13 Jan, 2012
1 commit
-
commit a9b3cd7f32 (rcu: convert uses of rcu_assign_pointer(x, NULL) to
RCU_INIT_POINTER) did a lot of incorrect changes, since it did a
complete conversion of rcu_assign_pointer(x, y) to RCU_INIT_POINTER(x,
y).We miss needed barriers, even on x86, when y is not NULL.
Signed-off-by: Eric Dumazet
CC: Stephen Hemminger
CC: Paul E. McKenney
Signed-off-by: David S. Miller
05 Jan, 2012
1 commit
-
kernel panic occurs when we create an IBSS mode and leave it for
sometime without any joiner and this is introduced by the
commit ec2b774e7c91094d8c00de579646f1162b87b01e where we don't
put proper braces for 'list_for_each_entry_safe' and we pass an
invalid 'sta' pointer to __sta_info_destroyEIP is at __list_add+0xe/0xa0
EAX: f3b63db4 EBX: 00000000 ECX: eab88c1c EDX: 00000000
ESI: 00000000 EDI: 00000246 EBP: f3b63d80 ESP: f3b63d58
DS: 007b ES: 007b FS: 00d8 GS: 00e0 SS: 0068
Process kworker/u:2 (pid: 198, ti=f3b62000 task=f3afbea0
task.ti=f3b62000)
Stack:
00000000 00000000 f9ef9821 00000000 00000000 eab88c30
f3b63d80 c017f623
eab88bf0 eab88bf0 f3b63dd0 c066f925 00000000 00000002
00000000 f9ef9821
f3b63da0 c0180a2b eab88c1c eab88c30 00000002 f3afbea0
eab88bf4 f3b63db4
Call Trace:
[] ? __ieee80211_stop_tx_ba_session+0x31/0x60
[mac80211]
[] ? debug_mutex_add_waiter+0x23/0x60
[] __mutex_lock_common+0xd5/0x390
[] ? __ieee80211_stop_tx_ba_session+0x31/0x60
[mac80211]
[] ? trace_hardirqs_off+0xb/0x10
[] mutex_lock_nested+0x47/0x60
[] ? __ieee80211_stop_tx_ba_session+0x31/0x60
[mac80211]
[] __ieee80211_stop_tx_ba_session+0x31/0x60
[mac80211]
[] ieee80211_sta_tear_down_BA_sessions+0x39/0x60 [mac80211]
[] __sta_info_destroy+0x57/0x780 [mac80211]
[] ieee80211_sta_expire+0x93/0xb0 [mac80211]
[] ieee80211_ibss_work+0x2d6/0x530 [mac80211]Cc: Marek Lindner
Signed-off-by: Mohammed Shafi Shajakhan
Signed-off-by: John W. Linville
22 Dec, 2011
1 commit
-
The IBSS merge code calls ieee80211_sta_expire() with a relatively
short expire timeout that purges other clients prematurely. The
expire function has to check that only the clients belonging to
the vif in question are purged.Signed-off-by: Marek Lindner
Signed-off-by: John W. Linville
20 Dec, 2011
1 commit
-
The commit 42624d4913a00219a8fdbb4bafd634d1d843be85
created following sparse warning
>net/mac80211/sta_info.c:965:24: warning: incorrect type in assignment (different address spaces)
>net/mac80211/sta_info.c:965:24: expected struct tid_ampdu_tx *tid_tx
>net/mac80211/sta_info.c:965:24: got struct tid_ampdu_tx [noderef] *Making use of rcu_dereference_protected to fix the problem.
V2:
- Replacing rcu_dereference with rcu_dereference_protected
as suggested by Johannes.
- Adding mutex_lock/unlock to satisfy the condition at
rcu_dereference_protectedCc: Nishant Sarmukadam
Reported-by: Johannes Berg
Signed-off-by: Yogesh Ashok Powar
Signed-off-by: John W. Linville
16 Dec, 2011
5 commits
-
Now that IBSS no longer needs to insert stations
from atomic context, we can get rid of all the
special cases for that, and even get rid of the
sta_lock (though it needs to stay as tim_lock.)This makes the station management code much more
straight-forward.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
In order to notify drivers and simplify the station
management code, defer IBSS station insertion to a
work item and don't do it directly while receiving
a frame.This increases the complexity in IBSS a little bit,
but it's pretty straight forward and it allows us
to reduce the station management complexity (next
patch) considerably.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
No real changes, just note that they are const.
Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
Currently, each AP interface will send multicast
traffic if any interface has a station entry even
if that station entry is allocated only. With the
new station state management we can easily fix it
by adding a counter that counts each authorized
station only and send multicast traffic only when
the correct interface has at least one authorized
station.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
Station entries can have various states, the most
important ones being auth, assoc and authorized.
This patch prepares us for telling the driver about
these states, we don't want to confuse drivers with
strange transitions, so with this we enforce that
they move in the right order between them (back and
forth); some transitions might happen before the
driver even knows about the station, but at least
runtime transitions will be ordered correctly.As a consequence, IBSS and MESH stations will now
have the ASSOC flag set (so they can transition to
AUTHORIZED), and we can get rid of a special case
in TX processing.When freeing a station, unwind the state so that
other parts of the code (or drivers later) can rely
on the transitions.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville
14 Dec, 2011
1 commit
-
When a station leaves suddenly while ampdu traffic to that station is still
running, there is a possibility that the ampdu pending queues are not freed due
to a race condition leading to memory leaks. In '__sta_info_destroy' when we
attempt to destroy the ampdu sessions in 'ieee80211_sta_tear_down_BA_sessions',
the driver calls 'ieee80211_stop_tx_ba_cb_irqsafe' to delete the ampdu
structures (tid_tx) and splice the pending queues and this job gets queued in
sdata workqueue. However, the sta entry can get destroyed before the above work
gets scheduled and hence the race.Purging the queues and freeing the tid_tx to avoid the leak. The better solution
would be to fix the race, but that can be taken up in a separate patch.Signed-off-by: Nishant Sarmukadam
Signed-off-by: Yogesh Ashok Powar
Signed-off-by: John W. Linville
29 Nov, 2011
1 commit
-
This fixes frequent WARN_ONs when using AP VLAN + aggregation, as these vifs
are virtual and not registered with drivers.
Use sta_info_get_bss instead of sta_info_get in aggregation callbacks, so
that these callbacks can find the station entry when called with the AP vif.Signed-off-by: Felix Fietkau
Signed-off-by: John W. Linville
08 Nov, 2011
1 commit
-
Set IEEE80211_FCTL_MOREDATA bit also in case we have buffered
frames (more than one) only for one AC.Signed-off-by: Janusz Dziedzic
Reviewed-by: Johannes Berg
Signed-off-by: John W. Linville
18 Oct, 2011
1 commit
15 Oct, 2011
1 commit
-
The queue mapping/TID for non-QoS null data
responses to is never set, making it default
to BK. Fix that.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville
12 Oct, 2011
1 commit
-
Conflicts:
Documentation/feature-removal-schedule.txt
01 Oct, 2011
11 commits
-
iwlwifi has a separate EOSP notification from
the device, and to make use of that properly
it needs to be passed to mac80211. To be able
to mix with tx_status_irqsafe and rx_irqsafe
it also needs to be an "_irqsafe" version in
the sense that it goes through the tasklet,
the actual flag clearing would be IRQ-safe
but doing it directly would cause reordering
issues.This is needed in the case of a P2P GO going
into an absence period without transmitting
any frames that should be driver-released as
in this case there's no other way to inform
mac80211 that the service period ended. Note
that for drivers that don't use the _irqsafe
functions another version of this function
will be required.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
iwlwifi needs to know the number of frames that are
going to be sent to a station while it is asleep so
it can properly handle the uCode blocking of that
station.Before uAPSD, we got by by telling the device that
a single frame was going to be released whenever we
encountered IEEE80211_TX_CTL_POLL_RESPONSE. With
uAPSD, however, that is no longer possible since
there could be more than a single frame.To support this model, add a new callback to notify
drivers when frames are going to be released.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
The flaglock in struct sta_info has long been
something that I wanted to get rid of, this
finally does the conversion to atomic bitops.The conversion itself is straight-forward in
most places, a few things needed to change a
bit since we can no longer use multiple bits
at the same time.On x86-64, this is a fairly significant code
size reduction:
text data bss dec hex
427861 23648 1008 452517 6e7a5 before
425383 23648 976 450007 6ddd7 afterSigned-off-by: Johannes Berg
Signed-off-by: John W. Linville -
If a PS-poll frame is retried (but was received)
there is no way to detect that since it has no
sequence number. As a consequence, the standard
asks us to not react to PS-poll frames until the
response to one made it out (was ACKed or lost).Implement this by using the WLAN_STA_SP flags to
also indicate a PS-Poll "service period" and the
IEEE80211_TX_STATUS_EOSP flag for the response
packet to indicate the end of the "SP" as usual.We could use separate flags, but that will most
likely completely confuse drivers, and while the
standard doesn't exclude simultaneously polling
using uAPSD and PS-Poll, doing that seems quite
problematic.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
For PS-poll, there's a possible race between
us expiring a frame and the station polling
for it -- send it a null frame in that case.For uAPSD, the standard says that we have to
send a frame in each SP, so send null if we
don't have any other frames.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
Add uAPSD support to mac80211. This is probably not
possible with all devices, so advertising it with
the cfg80211 flag will be left up to drivers that
want it.Due to my previous patches it is now a fairly
straight-forward extension. Drivers need to have
accurate TX status reporting for the EOSP frame.
