summaryrefslogtreecommitdiff
AgeCommit message (Collapse)AuthorFilesLines
2014-10-27ath9k: Do not start BA when scanningSujith Manoharan1-0/+7
mac80211 currently has a race which can be hit with this sequence: * Start a scan operation. * TX BA is initiated by ieee80211_start_tx_ba_session(). * Driver sets up internal state and calls ieee80211_start_tx_ba_cb_irqsafe(). * mac80211 adds a packet to sdata->skb_queue with type IEEE80211_SDATA_QUEUE_AGG_START. * ieee80211_iface_work() doesn't process the packet because scan is in progress. * ADDBA response timer expires and the sta/tid is torn down. * Driver receives BA stop notification and calls ieee80211_stop_tx_ba_cb_irqsafe(). * This is also added to the queue by mac80211. * Now, scan finishes. At this point, the queued up packets might be processed if some other operation schedules the sdata work. Since the tids have been cleaned up already, warnings are hit. If this doesn't happen, the packets are left in the queue until the interface is torn down. Since initiating a BA session when scan is in progress leads to flaky connections, especially in MCC mode, we can drop the TX BA request. This improves connectivity with legacy clients in MCC mode. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Improve flush() in mcc modeSujith Manoharan3-6/+28
The flush timeout in MCC mode is very small, since we are constrained by the time slice for each channel context, but since only the HW queues are flushed when switching contexts, it is acceptable. Since the SW queues are also emptied in the mac80211 flush() callback, a larger duration is needed. Add an override argument to __ath9k_flush() and set it when flush() is called in MCC mode. This allows the driver to drain both the SW and HW queues. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Fix HW scan abortSujith Manoharan2-17/+13
Instead of using ATH_CHANCTX_EVENT_ASSIGN to abort a HW scan when a new interface becomes active, use the mgd_prepare_tx() callback. This allows us to make sure that the GO's channel becomes operational by using flush_work(). Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Send oneshot NoASujith Manoharan2-0/+42
This patch makes sure that a GO interface sends out a new NoA schedule with 200ms duration when mgd_prepare_tx() is called. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Fix RoC expirationSujith Manoharan1-1/+2
mac80211 has to be notified when a RoC period expires in the driver. In MCC mode, since the offchannel/RoC timer is set with the requested duration, ieee80211_remain_on_channel_expired() needs to be called when the timer expires. But, currently it is done after we move back to the operating channel. This is incorrect - fix this by calling ieee80211_remain_on_channel_expired() when the RoC timer expires and in ath_roc_complete() when the RoC request is aborted. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Use a helper function for bmissSujith Manoharan1-19/+26
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Use a helper function to set NoASujith Manoharan1-25/+36
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Use a helper function for offchannel NoASujith Manoharan1-18/+27
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Check for active GO in mgd_prepare_tx()Sujith Manoharan3-7/+54
If a GO interface is active when we receive a mgd_prepare_tx() call, then we need to send out a new NoA before switching to a new context. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Add a function to check for an active GOSujith Manoharan2-0/+31
Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Fix address managementSujith Manoharan3-5/+5
Since both the arguments need to satisfy the alignment requirements of ether_addr_copy(), use memcpy() in cases where there will be no big performance benefit and make sure that ether_addr_copy() calls use properly aligned arguments. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Send AUTHORIZED event only for station modeSujith Manoharan1-4/+6
ATH_CHANCTX_EVENT_AUTHORIZED is required to trigger the MCC scheduler when a station interface becomes authorized. But, since the driver gets station state notifications when the current operating mode is AP too, make sure that we send ATH_CHANCTX_EVENT_AUTHORIZED only when the interface is in station mode. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Check for pending frames properlySujith Manoharan3-10/+19
Pending frames in the driver can be present either in the HW queues or SW. ath9k_has_pending_frames() currently checks for the HW queues first and then checks if any ACs are queued in the driver. In MCC mode, we need to check the HW queues alone, since the SW queues are just marked as 'stopped' - they will be processed in the next context switch. But since we don't differentiate this now, mention whether we want to check if there are frames in the SW queues. * The flush() callback checks both HW and SW queues. * The tx_frames_pending() callback does the same. * The call to __ath9k_flush() in MCC mode checks HW queues alone. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Fix offchannel flush timeoutSujith Manoharan1-0/+20
