09 Oct, 2012
2 commits
-
A few places free skbs using dev_kfree_skb even though they're called
after ieee80211_subif_start_xmit might have cloned it for tracking tx
status. Use ieee80211_free_txskb here to prevent skb leaks.Signed-off-by: Felix Fietkau
Cc: stable@vger.kernel.org
Signed-off-by: John W. Linville -
The call to drv_get/set_tsf() was put on the workqueue to perform tsf
adjustments since that function might sleep. However it ended up inside
a spinlock, whose critical section must be atomic. Do tsf adjustment
outside the spinlock instead, and get rid of a warning.Signed-off-by: Thomas Pedersen
Signed-off-by: John W. Linville
21 Sep, 2012
3 commits
-
There's no reason to send a delBA when the
peer refused our addBA, so change that.Signed-off-by: Johannes Berg
-
When a station is removed and we stop the aggregation
sessions, it's not useful to send delBA since this is
due to us or the station disassociating or dropping
the connection in some other way, so change that.Signed-off-by: Johannes Berg
-
When we disassociate, it's not really useful to
send delBA action frames since we're going to send
disassoc/deauth anyway, so change that.Signed-off-by: Johannes Berg
20 Sep, 2012
1 commit
-
In the IBSS auth TX debug message the BSSID and DA
address are reversed, fix that.Signed-off-by: Sylvain Roger Rieunier
[reword commit message and make it fit 72 cols]
Signed-off-by: Johannes Berg
19 Sep, 2012
2 commits
-
The name can't be modified by the driver,
make it const.Signed-off-by: Johannes Berg
-
The suspend/resume code depends on CONFIG_PM, so
the reset debugfs file can only be made available
if that is enabled.Fengguang Wu's zero-day build testing found this.
Signed-off-by: Johannes Berg
14 Sep, 2012
3 commits
-
Peer link which is blocked using the "iw mesh0 station
set plink_action block" is previously not able
to re-open using "iw mesh0 station set
plink_action open". This patch is intended to solve this.If the station plink state remains at OPN_SNT once open,
try block and open again should solve this problem.Signed-off-by: Chun-Yeow Yeoh
Signed-off-by: Johannes Berg -
Make the function acquire the necessary mutex itself
to simplify the callers.Signed-off-by: Johannes Berg
-
Currently, mac80211 uses the power constraint IE, and reduces
the regulatory max TX power by it. This can cause issues if
the AP is advertising a large power constraint value matching
a high TX power in its country IE, for example in this case:...
Country: US Environment: Indoor/Outdoor
...
Channels [157 - 157] @ 30 dBm
...
Power constraint: 13 dB
...What happened here is that our local regulatory TX power is
15 dBm, and gets reduced by 13 dB so we end up with only
2 dBm effective TX power, which is way too low.Instead, handle the country IE/power constraint IE combined
and restrict our TX power to the max of the regulatory power
and the maximum power advertised by the AP, in this case
17 dBm (= 30 dBm - 13 dB).Also print a message when this happens to let the user know
and help us debug issues with it.Reported-by: Carl A. Cook
Tested-by: Carl A. Cook
Signed-off-by: Johannes Berg
11 Sep, 2012
2 commits
-
The functions are only called if CONFIG_PM is set
as the callers are under an ifdef, so there's no
need to also define no-op functions.Signed-off-by: Johannes Berg
-
skb->dev might contain a stale reference to a device that was already
deleted, and using it unchecked can lead to invalid pointer accesses.
Since this is only used for nl80211 tx, iterate over active interfaces
to find a match for skb->dev, and discard the tx status if the device
is gone.Signed-off-by: Felix Fietkau
Signed-off-by: Johannes Berg
10 Sep, 2012
2 commits
-
mac80211 calls synchronize_rcu() on sta deletion,
which increase the roaming time significantly.Convert it into a call_rcu() mechanism, in order
to avoid blocking. Since some of the cleanup
functions might sleep, schedule from the call_rcu
callback a new work that will do the actual cleanup.In order to make sure the cleanup occurs before
the interface went down, flush local->workqueue
on ieee80211_do_stop().Signed-off-by: Yoni Divinsky
Signed-off-by: Eliad Peller
Signed-off-by: Johannes Berg -
Mark keys that might be used to receive management
frames so drivers can fall back on software crypto
for them if they don't support hardware offload.
As the new flag is only set correctly for RX keys
and the existing IEEE80211_KEY_FLAG_SW_MGMT flag
can only affect TX, also rename the latter to
IEEE80211_KEY_FLAG_SW_MGMT_TX.Signed-off-by: Johannes Berg
07 Sep, 2012
3 commits
-
Whenever a host gets an AUTH frame it first allocates a new
station and then replies with another AUTH frame. However,
if sta allocations fails the host should send a DEAUTH frame
instead to tell the other end that something went wrong.Signed-off-by: Antonio Quartulli
[reword commit message a bit]
Signed-off-by: Johannes Berg -
Move ieee80211_send_deauth_disassoc() to util.c to make it
available for the rest of the mac80211 code.Signed-off-by: Antonio Quartulli
[reword commit message]
Signed-off-by: Johannes Berg -
removes unnecessary semicolon
Found by Coccinelle: http://coccinelle.lip6.fr/
Signed-off-by: Peter Senna Tschudin
Signed-off-by: Johannes Berg
06 Sep, 2012
8 commits
-
Split functionality for further reuse.
