The netfilter conflicts were rather simple overlapping
changes.
However, the cls_tcindex.c stuff was a bit more complex.
On the 'net' side, Cong is fixing several races and memory
leaks. Whilst on the 'net-next' side we have Vlad adding
the rtnl-ness support.
What I've decided to do, in order to resolve this, is revert the
conversion over to using a workqueue that Cong did, bringing us back
to pure RCU. I did it this way because I believe that either Cong's
races don't apply with have Vlad did things, or Cong will have to
implement the race fix slightly differently.
Signed-off-by: David S. Miller <davem@davemloft.net>
When mac80211 requests the low level driver to stop an ongoing
Tx aggregation, the low level driver is expected to call
ieee80211_stop_tx_ba_cb_irqsafe() to indicate that it is ready
to stop the session. The callback in turn schedules a worker
to complete the session tear down, which in turn also handles
the relevant state for the intermediate Tx queue.
However, as this flow in asynchronous, the intermediate queue
should be stopped and not continue servicing frames, as in
such a case frames that are dequeued would be marked as part
of an aggregation, although the aggregation is already been
stopped.
Fix this by stopping the intermediate Tx queue, before
calling the low level driver to stop the Tx aggregation.
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This adds an API to mac80211 to handle scheduling of TXQs. The interface
between driver and mac80211 for TXQ handling is changed by adding two new
functions: ieee80211_next_txq(), which will return the next TXQ to schedule
in the current round-robin rotation, and ieee80211_return_txq(), which the
driver uses to indicate that it has finished scheduling a TXQ (which will
then be put back in the scheduling rotation if it isn't empty).
The driver must call ieee80211_txq_schedule_start() at the start of each
scheduling session, and ieee80211_txq_schedule_end() at the end. The API
then guarantees that the same TXQ is not returned twice in the same
session (so a driver can loop on ieee80211_next_txq() without worrying
about breaking the loop.
Usage of the new API is optional, so drivers can be ported one at a time.
In this patch, the actual scheduling performed by mac80211 is simple
round-robin, but a subsequent commit adds airtime fairness awareness to the
scheduler.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
[minor kernel-doc fix, propagate sparse locking checks out]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Add support for HE in mac80211 conforming with P802.11ax_D1.4.
Johannes: Fix another bug with the buf_size comparison in agg-rx.c.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: Ilan Peer <ilan.peer@intel.com>
Signed-off-by: Ido Yariv <idox.yariv@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Bump the IEEE80211_MAX_AMPDU_BUF size to 0x100 for HE support
and - for now - use IEEE80211_MAX_AMPDU_BUF_HT everywhere.
This is derived from my internal patch, parts of which Luca
had sent upstream.
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
2016 spec, section 10.24.2 specifies that the block ack
timeout in the ADD BA request is advisory.
That means we should check the value in the response and
act upon it (same as buffer size).
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This reverts commit e937b8da5a.
Turns out that a new driver (mt76) is coming in through
Kalle's tree, and will conflict with this. It also has some
conflicting requirements, so we'll revisit this later.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This adds an API to mac80211 to handle scheduling of TXQs and changes the
interface between driver and mac80211 for TXQ handling as follows:
- The wake_tx_queue callback interface no longer includes the TXQ. Instead,
the driver is expected to retrieve that from ieee80211_next_txq()
- Two new mac80211 functions are added: ieee80211_next_txq() and
ieee80211_schedule_txq(). The former returns the next TXQ that should be
scheduled, and is how the driver gets a queue to pull packets from. The
latter is called internally by mac80211 to start scheduling a queue, and
the driver is supposed to call it to re-schedule the TXQ after it is
finished pulling packets from it (unless the queue emptied).
The ath9k and ath10k drivers are changed to use the new API.
Signed-off-by: Toke Høiland-Jørgensen <toke@toke.dk>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Currently the restart flow enables RX back, and then proceeds
to tear down RX and TX aggregations.
