task_frag
  net/ulp: prevent ULP without clone op from entering the LISTEN status
  qed: allow sleep in qed_mcp_trace_dump()
  MAINTAINERS: Update maintainers for ptp_vmw driver
  usb: rndis_host: Secure rndis_query check against int overflow
  net: dpaa: Fix dtsec check for PCS availability
  octeontx2-pf: Fix lmtst ID used in aura free
  drivers/net/bonding/bond_3ad: return when there's no aggregator
  netfilter: ipset: Rework long task execution when adding/deleting entries
  netfilter: ipset: fix hash:net,port,net hang with /0 subnet
  net: sparx5: Fix reading of the MAC address
  vxlan: Fix memory leaks in error path
  net: sched: htb: fix htb_classify() kernel-doc
  net: sched: cbq: dont intepret cls results when asked to drop
  net: sched: atm: dont intepret cls results when asked to drop
  dt-bindings: net: marvell,orion-mdio: Fix examples
  dt-bindings: net: sun8i-emac: Add phy-supply property
  net: ipa: use proper endpoint mask for suspend
  selftests: net: return non-zero for failures reported in arp_ndisc_evict_nocarrier
  ...