03 May, 2019
1 commit
-
Three trivial overlapping conflicts.
Signed-off-by: David S. Miller
27 Apr, 2019
1 commit
-
…inux/kernel/git/jberg/mac80211-next
Johannes Berg says:
====================
Various updates, notably:
* extended key ID support (from 802.11-2016)
* per-STA TX power control support
* mac80211 TX performance improvements
* HE (802.11ax) updates
* mesh link probing support
* enhancements of multi-BSSID support (also related to HE)
* OWE userspace processing support
====================Signed-off-by: David S. Miller <davem@davemloft.net>
26 Apr, 2019
20 commits
-
The requirement for mesh link metric refreshing, is that from one
mesh point we be able to send some data frames to other mesh points
which are not currently selected as a primary traffic path, but which
are only 1 hop away. The absence of the primary path to the chosen node
makes it necessary to apply some form of marking on a chosen packet
stream so that the packets can be properly steered to the selected node
for testing, and not by the regular mesh path lookup.Tested-by: Pradeep Kumar Chitrapu
Signed-off-by: Rajkumar Manoharan
Signed-off-by: Johannes Berg -
Allows setting of control flags of skb cb - if needed -
when calling ieee80211_subif_start_xmit().Tested-by: Pradeep Kumar Chitrapu
Signed-off-by: Rajkumar Manoharan
Signed-off-by: Johannes Berg -
Mac80211 SW crypto handles replacing PTK keys correctly.
Don't trigger needless warnings or workarounds when the driver can only
use the known good SW crypto provided by mac80211.Signed-off-by: Alexander Wetzel
Signed-off-by: Johannes Berg -
Only enable Extended Key ID support for drivers which are not supporting
crypto offload and also do not support A-MPDU.While any driver using SW crypto from mac80211 is generally able to also
support Extended Key ID these drivers are likely to mix keyIDs in
AMPDUs when rekeying.According to IEEE 802.11-2016 "9.7.3 A-MPDU contents" this is not
allowed.Signed-off-by: Alexander Wetzel
[reword comment a bit, move ! into logic expression]
Signed-off-by: Johannes Berg -
The cfg80211_merge_profile() and ieee802_11_find_bssid_profile() are
a bit cleaner if we just pass the merged_ie pointer instead of a pointer
to the pointer.This isn't a functional change, it's just a clean up.
Signed-off-by: Dan Carpenter
Signed-off-by: Johannes Berg -
This patch introduce a new driver callback drv_sta_set_txpwr. This API will
copy the transmit power value passed from user space and call the driver
callback to set the tx power for the station.Co-developed-by: Balaji Pothunoori
Signed-off-by: Ashok Raj Nagarajan
Signed-off-by: Balaji Pothunoori
Signed-off-by: Johannes Berg -
There's no need to allocate than one queue in the iTXQs case
now that we no longer use ndo_select_queue to assign the AC.Signed-off-by: Johannes Berg
-
At Technical University of Munich we use MAC 802.11 TX status frames to
perform several measurements in MAC 802.11 setups.With ath based drivers this was possible until commit d94a461d7a7df6
("ath9k: use ieee80211_tx_status_noskb where possible") as the driver
ignored the IEEE80211_TX_CTL_REQ_TX_STATUS flag and always delivered
tx_status frames. Since that commit, this behavior was changed and the
driver now adheres to IEEE80211_TX_CTL_REQ_TX_STATUS.Due to performance reasons, IEEE80211_TX_CTL_REQ_TX_STATUS is not set for
data frames from interfaces in managed mode. Hence, frames that are sent
from a managed mode interface do never deliver tx_status frames. This
remains true even if a monitor mode interface (the measurement interface)
is added to the same ieee80211 physical device. Thus, there is no
possibility for receiving tx_status frames for frames sent on an interface
in managed mode, if the driver adheres to IEEE80211_TX_CTL_REQ_TX_STATUS.In order to force delivery of tx_status frames for research and debugging
purposes, implement a debugfs option force_tx_status for ieee80211 physical
devices. When this option is set for a physical device,
IEEE80211_TX_CTL_REQ_TX_STATUS is enabled in all packets sent from that
device. This option can be set via
/sys/kernel/debug/ieee80211//force_tx_status. The default is disabled.Co-developed-by: Charlie Groh
Signed-off-by: Charlie Groh
Signed-off-by: Julius Niedworok
Signed-off-by: Johannes Berg -
FullMAC STAs have no way to update bss channel after CSA channel switch
completion. As a result, user-space tools may provide inconsistent
channel info. For instance, consider the following two commands:
$ sudo iw dev wlan0 link
$ sudo iw dev wlan0 info
The latter command gets channel info from the hardware, so most probably
its output will be correct. However the former command gets channel info
from scan cache, so its output will contain outdated channel info.