For drivers that buffer themselves, the provided
APIs allow releasing the right number of frames,
but then drivers need to set EOSP and more-data
themselves. This is documented in more detail in
the new code itself.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
If there are frames for a station buffered in
the driver, mac80211 announces those in the TIM
IE but there's no way to release them. Add new
API to release such frames and use it when the
station polls for a frame.Since the API will soon also be used for uAPSD
it is easily extensible.Note that before this change drivers announcing
driver-buffered frames in the TIM bit actually
will respond to a PS-Poll with a potentially
lower priority frame (if there are any frames
buffered in mac80211), after this patch a driver
that hasn't been changed will no longer respond
at all. This only affects ath9k, which will need
to be fixed to implement the new API.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
For uAPSD support we'll need to have per-AC PS
buffers. As this is a major undertaking, split
the buffers before really adding support for
uAPSD. This already makes some reference to the
uapsd_queues variable, but for now that will
never be non-zero.Since book-keeping is complicated, also change
the logic for keeping a maximum of frames only
and allow 64 frames per AC (up from 128 for a
station).Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
mac80211 will expire normal PS-buffered frames, but
if the device rejected some frames for a sleeping
station, these won't be on the ps_tx_buf queue but
on the tx_filtered queue instead; this is done to
avoid reordering.However, mac80211 will not expire frames from the
filtered queue, let's fix that.Also add a more comments to what all this expiry is
doing and how it works.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
Currently, the TIM bit for a given station is set
and cleared all over the place. Since the logic to
set/clear it will become much more complex when we
add uAPSD support, as a first step let's collect
the entire logic in one place. This requires a few
small adjustments to other places.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville -
For uAPSD implementation, it is necessary to know on
which ACs frames are buffered. mac80211 obviously
knows about the frames it has buffered itself, but
with aggregation many drivers buffer frames. Thus,
mac80211 needs to be informed about this.For now, since we don't have APSD in any form, this
will unconditionally set the TIM bit for the station
but later with uAPSD only some ACs might cause the
TIM bit to be set.ath9k is the only driver using this API and I only
modify it in the most basic way, it won't be able
to implement uAPSD with this yet. But it can't do
that anyway since there's no way to selectively
release frames to the peer yet.Since drivers will buffer frames per TID, let them
inform mac80211 on a per TID basis, mac80211 will
then sort out the AC mapping itself.Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville
22 Sep, 2011
1 commit
-
Conflicts:
MAINTAINERS
drivers/net/Kconfig
drivers/net/ethernet/broadcom/bnx2x/bnx2x_link.c
drivers/net/ethernet/broadcom/tg3.c
drivers/net/wireless/iwlwifi/iwl-pci.c
drivers/net/wireless/iwlwifi/iwl-trans-tx-pcie.c
drivers/net/wireless/rt2x00/rt2800usb.c
drivers/net/wireless/wl12xx/main.c
21 Sep, 2011
1 commit
-
Conflicts:
drivers/net/wireless/iwlwifi/iwl-pci.c
drivers/net/wireless/iwlwifi/iwl-trans-pcie-tx.c
drivers/net/wireless/rt2x00/rt2800usb.c
drivers/net/wireless/wl12xx/main.c
20 Sep, 2011
1 commit
15 Sep, 2011
1 commit
-
Juuso optimised the timer to not run all the
time in commit 3393a608c4979a94d1887efc05b7.