The TX aggregation tear down calls synchronize_net(), which
waits for packet receiving to be done.
This is done for every session, while RX processing is already
active, and in some reproductions it takes up to 3 seconds.
Add a call once in the restart_work, before we have traffic
active again, and remove the subsequent calls when tearing
down the aggregation.
This requires to move down the code that turns off the
reconfig flag in order to be able to test it in
_ieee80211_stop_tx_ba_session().
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
There's no need to re-lookup the data structures now that
we actually get them immediately with from_timer(), just
avoid that. The struct has to be valid anyway, otherwise
the timer object itself would no longer be valid, and we
can't have a different version of the struct since only a
single session per TID is permitted.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When deleting a station or otherwise tearing down all aggregation
sessions, make sure to delete requested but not yet started ones,
to avoid the following scenario:
* session is requested, added to tid_start_tx[]
* ieee80211_ba_session_work() runs, gets past BLOCK_BA check
* ieee80211_sta_tear_down_BA_sessions() runs, locks &sta->ampdu_mlme.mtx,
e.g. while deleting the station - deleting all active sessions
* ieee80211_ba_session_work() continues since tear down flushes it, and
calls ieee80211_tx_ba_session_handle_start() for the new session, arms
the timer for it
* station deletion continues to __cleanup_single_sta() and frees the
session struct, while the timer is armed
Reported-by: Fengguang Wu <fengguang.wu@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
In preparation for unconditionally passing the struct timer_list pointer to
all timer callbacks, switch to using the new timer_setup() and from_timer()
to pass the timer pointer explicitly.
This removes the tid mapping array and expands the tid structures to
add a pointer back to the station, along with the tid index itself.
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: linux-wireless@vger.kernel.org
Cc: netdev@vger.kernel.org
Signed-off-by: Kees Cook <keescook@chromium.org>
[switch tid variables to u8, the valid range is 0-15 at most,
initialize tid_tx->sta/tid properly]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Since drv_wake_tx_queue() is normally called in the TX path, which
is already in an RCU critical section, we should call it the same
way in the aggregation code path, so if the driver expects to be
able to use RCU, it'll already be protected without having to enter
a nested critical section.
Additionally, disable soft-IRQs, since not doing so could cause
issues in a driver that relies on them already being disabled like
in the other path.
Fixes: ba8c3d6f16 ("mac80211: add an intermediate software queue implementation")
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Invoking ht_dbg() with too long of a string will print a warning.
Shorten the messages while retaining the printed patameters.
Signed-off-by: Sharon Dvir <sharon.dvir@intel.com>
Signed-off-by: Luca Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
There were many places that my previous spatch didn't find,
as pointed out by yuan linyu in various patches.
The following spatch found many more and also removes the
now unnecessary casts:
@@
identifier p, p2;
expression len;
expression skb;
type t, t2;
@@
(
-p = skb_put(skb, len);
+p = skb_put_zero(skb, len);
|
-p = (t)skb_put(skb, len);
+p = skb_put_zero(skb, len);
)
... when != p
(
p2 = (t2)p;
-memset(p2, 0, len);
|
-memset(p, 0, len);
)
@@
type t, t2;
identifier p, p2;
expression skb;
@@
t *p;
...
(
-p = skb_put(skb, sizeof(t));
+p = skb_put_zero(skb, sizeof(t));
|
-p = (t *)skb_put(skb, sizeof(t));
+p = skb_put_zero(skb, sizeof(t));
)
... when != p
(
p2 = (t2)p;
-memset(p2, 0, sizeof(*p));
|
-memset(p, 0, sizeof(*p));
)
@@
expression skb, len;
@@
-memset(skb_put(skb, len), 0, len);
+skb_put_zero(skb, len);
Apply it to the tree (with one manual fixup to keep the
comment in vxlan.c, which spatch removed.)