In fact, current bss channel info will not be updated until the
next [re-]connect.Note that mac80211 STAs have a workaround for this, but it requires
access to internal cfg80211 data, see ieee80211_chswitch_work:/* XXX: shouldn't really modify cfg80211-owned data! */
ifmgd->associated->channel = sdata->csa_chandef.chan;This patch suggests to convert mac80211 workaround into cfg80211 behavior
and to update current bss channel in cfg80211_ch_switch_notify.Signed-off-by: Sergey Matyukevich
Signed-off-by: Johannes Berg -
A per-group shift was added to reduce the size of the per-rate transmit
duration field to u16 without sacrificing a lot of precision
This patch changes the macros to automatically calculate the best value for
this shift based on the lowest rate within the group.
This simplifies adding more groups and slightly improves accuracy for some of
the existing groups.Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
This is needed for the upcoming driver for MT7615 4x4 802.11ac chipsets
Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
When using iTXQ, the network stack does not need the real queue number, since
mac80211 is using its internal queues anyway. In that case we can defer
selecting the queue and remove a redundant station lookup in the tx path to save
some CPU cycles.Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
Add support for Extended Key ID as defined in IEEE 802.11-2016.
- Implement the nl80211 API for Extended Key ID
- Extend mac80211 API to allow drivers to support Extended Key ID
- Enable Extended Key ID by default for drivers only supporting SW
crypto (e.g. mac80211_hwsim)
- Allow unicast Tx usage to be supressed (IEEE80211_KEY_FLAG_NO_AUTO_TX)
- Select the decryption key based on the MPDU keyid
- Enforce existing assumptions in the code that rekeys don't change the
cipherSigned-off-by: Alexander Wetzel
[remove module parameter]
Signed-off-by: Johannes Berg -
Optimize/cleanup the delay tailroom checks and adds one missing tailroom
update.Signed-off-by: Alexander Wetzel
Signed-off-by: Johannes Berg -
When using iTXQ, tx sequence number allocation and statistics are run at
dequeue time. Because of that, it is safe to enable NETIF_F_LLTX, which
allows tx handlers to run on multiple CPUs in parallel.Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg -
Reduces lock contention on enqueue/dequeue of iTXQ packets
Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg -
Reduces lock contention on enqueue/dequeue of iTXQ packets
Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg -
The callers of these functions are all within RCU locked sections
Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
Since an element is limited to 255 octets, a profile may be split
split to several elements. Support the split as defined in the 11ax
draft 3.Signed-off-by: Sara Sharon
Signed-off-by: Luca Coelho
Signed-off-by: Johannes Berg -
Subelement profile may specify element IDs it doesn't inherit
from the management frame. Support it.Signed-off-by: Sara Sharon
Signed-off-by: Luca Coelho
Signed-off-by: Johannes Berg
23 Apr, 2019
2 commits
-
We need to dereference the directory to get its parent to
be able to rename it, so it's clearly not safe to try to
do this with ERR_PTR() pointers. Skip in this case.It seems that this is most likely what was causing the
report by syzbot, but I'm not entirely sure as it didn't
come with a reproducer this time.Cc: stable@vger.kernel.org
Reported-by: syzbot+4ece1a28b8f4730547c9@syzkaller.appspotmail.com
Signed-off-by: Johannes Berg -
The txq of vif is added to active_txqs list for ATF TXQ scheduling
in the function ieee80211_queue_skb(), but it was not properly removed
before freeing the txq object. It was causing use after free of the txq
objects from the active_txqs list, result was kernel panic
due to invalid memory access.Fix kernel invalid memory access by properly removing txq object
from active_txqs list before free the object.Signed-off-by: Bhagavathi Perumal S
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg
18 Apr, 2019
1 commit
-
Conflict resolution of af_smc.c from Stephen Rothwell.
Signed-off-by: David S. Miller
10 Apr, 2019
1 commit
-
The original patch neglected to take byte order conversions
into account, fix that.Fixes: d9bb410888ce ("mac80211: allow overriding HT STBC capabilities")
Signed-off-by: Johannes Berg
Reviewed-by: Sergey Matyukevich
Signed-off-by: Johannes Berg
09 Apr, 2019
1 commit
-
Restore SW_CRYPTO_CONTROL operation on AP_VLAN interfaces for unicast
keys, the original override was intended to be done for group keys as
those are treated specially by mac80211 and would always have been
rejected.Now the situation is that AP_VLAN support must be enabled by the driver
if it can support it (meaning it can support software crypto GTK TX).Thus, also simplify the code - if we get here with AP_VLAN and non-
pairwise key, software crypto must be used (driver doesn't know about
the interface) and can be used (driver must've advertised AP_VLAN if
it also uses SW_CRYPTO_CONTROL).Fixes: db3bdcb9c3ff ("mac80211: allow AP_VLAN operation on crypto controlled devices")
Signed-off-by: Alexander Wetzel
[rewrite commit message]
Signed-off-by: Johannes Berg
08 Apr, 2019
2 commits
-
Add support for mesh airtime link metric attribute
NL80211_STA_INFO_AIRTIME_LINK_METRIC.Signed-off-by: Narayanraddi Masti
Signed-off-by: Johannes Berg -
Currently there is no way for the driver to signal to mac80211 that it should
schedule a TXQ even if there are no packets on the mac80211 part of that queue.
