- 18 Nov, 2009 40 commits
-
-
Janne Ylalehto authored
Filter out unwanted events to reduce wakeups. Signed-off-by:
Janne Ylalehto <janne.ylalehto@nokia.com> Reviewed-by:
Vidhya Govindan <vidhya.govindan@nokia.com> Reviewed-by:
Juuso Oikarinen <juuso.oikarinen@nokia.com> Signed-off-by:
Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by:
Kalle Valo <kalle.valo@nokia.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Janne Ylalehto authored
Add support for IRQ looping. Helps in the case that we have e.g. multiple packets coming from the network when we wake up from the ELP. Signed-off-by:
-
Vidhya Govindan authored
This patch increases the number of beacons to be missed before generating SYNC TIMEOUT event. It increases the beacon timeout period to 500 microseconds, which gives enough time for the firmware to receive probe response or beacon. Also handled the regain event from firmware once it receives a probe response or beacon. Signed-off-by:
-
Vidhya Govindan authored
This patch sets the dtim period obtained from the mac80211 to firmware. Signed-off-by:
-
Vidhya Govindan authored
The dtim period obtained from the mac80211 is not set to the firmware. This patch implements the acx command to set correct tbtt and dtim value to the firmware. Signed-off-by:
-
Kalle Valo authored
In ps-poll template aid bits 14 and 15 were not masked as required by the standard. Mask them so that aid is sent in correct format. Signed-off-by:
-
Kalle Valo authored
Earlier firmware was stored to a memory area allocated with kmalloc() but finding a a contiguous area of memory long enough for the firmware is very difficult in certain cases. better to allocate the memory for firmware with vmalloc() instead and use a small buffer for DMA transfers. Thanks to Eero Tamminen for the idea. Signed-off-by:
-
Juuso Oikarinen authored
Implement (slightly) delayed entry into ELP. This will cure several problems: - It works around a firmware race condition if ELP is entered too fast after commands (resulting in ELP timeout -traces) - It will reduce the number of sleep-wake cycles between already scheduled events such as interrupts and tx, hence improving performance (less delay in switching between RX and TX) Signed-off-by: -
Juuso Oikarinen authored
Enable beacon filtering when PSM is enabled Signed-off-by:
-
Juuso Oikarinen authored
Enable beacon filtering with the mac80211 stack. Signed-off-by:
-
Juuso Oikarinen authored
Add configuration for connection monitor (number of allowed beacons, and timeout after last received beacon.) Signed-off-by:
-
Thomas Klute authored
At least two revisions of the D-Link DWA 160 exist, called A1 and A2. A1 (USB-ID 07d1:3c10) is already listed in usb.c as D-Link DWA 160A. A2 (USB-ID 07d1:3a09) works if added to ar9170_usb_ids. I didn't do much testing until now, but I was able to connect to APs using WPA or WEP and transmit data. Summary: * Add model revision number to the comment for D-Link DWA 160 A1 (07d1:3c10) * Add support for D-Link DWA 160 A2 (07d1:3a09) Signed-off-by:
Thomas Klute <thomas2.klute@uni-dortmund.de> Signed-off-by:
-
Xose Vazquez Perez authored
taken from windows inf file (09/15/2009, 1.04.07.0000) Signed-off-by:
Xose Vazquez Perez <xose.vazquez@gmail.com> Acked-by:
Ivo van Doorn <IvDoorn@gmail.com> Signed-off-by:
-
Johannes Berg authored
The netdev broadcast address cannot change from all-ones so there's no need to use it; we can instead hard-code it. Since we already have an instance in tkip.c, which will be shared if it is marked static const, doing this reduces text size at no data/bss cost. The real motivation for this is, of course, the desire to get rid of almost all uses of netdevs in mac80211 so that auditing their use becomes easier. Signed-off-by:
Johannes Berg <johannes@sipsolutions.net> Signed-off-by:
-
Johannes Berg authored
If we move the rcu sections a little, there's no need to touch the device refcount. Signed-off-by:
-
Johannes Berg authored
Some code currently assumes that there's a valid rate pointer even in the HT case, but there can't be. To reduce reliance on that, remove the rate pointer from the RX data struct and pass it where it's needed. Also, for now, in radiotap announce HT frames as having a DYN channel type, and remove their rate from cooked monitor radiotap completely (it isn't present in the regular monitor radiotap either.) Signed-off-by:
-
Johannes Berg authored
The RX data contains the netdev, which is duplicated since we have the sdata, and the RX status pointer, which is duplicate since we have the skb. Remove those two fields to have fewer fields that depend on each other and simply load them as necessary. Signed-off-by:
-
Johannes Berg authored
The reorder buffer handling is written in a quite peculiar style (especially comments) and also has a quirk where it invokes the entire reorder code in ieee80211_sta_manage_reorder_buf() for just a handful of lines in it with a special argument. Split out ieee80211_release_reorder_frames which can then be invoked from BAR handling and other reordering code, clean up code and comments and remove function arguments that are now unused from ieee80211_sta_manage_reorder_buf(). Signed-off-by:
-
Johannes Berg authored
When we receive a michael MIC failure report from the hardware we currently do not check whether it is actually reported on a frame that is destined to us. It shouldn't be possible to get a michael MIC failure report on other frames, but it also doesn't hurt to verify. Also, since we then don't need the station struct that early, move looking it up a bit later in the RX path. Finally, while at it, a few code cleanups in the area. Signed-off-by:
-
Johannes Berg authored
The entire aggregation code currently operates on the hw pointer and station addresses, but that needs to change to make stations purely per-vif; As one step preparing for that make the aggregation code callable with the station, or by the combination of virtual interface and station address. Signed-off-by:
-
Johannes Berg authored
Instead of filtering by device, directly look up by sdata. Signed-off-by:
-
Sujith authored
Check for AR5416 ver 1.0 before calibrating 3 chains for multi-chain. This is a WAR for calibration failure. Signed-off-by:
Sujith <Sujith.Manoharan@atheros.com> Signed-off-by:
-
Sujith authored
ATH9K_ANT_VARIABLE is the default diversity control used. Consequently ath9k_hw_decrease_chain_power() does nothing. ath9k_hw_setantennaswitch() is unused too. Also, gbeacon_rate is unused. Signed-off-by:
-
Sujith authored
axq_linkbuf, axq_aggr_depth, axq_lastdsWithCTS and axq_gatingds are unused. Signed-off-by:
-
Sujith authored
* Remove a code chunk dealing with operating mode changes. As noted, all such policy changes are to be done in add_interface. * Remove pointless check for empty BSSID. Also, remove mode checks - mac80211 does all the needed checks. * Handle enabling/disabling beacon transmission properly. * Handle beacon interval changes for AP mode. The original code depended on config_interface() to update the HW TSF. Since that callback has been removed, handle it properly. * Remove unneeded code dealing with key/privacy. * Set the chainmasks to 1x1 for IBSS when the BSSID is set. This was happening uncondionally before. Signed-off-by:
-
Felix Fietkau authored
This patch removes the need for separately allocated private tx info data in ath9k and brings the driver one small step closer to using the mac80211 rate control API properly. Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
-
Felix Fietkau authored
In order to handle association and authentication in AP mode, hostapd needs access to the tx status info of its own frames through a cooked monitor interface. Without this patch the cooked monitor interfaces also passed on tx status information for packets from other virtual interfaces. This creates a significant performance issue on embedded system. Hostapd tries to work around this by installing a Linux Socket Filter that only captures the frames it's interested in, however data duplication and socket filter matching still uses up enough CPU cycles to be very noticeable on small systems. This patch ensures that tx status information of non-injected frames does not make it to cooked monitor interfaces. Signed-off-by:
-
Christian Lamparter authored
This patch enables the driver to process all incoming dupofdm-modulated frames when operating in HT40 mode. Signed-off-by:
Christian Lamparter <chunkeey@googlemail.com> Signed-off-by:
-
Ben Cahill authored
Also regroup CSR_EEPROM and CSR_OTP bit field definitions. Signed-off-by:
Ben Cahill <ben.m.cahill@intel.com> Signed-off-by:
Reinette Chatre <reinette.chatre@intel.com> Signed-off-by:
-
Johannes Berg authored