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: David S. Miller <davem@davemloft.net>
When starting or stopping an aggregation session, one of the steps
is that the driver calls back to mac80211 that the start/stop can
proceed. This is handled by queueing up a fake SKB and processing
it from the normal iface/sdata work. Since this isn't flushed when
disassociating, the following race is possible:
* associate
* start aggregation session
* driver callback
* disassociate
* associate again to the same AP
* callback processing runs, leading to a WARN_ON() that
the TID hadn't requested aggregation
If the second association isn't to the same AP, there would only
be a message printed ("Could not find station: <addr>"), but the
same race could happen.
Fix this by not going the whole detour with a fake SKB etc. but
simply looking up the aggregation session in the driver callback,
marking it with a START_CB/STOP_CB bit and then scheduling the
regular aggregation work that will now process these bits as well.
This also simplifies the code and gets rid of the whole problem
with allocation failures of said skb, which could have left the
session in limbo.
Reported-by: Jouni Malinen <j@w1.fi>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Use setup_timer() and setup_deferrable_timer() to set the data and
function timer fields. It makes the code cleaner and will allow for
easier change of the timer struct internals.
Signed-off-by: Ondřej Lysoněk <ondrej.lysonek@seznam.cz>
Signed-off-by: Jiri Slaby <jslaby@suse.cz>
Cc: Johannes Berg <johannes@sipsolutions.net>
Cc: "David S. Miller" <davem@davemloft.net>
Cc: <linux-wireless@vger.kernel.org>
Cc: <netdev@vger.kernel.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Since mac80211 doesn't currently support TSIDs 8-15 which can
only be used after QoS TSPEC negotiation (and not even after
WMM negotiation), reject attempts to set up aggregation
sessions for them, which might confuse drivers. In mac80211
we do correctly handle that, but the TSIDs should never get
used anyway, and drivers might not be able to handle it.
Cc: stable@vger.kernel.org
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
mac80211's software queues were designed to work
very closely with device tx queues. They are
required to make use of 802.11 packet aggregation
easily and efficiently.
Due to the way 802.11 aggregation is designed it
only makes sense to keep fair queuing as close to
hardware as possible to reduce induced latency and
inertia and provide the best flow responsiveness.
This change doesn't translate directly to
immediate and significant gains. End result
depends on driver's induced latency. Best results
can be achieved if driver keeps its own tx
queue/fifo fill level to a minimum.
Signed-off-by: Michal Kazior <michal.kazior@tieto.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Requires software tx queueing and fast-xmit support. For good
performance, drivers need frag_list support as well. This avoids the
need for copying data of aggregated frames. Running without it is only
supported for debugging purposes.
To avoid performance and packet size issues, the rate control module or
driver needs to limit the maximum A-MSDU size by setting
max_rc_amsdu_len in struct ieee80211_sta.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
[fix locking issue]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Currently mac80211 does not inform the driver of the session
block ack timeout when starting a rx aggregation session.
Drivers that manage the reorder buffer need to know this
parameter.
Seeing that there are now too many arguments for the
drv_ampdu_action() function, wrap them inside a structure.
Signed-off-by: Sara Sharon <sara.sharon@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Advertising reordering window in ADDBA less than 64 can crash some APs,
an example is LinkSys WRT120N (with FW v1.0.07 build 002 Jun 18 2012).
On the other hand, a driver may need to limit Tx A-MPDU size for its own
reasons, like specific HW limitations.
Signed-off-by: Gregory Greenman <gregory.greenman@intel.com>
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Advertise the capability to send A-MSDU within A-MPDU
in the AddBA request sent by mac80211. Let the driver
know about the peer's capabilities.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
As we're running out of hardware capability flags pretty quickly,
convert them to use the regular test_bit() style unsigned long
bitmaps.
This introduces a number of helper functions/macros to set and to
test the bits, along with new debugfs code.
The occurrences of an explicit __clear_bit() are intentional, the
drivers were never supposed to change their supported bits on the
fly. We should investigate changing this to be a per-frame flag.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
This allows drivers to request per-vif and per-sta-tid queues from which
they can pull frames. This makes it easier to keep the hardware queues
short, and to improve fairness between clients and vifs.
