Delete the redundant word 'was'.
Signed-off-by: Jilin Yuan <yuanjilin@cdjrlc.com>
Link: https://lore.kernel.org/r/20220708151538.51483-1-yuanjilin@cdjrlc.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

The purpose of the netif_is_down() helper was to ensure that the network
interface used was still up when performing the transmission. What it
actually did was to check if _all_ interfaces were up. This was not
noticed at that time because I did not use interfaces at all before
discussing with Alexander Aring about how to handle coordinators
properly.

Drop the helper and call netif_running() on the right sub interface
object directly.

Fixes: 4f790184 ("net: mac802154: Add a warning in the slow path")
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220617192914.1275611-1-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

We need to call wake_up() when hold_txs reaches zero. The semantic of
atomic_dec_and_test() is that it returns true when it's zero.

Fixes: f0feb349 ("net: mac802154: Introduce a tx queue flushing mechanism")
Signed-off-by: Alexander Aring <aahringo@redhat.com>
Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/r/20220613043735.1039895-3-aahringo@redhat.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

The semantic of atomic_dec_and_test() is to return true if zero is
reached and we need call ieee802154_wake_queue() when zero is reached.

Fixes: 20a19d1d ("net: mac802154: Bring the ability to hold the transmit queue")
Signed-off-by: Alexander Aring <aahringo@redhat.com>
Reviewed-by: Miquel Raynal <miquel.raynal@bootlin.com>
Link: https://lore.kernel.org/r/20220613043735.1039895-2-aahringo@redhat.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

In order to be able to detect possible conflicts between the net
interface core and the ieee802154 core, let's add a warning in the slow
path: we want to be sure that whenever we start an asynchronous MLME
transmission (which can be fully asynchronous) the net core somehow
agrees that this transmission is possible, ie. the device was not
stopped. Warning in this case would allow us to track down more easily
possible issues with the MLME logic if we ever get reports.

Unlike in the hot path, such a situation cannot be handled.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-12-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

We should never start a transmission after the queue has been stopped.

But because it might work we don't kill the function here but rather
warn loudly the user that something is wrong.

Set a flag when the queue should remain stopped. Reset this flag when
the queue actually gets restarded. Just check this value to know if a
transmission is legitimate, warn if it is not.

Turn the flags variable into an unsigned long to allow the use of atomic
helpers on it.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-11-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

This is the slow path, we need to wait for each command to be processed
before continuing so let's introduce an helper which does the
transmission and blocks until it gets notified of its asynchronous
completion. This helper is going to be used when introducing scan
support.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-10-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Right now we are able to stop a queue but we have no indication if a
transmission is ongoing or not.

Thanks to recent additions, we can track the number of ongoing
transmissions so we know if the last transmission is over. Adding on top
of it an internal wait queue also allows to be woken up asynchronously
when this happens. If, beforehands, we marked the queue to be held and
stopped it, we end up flushing and stopping the tx queue.

Thanks to this feature, we will soon be able to introduce a synchronous
transmit API.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-9-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Sometimes calling the stop queue helper is not enough because it does
not hold any lock. In order to be safe and avoid racy situations when
trying to (soon) sync the Tx queue, for instance before sending an MLME
frame, let's now introduce an helper which actually hold the necessary
locks when doing so.
Suggested-by: Alexander Aring <alex.aring@gmail.com>
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-8-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Let's rename the current Tx path to show that this is the "hot" Tx
path. We will soon introduce a slower Tx path for MLME commands.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-7-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Create a hold_txs atomic variable and increment/decrement it when
relevant, ie. when we want to hold the queue or release it: currently
all the "stopped" situations are suitable, but very soon we will more
extensively use this feature for MLME purposes.

Upon release, the atomic counter is decremented and checked. If it is
back to 0, then the netif queue gets woken up. This makes the whole
process fully transparent, provided that all the users of
ieee802154_wake/stop_queue() now call ieee802154_hold/release_queue()
instead.

In no situation individual drivers should call any of these helpers
manually in order to avoid messing with the counters. There are other
functions more suited for this purpose which have been introduced, such
as the _xmit_complete() and _xmit_error() helpers which will handle all
that for them.