However, after that it will still run once
more even if all frames just expired. Fixing
that also makes the function return value a
little clearer in the process.Also, while at it, change the return value
to bool (instead of int).Cc: Juuso Oikarinen
Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville
14 Sep, 2011
1 commit
-
Since my commit 34e895075e21be3e21e71d6317440d1ee7969ad0
("mac80211: allow station add/remove to sleep") there is
a race in mac80211 when it clears the TIM bit because a
sleeping station disconnected, the spinlock isn't held
around the relevant code any more. Use the right API to
acquire the spinlock correctly.Cc: stable@kernel.org [2.6.34+]
Signed-off-by: Johannes Berg
Signed-off-by: John W. Linville
30 Aug, 2011
1 commit
-
…wireless-next into for-davem
26 Aug, 2011
2 commits
-
When associating to an AP, the station might miss the first EAP
packet that the AP sends due to a race condition between the association
success procedure and the rx flow in mac80211.
In such cases, the packet might fall in ieee80211_rx_h_check due to
the fact that the relevant rx->sta wasn't allocated yet.
Allocation of the relevant station info struct before actually
sending the association request and setting it with a new
dummy_sta flag solve this problem.
The station will accept only EAP packets from the AP while it
is in the pre-association/dummy state.
This dummy station entry is not seen by normal sta_info_get()
calls, only by sta_info_get_bss_rx().
The driver is not notified for the first insertion of the
dummy station. The driver is notified only after the association
is complete and the dummy flag is removed from the station entry.
That way, all the rest of the code flow should be untouched by
this change.Signed-off-by: Guy Eilam
Signed-off-by: John W. Linville -
Divided the sta_info_insert_rcu function to 3 mini-functions:
sta_info_insert_check - the initial checks done when inserting
a new station
sta_info_insert_ibss - the function that handles the station
addition for IBSS interfaces
sta_info_insert_non_ibss - the function that handles the station
addition in other casesThe outer API was not changed.
The refactoring was done for better usage of the different
stages in the station addition in new scenarios added
in the next commit.Signed-off-by: Guy Eilam
Signed-off-by: John W. Linville
23 Aug, 2011
1 commit
-
…wireless-next into for-davem
Conflicts:
drivers/staging/ath6kl/miscdrv/ar3kps/ar3kpsparser.c
drivers/staging/ath6kl/os/linux/ar6000_drv.c
12 Aug, 2011
1 commit
-
Add a comment pointing out the use of enum station_info_flags for
all new struct station_info fields. In addition, memset the sinfo
buffer to zero before use on all paths in the current tree to avoid
leaving uninitialized pointers in the data.Signed-off-by: Jouni Malinen
Signed-off-by: John W. Linville
02 Aug, 2011
1 commit
-
When assigning a NULL value to an RCU protected pointer, no barrier
is needed. The rcu_assign_pointer, used to handle that but will soon
change to not handle the special case.Convert all rcu_assign_pointer of NULL value.
//smpl
@@ expression P; @@- rcu_assign_pointer(P, NULL)
+ RCU_INIT_POINTER(P, NULL)//
Signed-off-by: Stephen Hemminger
Acked-by: Paul E. McKenney
Signed-off-by: David S. Miller
09 Jul, 2011
1 commit
-
Since ca5ecddf (rcu: define __rcu address space modifier for sparse)
rcu_dereference_check use rcu_read_lock_held as a part of condition
automatically so callers do not have to do that as well.Signed-off-by: Michal Hocko
Acked-by: Paul E. McKenney
Signed-off-by: Jiri Kosina
17 May, 2011
2 commits
-
…wireless-next-2.6 into for-davem
Conflicts:
drivers/net/wireless/iwlwifi/iwl-agn-tx.c
net/mac80211/sta_info.h -
These definitions need to be exposed now that we can set the peer link
states via NL80211_ATTR_STA_PLINK_STATE. They were already being
(opaquely) reported by NL80211_STA_INFO_PLINK_STATE.Signed-off-by: Javier Cardona
Signed-off-by: John W. Linville