The task of scheduling packet transmission is left up to the driver -
queueing is controlled by mac80211. Drivers can only dequeue packets by
calling ieee80211_tx_dequeue. This makes it possible to add active queue
management later without changing drivers using this code.
This can also be used as a starting point to implement A-MSDU
aggregation in a way that does not add artificially induced latency.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
[resolved minor context conflict, minor changes, endian annotations]
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
If a peer or some local agent (rate control, ...) decides to start
an aggregation session but doesn't support HT (which also implies
QoS), reject it.
This is mostly a corner case as such peers normally won't try to
use block-ack sessions and rate control wouldn't start them, but
technically QoS stations could request it according to the spec.
However, since drivers don't really support such non-HT sessions
it's better to reject them.
Also, while at it, move the tracing for TX sessions earlier so it
captures the error cases as well.
Reviewed-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The rate control locking caused a potential deadlock here due to the
locks being acquired in different orders, so that change cannot yet
be applied. However, there's no fundamental reason for this code to
hold the sta->lock while transmitting frames.
Clearly it's better not to hold the lock for longer periods of time,
which can happen here since we call all the way down to the driver.
Change the code a bit to not hold it while doing that.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
In TDLS (e.g., TDLS off-channel) there is a requirement for
some drivers to supply an unused TID between the AP and the
device to the FW, to allow sending PTI requests and to allow
the FW to aggregate on a specific TID for better throughput.
To ensure that the allocated TID is indeed unused, this patch
introduces an API for blocking the driver from TXing on that
TID.
Signed-off-by: Liad Kaufman <liad.kaufman@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Use the currently existing APIs between mac80211 and the low
level driver to implement WMM admission control.
The low level driver needs to report the media time used by
each transmitted packet in ieee80211_tx_status. Based on that
information, mac80211 will modify the QoS parameters of the
admission controlled Access Category when the limit is
reached. Once the original QoS parameters can be restored,
mac80211 will do so.
One issue with this approach is that management frames will
also erroneously be downgraded, but the upside is that the
implementation is simple. In the future, it can be extended
to driver- or device-based implementations that are better.
Signed-off-by: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Sometimes different vifs may be stopping the queues for the same
reason (e.g. when several interfaces are performing a channel switch).
Instead of using a bitmask for the reasons, use an integer that holds
a refcount instead. In order to keep it backwards compatible,
introduce a boolean in some functions that tell us whether the queue
stopping should be refcounted or not. For now, use not refcounted for
all calls to keep it functionally the same as before.
Signed-off-by: Luciano Coelho <luciano.coelho@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
ATM, {ADD,DEL}BA and BAR frames are sent on the AC matching the TID of
the BA parameters. In the discussion [1] about this patch, Johannes
recalled that it fixed some races with the DELBA and indeed this
behavior was introduced in [2].
While [2] is right for the BARs, the part queueing the {ADD,DEL}BAs on
their BA params TID AC violates the spec and is more a workaround for
some drivers. Helmut expressed some concerns wrt such drivers, in
particular DELBAs in rt2x00.
ATM, DELBAs are sent after a driver has called (hence "purposely")
ieee80211_start_tx_ba_cb_irqsafe and Johannes and Emmanuel gave some
details wrt intentions behind the split of the IEEE80211_AMPDU_TX_STOP_*
given to the driver ampdu_action supposed to call this function, which
could prove handy to people trying to do the right thing in faulty
drivers (if their fw/hw don't get in their way).