An offchannel operation also needs to have a flush timeout that doesn't exceed the NoA absence duration of a GO context, so use channel_switch_time. The first offchannel operation is set a flush timeout of 10ms since channel_switch_time will be zero. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Fix MCC flush timeoutSujith Manoharan1-1/+15
In MCC mode, the duration for a channel context is half the beacon interval and having a large flush timeout will adversely affect GO operation, since the default value of 200ms will overshoot the advertised NoA absence duration. The scheduler initiates a channel context switch only when the slot duration for the current context expires, so there is no possibility of having a fixed timeout for flush. Since the channel_switch_time is added to the absence duration when the GO sets up the NoA attribute, this is the maximum time that we have to flush the TX queues. The duration is very small, but we don't have a choice in MCC mode. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Use configurable timeout for flushSujith Manoharan3-1/+10
The timeout value for flushing the TX queues is hardcoded at 200ms right now. Use a channel context-specific value instead to allow adjustments to the timeout in case MCC is enabled. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Clear NoA schedule properlySujith Manoharan1-0/+15
When an active context transitions to inactive state, the NoA schedule needs to be removed for the context that has beaconing enabled. Not doing this will affect p2p clients. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Disable beacon tasklet during resetSujith Manoharan1-0/+2
When a chip reset is done, all running timers, tasklets etc. are stopped but the beacon tasklet is left running. Fix this. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Set ATH_OP_HW_RESET before HW resetSujith Manoharan1-0/+12
When a HW reset is done, the interrupt tasklet is disabled before ISRs are disabled in the HW. This allows a small window where the HW can still generate interrupts. Since the tasklet is disabled and not killed, it is not scheduled but deferred for execution at a later time. This happens because ATH_OP_HW_RESET is not set when ath_reset() is called. When the hw_reset_work workqueue is used, this problem doesn't arise because ATH_OP_HW_RESET is set and the ISR bails out. Set ATH_OP_HW_RESET properly in ath_reset() to avoid this race - all the ath_reset_internal() callers have been converted to use ath_reset() in the previous patch. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Unify reset APISujith Manoharan4-10/+11
Instead of having ath_reset_internal() and ath_reset() as two separate calls to perform a HW reset, have one function. This makes sure that the behavior will be the same at all callsites. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Process beacons properlySujith Manoharan1-1/+6
When the current operating channel context has been marked as ATH_CHANCTX_STATE_FORCE_ACTIVE, do not process beacons that might be received, since we have to wait for the station to become authorized. Also, since the cached TSF value will be zero initially do not rearm the timer in this case when a beacon is received, since it results in spurious values. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Enable multi-channel properlySujith Manoharan3-9/+10
In MCC mode, currently the decision to enable the multi-channel state machine is done based on the association status if one of the interfaces assigned to a context is in station mode. This allows the driver to switch to the other context before the current station is able to complete the 4-way handshake in case it is required and this causes problems. Instead, enable multi-channel mode when the station moves to the authorized state. This disallows an early switch to the other channel. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: Use sta_state() callbackSujith Manoharan1-2/+26
Instead of using the sta_add()/sta_remove() callbacks, use the sta_state() callback since this gives more fine-grained control. Signed-off-by: Sujith Manoharan <c_manoha@qca.qualcomm.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ath9k: do not overwrite AR_PHY_RADAR_1 MSBLorenzo Bianconi2-2/+8
Do not overwrite AR_PHY_RADAR_1 most significant byte default value Signed-off-by: Lorenzo Bianconi <lorenzo.bianconi83@gmail.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ipw2x00: clean up a conditionDan Carpenter1-1/+1
The original condition was "(PAGE_SIZE - len)" when "(len < PAGE_SIZE)" is intended. This condition is not really sufficient, but also not really needed... If "len > PAGE_SIZE" then it we will print a warning message in dmesg but there are no other effects. Maybe we should just remove the condition? Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27ipw2x00: remove unused ->ibss_dfs pointerDan Carpenter3-38/+3