This is problematic if the driver has an internal retry queue to deal with
software A-MPDU retry.This patch changes the behavior of ieee80211_schedule_txq to always schedule
the queue, as its only user (ath9k) seems to expect such behavior already:
it calls this function on tx status and on powersave wakeup whenever its
internal retry queue is not empty.Also add an extra argument to ieee80211_return_txq to get the same behavior.
This fixes an issue on ath9k where tx queues with packets to retry (and no
new packets in mac80211) would not get serviced.Fixes: 89cea7493a346 ("ath9k: Switch to mac80211 TXQ scheduling and airtime APIs")
Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg
29 Mar, 2019
6 commits
-
Holding the lock around the entire duration of tx scheduling can create
some nasty lock contention, especially when processing airtime information
from the tx status or the rx path.
Improve locking by only holding the active_txq_lock for lookups / scheduling
list modifications.Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg -
There are several scenarios in which mac80211 can call drv_wake_tx_queue
after ieee80211_restart_hw has been called and has not yet completed.
Driver private structs are considered uninitialized until mac80211 has
uploaded the vifs, stations and keys again, so using private tx queue
data during that time is not safe.The driver can also not rely on drv_reconfig_complete to figure out when
it is safe to accept drv_wake_tx_queue calls again, because it is only
called after all tx queues are woken again.To fix this, bail out early in drv_wake_tx_queue if local->in_reconfig
is set.Cc: stable@vger.kernel.org
Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
skb->truesize can change due to memory reallocation or when adding extra
fragments. Adjust fq->memory_usage accordinglySigned-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg -
Looks that 100 chars isn't enough for messages, as we keep getting
warnings popping from different places due to message shortening.
Instead of trying to shorten the prints, just increase the buffer size.Signed-off-by: Andrei Otcheretianski
Signed-off-by: Luca Coelho
Signed-off-by: Johannes Berg -
The pointer to the last four bytes of the address is not guaranteed to be
aligned, so we need to use __get_unaligned_cpu32 hereSigned-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
Once a station enters powersave, its queues should not be returned by
ieee80211_next_txq() anymore. They will be re-scheduled again after the
station has woken up againFixes: 1866760096bf4 ("mac80211: Add TXQ scheduling API")
Signed-off-by: Felix Fietkau
Acked-by: Toke Høiland-Jørgensen
Signed-off-by: Johannes Berg
21 Mar, 2019
1 commit
-
After the previous patch, all the callers of ndo_select_queue()
provide as a 'fallback' argument netdev_pick_tx.
The only exceptions are nested calls to ndo_select_queue(),
which pass down the 'fallback' available in the current scope
- still netdev_pick_tx.We can drop such argument and replace fallback() invocation with
netdev_pick_tx(). This avoids an indirect call per xmit packet
in some scenarios (TCP syn, UDP unconnected, XDP generic, pktgen)
with device drivers implementing such ndo. It also clean the code
a bit.Tested with ixgbe and CONFIG_FCOE=m
With pktgen using queue xmit:
threads vanilla patched
(kpps) (kpps)
1 2334 2428
2 4166 4278
4 7895 8100v1 -> v2:
- rebased after helper's name changeSigned-off-by: Paolo Abeni
Signed-off-by: David S. Miller
25 Feb, 2019
1 commit
-
Three conflicts, one of which, for marvell10g.c is non-trivial and
requires some follow-up from Heiner or someone else.The issue is that Heiner converted the marvell10g driver over to
use the generic c45 code as much as possible.However, in 'net' a bug fix appeared which makes sure that a new
local mask (MDIO_AN_10GBT_CTRL_ADV_NBT_MASK) with value 0x01e0
is cleared.Signed-off-by: David S. Miller
22 Feb, 2019
3 commits
-
Forwarded packets enter the tx path through ieee80211_add_pending_skb,
which skips the ieee80211_skb_resize call.
Fixes WARN_ON in ccmp_encrypt_skb and resulting packet loss.Cc: stable@vger.kernel.org
Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg -
The code in mesh_path_add tries to handle the case where a duplicate
entry is added to the rhashtable by doing a lookup after a failed
insertion. It also tries to handle races by repeating the insertion
should the lookup fail.This is now unnecessary as we have rhashtable API functions that can
directly return the mathcing object.Signed-off-by: Herbert Xu
Signed-off-by: Johannes Berg -
Merge net-next to resolve a conflict and to get the mac80211
rhashtable fixes so further patches can be applied on top.Signed-off-by: Johannes Berg