[1] http://mid.gmane.org/1390391564-18481-1-git-send-email-karl.beldan@gmail.com
[2] Commit: cf6bb79ad8 ("mac80211: Use appropriate TID for sending BAR, ADDBA and DELBA frames")
Signed-off-by: Karl Beldan <karl.beldan@rivierawaves.com>
Cc: Helmut Schaa <helmut.schaa@googlemail.com>
Cc: Emmanuel Grumbach <emmanuel.grumbach@intel.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
For drivers that don't actually flush their queues when
aggregation stop with the IEEE80211_AMPDU_TX_STOP_FLUSH
or IEEE80211_AMPDU_TX_STOP_FLUSH_CONT reasons is done,
like iwlwifi or iwlegacy, mac80211 can then transmit on
a TID that the driver still considers busy. This happens
in the following way:
- IEEE80211_AMPDU_TX_STOP_FLUSH requested
- driver marks TID as emptying
- mac80211 removes tid_tx data, this can copy packets
to the TX pending queues and also let new packets
through to the driver
- driver gets unexpected TX as it wasn't completely
converted to the new API
In iwlwifi, this lead to the following warning:
WARNING: at drivers/net/wireless/iwlwifi/dvm/tx.c:442 iwlagn_tx_skb+0xc47/0xce0
Tx while agg.state = 4
Modules linked in: [...]
Pid: 0, comm: kworker/0:0 Tainted: G W 3.1.0 #1
Call Trace:
[<c1046e42>] warn_slowpath_common+0x72/0xa0
[<c1046f13>] warn_slowpath_fmt+0x33/0x40
[<fddffa17>] iwlagn_tx_skb+0xc47/0xce0 [iwldvm]
[<fddfcaa3>] iwlagn_mac_tx+0x23/0x40 [iwldvm]
[<fd8c98b6>] __ieee80211_tx+0xf6/0x3c0 [mac80211]
[<fd8cbe00>] ieee80211_tx+0xd0/0x100 [mac80211]
[<fd8cc176>] ieee80211_xmit+0x96/0xe0 [mac80211]
[<fd8cc578>] ieee80211_subif_start_xmit+0x348/0xc80 [mac80211]
[<c1445207>] dev_hard_start_xmit+0x337/0x6d0
[<c145eee9>] sch_direct_xmit+0xa9/0x210
[<c14462c0>] dev_queue_xmit+0x1b0/0x8e0
Fortunately, solving this problem is easy as the station
is being destroyed, so such transmit packets can only
happen due to races. Instead of trying to close the race
just let the race not reach the drivers by making two
changes:
1) remove the explicit aggregation session teardown in
the managed mode code, the same thing will be done
when the station is removed, in __sta_info_destroy.
2) When aggregation stop with AGG_STOP_DESTROY_STA is
requested, leave the tid_tx data around as stopped.
It will be cleared and freed in cleanup_single_sta
later, but until then any racy packets will be put
onto the tid_tx pending queue instead of transmitted
which is fine since the station is being removed.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
A lot of the aggregation messages don't indicate the
station so they're hard to understand if there are
multiple sessions in progress. Make that easier by
adding the MAC address to most messages. Also add
the TID if it wasn't already there.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
When TX aggregation is stopped, there are a few
different cases:
- connection with the peer was dropped
- session stop was requested locally
- session stop was requested by the peer
- connection was dropped while a session is stopping
The behaviour in these cases should be different, if
the connection is dropped then the driver should drop
all frames, otherwise the frames may continue to be
transmitted, aggregated in the case of a locally
requested session stop or unaggregated in the case of
the peer requesting session stop.
Split these different cases so that the driver can
act accordingly; however, treat local and remote stop
the same way and ask the driver to not send frames as
aggregated packets any more.
In the case of connection drop, the stop callback the
driver is otherwise supposed to call is no longer
required.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Create the function ieee80211_remove_tid_tx to call
it from ___ieee80211_stop_tx_ba_session later.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
The initiator/tx doesn't really identify why an
aggregation session is stopped, give a reason
for stopping that more clearly identifies what's
going on. This will help tell the driver clearly
what is expected of it.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
New drivers that might not support ampdu_action yet while in
development cause a lot of warnings, use WARN_ON_ONCE instead.