Will prevent code duplication when channel context
channel_type merging is introduced.Signed-off-by: Michal Kazior
Signed-off-by: Johannes Berg -
__ieee80211_key_destroy() calls synchronize_rcu() in
order to sync the tx path before destroying the key.However, synching the tx path can be done with
synchronize_net() as well, which is usually faster
(the timing might be important for roaming scenarios).Signed-off-by: Eliad Peller
Signed-off-by: Johannes Berg -
The power constraint IE is always a single byte
so check the size when parsing instead of later.Signed-off-by: Johannes Berg
-
Disconnect from the AP if channel switching in the
driver failed or if the new channel is unavailable.Signed-off-by: Johannes Berg
-
There's no need to keep a copy of the scheduled
scan IEs after the driver has been told, if it
requires a copy it must make one. Therefore, we
can move sched_scan_ies into the function.Signed-off-by: Johannes Berg
-
Pull in mac80211.git to let the next patch apply
without conflicts, also resolving a hwsim conflict.Conflicts:
drivers/net/wireless/mac80211_hwsim.cSigned-off-by: Johannes Berg
-
Using list_move() instead of list_del() + list_add().
spatch with a semantic match is used to found this problem.
(http://coccinelle.lip6.fr/)Signed-off-by: Wei Yongjun
Signed-off-by: Johannes Berg
05 Sep, 2012
2 commits
-
Various small fixes for net/mac80211/cfg.c:mpath_set_pinfo():
Initialize *pinfo before filling members in, handle MESH_PATH_RESOLVED
correctly, and remove bogus assignment; result in correct display
of FLAGS values and meaningful EXPTIME for expired paths in iw utility.Signed-off-by: Yoichi Shinoda
Signed-off-by: Johannes Berg -
Doing so creates warnings, but the function is internal and
not part of the 802.11 docbooks, so it from kerneldoc.Signed-off-by: Johannes Berg
04 Sep, 2012
2 commits
-
ifmgd->bssid wasn't cleared properly in some
auth/assoc failure cases, causing mac80211 and
the low-level driver to go out of sync.Clear ifmgd->bssid on failure, and notify the driver.
Cc: stable@kernel.org # 3.4+
Signed-off-by: Eliad Peller
Signed-off-by: Johannes Berg -
A P2P Device interface does not have a netdev, and is not
expected to be used for transmitting data, so there is no
need to assign hw queues for it.Signed-off-by: Ilan Peer
Signed-off-by: Johannes Berg
23 Aug, 2012
1 commit
22 Aug, 2012
3 commits
-
Commit mac80211: avoid using synchronize_rcu in ieee80211_set_probe_resp
changed the return value when the probe response template is not present.
Revert to the earlier value of 1 - this fixes AP mode for drivers like
ath9k.Signed-off-by: Sujith Manoharan
Signed-off-by: Johannes Berg -
The destination address of unicast frames forwarded through a mesh gate
was being replaced with the broadcast address. Instead leave the
original destination address as the mesh DA. If the nexthop address is
not in the mpath table it will be resolved. If that fails, the frame
will be forwarded to known mesh gates.Reported-by: Cedric Voncken
Signed-off-by: Thomas Pedersen
Signed-off-by: Johannes Berg -
Conflicts:
drivers/net/wireless/mac80211_hwsim.c
20 Aug, 2012
6 commits
-
There's no need to carry around a return value
that is always NETDEV_TX_OK anyway.Signed-off-by: Johannes Berg
-
In multi-channel scenarios, the channel that we will
transmit a probe request on isn't always the current
channel (which will be NULL anyway) but will instead
be the channel that the AP is on. Pass the channel
to the ieee80211_send_probe_req() function so it can
be used in the different scenarios. The scan code
continues to pass the current channel, of course.Signed-off-by: Johannes Berg
-
There's no need to BUG_ON when a driver registers
invalid operations, warn and return an error.Signed-off-by: Johannes Berg
-
The optimisation of scanning only on the current
channel should check the operating channel. Also
modify it to compare channel pointer rather than
the frequency.Signed-off-by: Johannes Berg
-
Even for single-channel devices it is possible that we
switch the channel temporarily (e.g. for scanning) but
while doing so process a received frame that was still
received on the old channel, so checking the current
band is racy. Use the band from status instead.Signed-off-by: Johannes Berg
-
With the move to multi-channel and away from
drv_config(), hw.conf.channel will not always
be set, only for devices using the current API
instead of the new channel context APIs. Check
the channel is set before adding its frequency
to the trace data.Also break some overly long lines in the code.
Signed-off-by: Johannes Berg