The ->ibss_dfs pointer is always allocated with a user controlled length. This caused a static checker warning because what if the length was zero? In that case, any dereference of ->ibss_dfs would lead to an Oops. It turns out that this isn't a problem because the ->ibss_dfs pointer is never used. This patch deletes it along with all the related code. In particular the entire libipw_network_reset() function can be removed. Signed-off-by: Dan Carpenter <dan.carpenter@oracle.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27rt2x00: tune multi-registers I/O timeoutStanislaw Gruszka3-36/+17
We provide timeout value to rt2x00usb_vendor_request_buff() based on number of registers to process. That value is passed down to rt2x00usb_vendor_req_buff_lock() and ends in usb_control_msg(). But we do not read/write all registers in rt2x00usb_vendor_req_buff_lock() at once. We read/write them in chunks of 64 bytes in the loop, hence passed timeout value to low level is too big. Patch removes timeout argument from rt2x00usb_vendor_request_buff() and use short REGISTER_TIMEOUT in rt2x00usb_vendor_req_buff_lock(). That timeout value should be fine for 64 bytes and smaller requests. For EEPROM read we introduced new timeout value equal to 2 seconds. Patch fixes process uninterruptible sleep stalls for long period, when USB bus has problem to satisfy a request and we wait very long time on usb_start_wait_urb(). Reported-and-tested-by: Sakari Ailus <sakari.ailus@iki.fi> Signed-off-by: Stanislaw Gruszka <sgruszka@redhat.com> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27bcma: fill core details for every deviceRafał Miłecki3-3/+8
We were setting things like dma_dev, IRQ, etc. during core registration only. We need such info for cores handled internally (e.g. ChipCommon) as well. Signed-off-by: Rafał Miłecki <zajec5@gmail.com> Acked-by: Hauke Mehrtens <hauke@hauke-m.de> Signed-off-by: John W. Linville <linville@tuxdriver.com>
2014-10-27Merge tag 'for-linville-20141024' of git://github.com/kvalo/athJohn W. Linville24-1373/+2258
Conflicts: drivers/net/wireless/ath/wil6210/wil6210.h
2014-10-27mac802154: rx: change naming conventionAlexander Aring1-16/+17
This patch changes the naming convention of mac802154 rx file. It should be more named like mac80211 stack. Furthermore we introduce a new frame parsing implementation which is much similar the mac80211 implementation. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27at86rf230: improve receive handlingAlexander Aring1-7/+3
Current behaviour it to copy the frame inclusive CRC into a skb, then remove the CRC from the skb. This patch optimizes this by not copying the CRC at the first place. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27at86rf230: use ieee802154_is_valid_psdu_len helperAlexander Aring1-18/+8
This patch adds the ieee802154_is_valid_psdu_len function to validate the psdu length. If the psdu length is invalid we use the maximum payload to receive also corrupted frames in monitor mode. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27ieee802154: add valid psdu length helperAlexander Aring1-0/+11
This patch adds a generic valid psdu length check function helper. This is useful to check the length field after receiving. For example the at86rf231 doesn't filter invalid psdu length. Sometimes the CRC can also be correct. If we get the lqi value with an invalid frame length the kernel may crash because we dereference an invalid pointer in the receive buffer. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: move rcu lockingAlexander Aring1-4/+4
Instead of twice lock and unlock mechanism this patch hold these locks only once at one position. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: move skb_reset_mac_headerAlexander Aring1-2/+3
This patch moves the skb_reset_mac_header call before frame parsing while wpan rx and before monitor deliver functionality. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: add monitor pkt_type informationAlexander Aring1-0/+1
This patch adds a PACKET_OTHERHOST setting when a monitor interface receives a skb. All receiving skb's to the monitor interface should be PACKET_OTHERHOST. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: add CHECKSUM_UNNECESSARYAlexander Aring1-0/+2
This patch adds CHECKSUM_UNNECESSARY to skb->ip_summed before delivery. There exist no transceiver with IP checksum functionality. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: move skb->protocol settingAlexander Aring1-1/+4
This patch moves the skb->protocol setting to the position when it's needed. It's only needed when frame parsing was successful. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: rename remove mac802154_subif_rxAlexander Aring1-9/+3
This patch removes the mac802154_subif_rx function and do the necessary calls inside of ieee802154_rx function. The ieee802154_rx is small enough to move the functionality inside this function. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: tx: remove monitor receive while xmitAlexander Aring3-4/+2