Signed-off-by: T Krushna Chaitanya <chaitanyatk@posedge.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Introduce IEEE80211_NUM_TIDS in the generic 802.11
header file and use it in place of STA_TID_NUM and
NUM_RX_DATA_QUEUES which are both really the number
of TIDs.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Because ieee80211_tx_status in status.c checks if
outgoing BlockAck requests have been acked, it is
necessary to tell the driver that tx feedback for
this sort of frame is important.
Otherwise, the stack will continue to send the same
BlockAck request over and over, which can cause
the receiver to flush or clean its reorder buffer
over and over.
Signed-off-by: Christian Lamparter <chunkeey@googlemail.com>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
There are a few things that make the logging and
debugging in mac80211 less useful than it should
be right now:
* a lot of messages should be pr_info, not pr_debug
* wholesale use of pr_debug makes it require *both*
Kconfig and dynamic configuration
* there are still a lot of ifdefs
* the style is very inconsistent, sometimes the
sdata->name is printed in front
Clean up everything, introducing new macros and
separating out the station MLME debugging into
a new Kconfig symbol.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Simplify the use of #ifdef CONFIG_MAC80211_HT_DEBUG/#endif
by adding a logging macro to encapsulate the test.
Convert the appropriate uses too.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Standardize the debugging to be able to use dynamic_debug.
Coalesce formats, align arguments.
Signed-off-by: Joe Perches <joe@perches.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
In commit 12d3952fc4
("mac80211: optimize aggregation session timeout handling")
two bugs were introduced:
1) RCU usage was completely broken since no locks are held
2) the timer must not rearm when agg session is stopping
Reported-and-tested-by: Larry Finger <Larry.Finger@lwfinger.net>
Cc: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
mac80211 currently only supports one hardware queue
per AC. This is already problematic for off-channel
uses since if we go off channel while the BE queue
is full and then try to send an off-channel frame
the frame will never go out. This will become worse
when we support multi-channel since then a queue on
one channel might be full, but we have to stop the
software queue for all channels. That is obviously
not desirable.
To address this problem allow drivers to register
more hardware queues, and allow them to map them to
virtual interfaces. When they stop a hardware queue
the corresponding AC software queues on the correct
interfaces will be stopped as well. Additionally,
there's an off-channel queue to solve that problem
and a per-interface after-DTIM beacon queue. This
allows drivers to manage software queues closer to
how the hardware works.
Currently, there's a limit of 16 hardware queues.
This may or may not be sufficient, we can adjust it
as needed.
Signed-off-by: Johannes Berg <johannes.berg@intel.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Calling mod_timer from the rx/tx hotpath is somewhat expensive, and the
timeout doesn't need to be so precise.
Switch to a different strategy: Schedule the timer initially, store jiffies
of all last rx/tx activity which would previously modify the timer, and
let the timer re-arm itself after checking the last rx/tx timestamp.
Make the session timers deferrable to avoid causing extra wakeups on systems
running on battery.
This visibly reduces CPU load under high network load on small embedded
systems.
Signed-off-by: Felix Fietkau <nbd@openwrt.org>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently code allows three (HT_AGG_MAX_RETRIES) unanswered addba
requests. When this limit is reached aggregation is turned off for
given TID permanently. This doesn't seem right: three requests is
not that much, some 'blackout' can happen, but effect of it affects
whole connection indefinitely.
This patch increases number of retries to 15. Also, when there have
been 3 or more retries it splits further retries apart by 15 seconds
instead of sending them in very short period of time.
Signed-off-by: Nikolay Martynov <mar.kolya@gmail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
Currently BAR, ADDBA and DELBA frames are always sent using AC_VO. If
the TID for which a BA session is established is assigned to a different
queue BAR, ADDBA and DELBA frames can "overtake" frames of the according
BA session.
Hence, always put BA session related frames into the same queue as the
BA sessions data frames.
Signed-off-by: Helmut Schaa <helmut.schaa@googlemail.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
David S. Miller