One advantage is that, as no more drivers call the stop/wake helpers
directly, we can safely stop exporting them and only declare the
hold/release ones in a header only accessible to the core.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-6-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

In order to create a synchronous API for MLME command purposes, we need
to be able to track the end of the ongoing transmissions. Let's
introduce an atomic variable which is incremented when a transmission
starts and decremented when relevant so that we know at any moment
whether there is an ongoing transmission.

The counter gets decremented in the following situations:
- The operation is asynchronous and there was a failure during the
  offloading process.
- The operation is synchronous and the synchronous operation failed.
- The operation finished, either successfully or not.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-5-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Before adding more logic in the error path, let's move the wake queue
call there, rename the default label and create an additional one.

There is no functional change.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-4-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

This entry is dedicated to synchronous transmissions done by drivers
without async hook. Make this clearer that this is not a work that any
driver can use by at least prefixing it with "sync_". While at it, let's
enhance the comment explaining why we choose one or the other.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-3-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

There are currently two driver hooks: one is synchronous, the other is
not. We cannot rely on driver implementations to provide a synchronous
API (which is related to the bus medium more than a wish to have a
synchronized implementation) so we are going to introduce a sync API
above any kind of driver transmit function. In order to clarify what
this worker is for (synchronous driver implementation), let's rename it
so that people don't get bothered by the fact that their driver does not
make use of the "xmit worker" which is a too generic name.
Signed-off-by: Miquel Raynal <miquel.raynal@bootlin.com>
Acked-by: Alexander Aring <aahringo@redhat.com>
Link: https://lore.kernel.org/r/20220519150516.443078-2-miquel.raynal@bootlin.comSigned-off-by: Stefan Schmidt <stefan@datenfreihafen.org>

Merge tag 'ieee802154-for-net-next-2022-06-09' of git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan-next

Stefan Schmidt says:

====================
pull-request: ieee802154-next 2022-06-09

This is a separate pull request for 6lowpan changes. We agreed with the
bluetooth maintainers to switch the trees these changing are going into
from bluetooth to ieee802154.

Jukka Rissanen stepped down as a co-maintainer of 6lowpan (Thanks for the
work!). Alexander is staying as maintainer.

Alexander reworked the nhc_id lookup in 6lowpan to be way simpler.
Moved the data structure from rb to an array, which is all we need in this
case.

* tag 'ieee802154-for-net-next-2022-06-09' of git://git.kernel.org/pub/scm/linux/kernel/git/sschmidt/wpan-next:
  MAINTAINERS: Remove Jukka Rissanen as 6lowpan maintainer
  net: 6lowpan: constify lowpan_nhc structures
  net: 6lowpan: use array for find nhc id
  net: 6lowpan: remove const from scalars
====================

Link: https://lore.kernel.org/r/20220609202956.1512156-1-stefan@datenfreihafen.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Jonathan Toppins says:

====================
bonding: netlink errors and cleanup

The first patch attempts to set helpful error messages when
configuring bonds via netlink. The second patch removes redundant
init code for RLB mode which is already done in bond_open.
====================

Link: https://lore.kernel.org/r/cover.1654711315.git.jtoppins@redhat.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Setting RLB_NULL_INDEX is not needed as this is done in bond_alb_initialize
which is called by bond_open.

Also reduce the number of rtnl_unlock calls by just using the standard
goto cleanup path.
Signed-off-by: Jonathan Toppins <jtoppins@redhat.com>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Add support for reporting errors via extack in both bond_newlink
and bond_changelink.

Instead of having to look in the kernel log for why an option was not
correct just report the error to the user via the extack variable.

What is currently reported today:
  ip link add bond0 type bond
  ip link set bond0 up
  ip link set bond0 type bond mode 4
 RTNETLINK answers: Device or resource busy

After this change:
  ip link add bond0 type bond
  ip link set bond0 up
  ip link set bond0 type bond mode 4
 Error: unable to set option because the bond is up.
Signed-off-by: Jonathan Toppins <jtoppins@redhat.com>
Acked-by: Jay Vosburgh <jay.vosburgh@canonical.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Tony Nguyen says:

====================
100GbE Intel Wired LAN Driver Updates 2022-06-08

Michal prevents setting of VF VLAN capabilities in switchdev mode and
removes, not needed, specific switchdev VLAN operations.