This removes the call of monitor receive funktion when any interface type call xmit. There exist no such use case that a monitor interface should receive the actual sending frame. One use case could be that a wpan interface and monitor interface could be running at the same time on one phy. Then the monitor interface receives the wpan frames also. Furthermore we adding support for promiscous mode setting. With promiscous mode setting we can't run a wpan and monitor interface at the same time. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: move receive handling into rx.cAlexander Aring4-214/+214
This patch removes all relevant receiving functions inclusive frame parsing into rx file. Like mac80211 we should implement the complete receive handling and parsing in this file. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: document ieee802154_rx() context requirementAlexander Aring1-0/+2
This patch is similar like d20ef63d32461332958661df73e21c0ca42601b0 ("mac80211: document ieee80211_rx() context requirement"). The netif_receive_skb call requires with softirqs disabled. This patch adds a warning if softirqs are pending while calling ieee802154_rx. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: rx: use tasklet instead workqueueAlexander Aring5-41/+47
Tasklets have much less overhead than workqueues. This patch also removes the heap allocation for the worker on receiving path. Like mac80211 we should prefer use a tasklet here instead a workqueue to getting fast out of interrupt context when ieee802154_rx_irqsafe is called by driver. Like wireless inside the tasklet context we should call netif_receive_skb instead netif_rx_ni anymore. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27ieee802154: drivers: use dev_alloc_skbAlexander Aring3-3/+3
This patch change the allocation of skb inside the ieee802154 driver layer to dev_alloc_skb. This changes also the gfp mask to GFP_ATOMIC which is needed for upcomming change that the receiving is done by a tasklet and not a workqueue anymore. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Cc: Alan Ott <alan@signal11.us> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27mac802154: tx: use put_unaligned_le16 for copy crcAlexander Aring1-2/+3
This patch replaces the memcpy with a put_unaligned_le16. The placement of crc inside of PSDU can also be unaligned. With memcpy this can fail on some architectures. Signed-off-by: Alexander Aring <alex.aring@gmail.com> Reported-by: Marcel Holtmann <marcel@holtmann.org> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27ieee802154: 6lowpan: rename process_data and lowpan_process_dataMartin Townsend4-23/+29
As we have decouple decompression from data delivery we can now rename all occurences of process_data in receive path. Signed-off-by: Martin Townsend <mtownsend1973@gmail.com> Acked-by: Alexander Aring <alex.aring@gmail.com> Acked-by: Jukka Rissanen <jukka.rissanen@linux.intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-27bluetooth:6lowpan: use consume_skb when packet processed successfullyMartin Townsend1-3/+4
Signed-off-by: Martin Townsend <mtownsend1973@gmail.com> Acked-by: Alexander Aring <alex.aring@gmail.com> Acked-by: Jukka Rissanen <jukka.rissanen@linux.intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-276lowpan: fix process_data return valuesMartin Townsend2-4/+4
As process_data now returns just error codes fix up the calls to this function to only drop the skb if an error code is returned. Signed-off-by: Martin Townsend <mtownsend1973@gmail.com> Acked-by: Alexander Aring <alex.aring@gmail.com> Acked-by: Jukka Rissanen <jukka.rissanen@linux.intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-276lowpan: remove skb_deliver from IPHCMartin Townsend4-46/+45
Separating skb delivery from decompression ensures that we can support further decompression schemes and removes the mixed return value of error codes with NET_RX_FOO. Signed-off-by: Martin Townsend <mtownsend1973@gmail.com> Acked-by: Alexander Aring <alex.aring@gmail.com> Acked-by: Jukka Rissanen <jukka.rissanen@linux.intel.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
2014-10-26mac802154: tx: make worker information staticAlexander Aring1-12/+5
This patch moves the worker information struct out of skb control block. Instead control block we declare it static inside of tx.c file. We can do that, because the worker can't be used twice at the same time. It's protected by stop and wake netdev queue. This patch fix an issue that the "struct ieee802154_xmit_cb" doesn't fit into the skb control block on some kernel configuartion reported by kbuild test robot. It was introduced by commit fe24371d6645b766c59ec664c59d0a9c310ad455 ("mac802154: tx: remove kmalloc in xmit hotpath"). Signed-off-by: Alexander Aring <alex.aring@gmail.com> Signed-off-by: Marcel Holtmann <marcel@holtmann.org>
generated by cgit v1.2.3 (git 2.39.0) at 2025-09-26 20:04:04 +0300