Karol converts u16 variables to unsigned int for GNSS calculations.

Christophe Jaillet corrects the parameter order for a couple of
devm_kcalloc() calls.

* '100GbE' of git://git.kernel.org/pub/scm/linux/kernel/git/tnguy/next-queue:
  ice: Use correct order for the parameters of devm_kcalloc()
  ice: remove u16 arithmetic in ice_gnss
  ice: remove VLAN representor specific ops
  ice: don't set VF VLAN caps in switchdev
====================

Link: https://lore.kernel.org/r/20220608160757.2395729-1-anthony.l.nguyen@intel.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

Eric Dumazet says:

====================
net: few debug refinements

Adopt DEBUG_NET_WARN_ON_ONCE() or WARN_ON_ONCE()
in some places where it makes sense.

Add checks in napi_consume_skb() and __napi_alloc_skb()

Make sure napi_get_frags() does not use page fragments
for skb->head.
====================

Link: https://lore.kernel.org/r/20220608160438.1342569-1-eric.dumazet@gmail.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

This is a follow up of commit 3226b158
("net: avoid 32 x truesize under-estimation for tiny skbs")

When/if we increase MAX_SKB_FRAGS, we better make sure
the old bug will not come back.

Adding a check in napi_get_frags() would be costly,
even if using DEBUG_NET_WARN_ON_ONCE().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Commit 6454eca8 ("net: Use lockdep_assert_in_softirq()
in napi_consume_skb()") added a check in napi_consume_skb()
which is a bit weak.

napi_consume_skb() and __napi_alloc_skb() should only
be used from BH context, not from hard irq or nmi context,
otherwise we could have races.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Remove this check from fast path unless CONFIG_DEBUG_NET=y
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Replace four WARN_ON() that have not triggered recently
with DEBUG_NET_WARN_ON_ONCE().
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

sk_stream_kill_queues() has three checks which have been
useful to detect kernel bugs in the past.

However they are potentially a problem because they
could flood the syslog.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

inet_sock_destruct() has four warnings which have been
useful to point to kernel bugs in the past.

However they are potentially a problem because they
could flood the syslog.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

One check in dev_loopback_xmit() has not caught issues
in the past.

Keep it for CONFIG_DEBUG_NET=y builds only.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Check against skb dst in socket backlog has never triggered
in past years.

Keep the check omly for CONFIG_DEBUG_NET=y builds.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Eric Dumazet says:

====================
net: adopt u64_stats_t type

While KCSAN has not raised any reports yet, we should address the
potential load/store tearing problem happening with per cpu stats.

This series is not exhaustive, but hopefully a step in the right
direction.
====================

Link: https://lore.kernel.org/r/20220608154640.1235958-1-eric.dumazet@gmail.comSigned-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

We have a convenient helper, let's use it.
This will make the following patch easier to review and smaller.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Reviewed-by: Jason A. Donenfeld <Jason@zx2c4.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

We have a convenient helper, let's use it.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

We have a convenient helper, let's use it.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.

Add READ_ONCE() when reading rx_errs & tx_drps.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

As explained in commit 316580b6 ("u64_stats: provide u64_stats_t type")
we should use u64_stats_t and related accessors to avoid load/store tearing.

Add READ_ONCE() when reading rx_errors & tx_dropped.
Signed-off-by: Eric Dumazet <edumazet@google.com>
Signed-off-by: Jakub Kicinski <kuba@kernel.org>

Netdev reference helpers have a dev_ prefix for historic
reasons. Renaming the old helpers would be too much churn
but we can rename the tracking ones which are relatively
recent and should be the default for new code.

Rename:
 dev_hold_track()    -> netdev_hold()
 dev_put_track()     -> netdev_put()
 dev_replace_track() -> netdev_ref_replace()

Link: https://lore.kernel.org/r/20220608043955.919359-1-kuba@kernel.orgSigned-off-by: Jakub Kicinski <kuba@kernel.org>