Using the new mac80211 functionality, this makes iwlwifi handle unicast PS buffering correctly. The device works like this: * when a station goes to sleep, the microcode notices this and marks the station as asleep * when the station is marked asleep, the microcode refuses to transmit to the station and rejects all frames queued to it with the failure status code TX_STATUS_FAIL_DEST_PS (a previous patch handled this correctly) * when we need to send frames to the station _although_ it is asleep, we need to tell the ucode how many, and this is asynchronous with sending so we cannot just send the frames, we need to wait for all other frames to be flushed, and then update the counter before sending out the poll response frames. This is handled partially in the driver and partially in mac80211. In order to do all this correctly, we need to * keep track of how many frames are pending for each associated client station (avoid doing it for other stations to avoid the atomic ops) * tell mac80211 that we driver-block the PS status while there are still frames pending on the queues, and once they are all rejected (due to the dest sta being in PS) unblock mac80211 Signed-off-by:
-
Johannes Berg authored
This field was marked as reserved before since we didn't use it, but is present in all released firmwares afaict. We're going to need it soon, so add it now. Signed-off-by:
-
Johannes Berg authored
When a frame is sent to a sleeping station, the microcode reports TX_STATUS_FAIL_DEST_PS as its status -- we need to translate that to the flag that mac80211 expects. Signed-off-by:
-
Daniel C Halperin authored
Handle BSS_CHANGED_BEACON_ENABLED to enable the sending of beacons. Also set the correct HT RXON and QoS config. Signed-off-by:
Daniel C Halperin <daniel.c.halperin@intel.com> Signed-off-by:
-
Daniel C Halperin authored
When sending beacon commands to the uCode, we must inform it of the offset in the beacon frame of the TIM Element so it can transmit packets from the correct queue. This functionality is implemented in iwl_set_beacon_tim(). Fix a bug setting the rate_n_flags for the beacon packet. First, it should not use the station table's rate (it's a management frame), and second it needs to properly configure the TX antennas. Finally, also, clean up and comment relevant functions. Signed-off-by:
-
Reinette Chatre authored
A recent change optimized the power usage by the device by only powering it up during EEPROM load if it is required (for OTP devices). This change causes an error on the 1000 series devices during module load. The error looks as follows: [ 1624.024524] iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux, 1.3.27kds [ 1624.024527] iwlagn: Copyright(c) 2003-2009 Intel Corporation [ 1624.024711] iwlagn 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [ 1624.024749] iwlagn 0000:01:00.0: setting latency timer to 64 [ 1624.024909] iwlagn 0000:01:00.0: Detected Intel Wireless WiFi Link 1000 Series BGN REV=0x6C [ 1624.081263] iwlagn 0000:01:00.0: MAC is in deep sleep!. CSR_GP_CNTRL = 0x080003D8 [ 1624.092967] iwlagn 0000:01:00.0: OTP is empty [ 1624.092988] iwlagn 0000:01:00.0: Unable to init EEPROM [ 1624.093033] iwlagn 0000:01:00.0: PCI INT A disabled [ 1624.093065] iwlagn: probe of 0000:01:00.0 failed with error -2 Adding a dump_stack() to where that error is printed shows the following: [ 1624.024524] iwlagn: Intel(R) Wireless WiFi Link AGN driver for Linux, 1.3.27kds [ 1624.024527] iwlagn: Copyright(c) 2003-2009 Intel Corporation [ 1624.024711] iwlagn 0000:01:00.0: PCI INT A -> GSI 16 (level, low) -> IRQ 16 [ 1624.024749] iwlagn 0000:01:00.0: setting latency timer to 64 [ 1624.024909] iwlagn 0000:01:00.0: Detected Intel Wireless WiFi Link 1000 Series BGN REV=0x6C [ 1624.081263] iwlagn 0000:01:00.0: MAC is in deep sleep!. CSR_GP_CNTRL = 0x080003D8 [ 1624.081263] Pid: 3073, comm: work_for_cpu Tainted: G W 2.6.31.5 #4 [ 1624.081263] Call Trace: [ 1624.081263] [<ffffffffa02395db>] T.726+0x22b/0x420 [iwlcore] [ 1624.081263] [<ffffffffa023985a>] iwlcore_eeprom_acquire_semaphore+0x8a/0x190 [iwlcore] [ 1624.081263] [<ffffffff81110c94>] ? __kmalloc+0x194/0x1c0 [ 1624.081263] [<ffffffffa02391f5>] ? iwlcore_eeprom_verify_signature+0x25/0xf0 [iwlcore] [ 1624.081263] [<ffffffffa0239c67>] iwl_eeprom_init+0x107/0xf40 [iwlcore] [ 1624.081263] [<ffffffffa026ab9c>] ? iwl_prepare_card_hw+0x11c/0x470 [iwlagn] [ 1624.081263] [<ffffffff8127e2a4>] ? pci_bus_write_config_byte+0x64/0x80 [ 1624.081263] [<ffffffffa026b1f8>] iwl_pci_probe+0x308/0xac0 [iwlagn] [ 1624.081263] [<ffffffff810710a0>] ? do_work_for_cpu+0x0/0x30 [ 1624.081263] [<ffffffff81284912>] local_pci_probe+0x12/0x20 [ 1624.081263] [<ffffffff810710b3>] do_work_for_cpu+0x13/0x30 [ 1624.081263] [<ffffffff81075826>] kthread+0xa6/0xb0 [ 1624.081263] [<ffffffff81012fea>] child_rip+0xa/0x20 [ 1624.081263] [<ffffffff81075780>] ? kthread+0x0/0xb0 [ 1624.081263] [<ffffffff81012fe0>] ? child_rip+0x0/0x20 [ 1624.092967] iwlagn 0000:01:00.0: OTP is empty [ 1624.092988] iwlagn 0000:01:00.0: Unable to init EEPROM [ 1624.093033] iwlagn 0000:01:00.0: PCI INT A disabled [ 1624.093065] iwlagn: probe of 0000:01:00.0 failed with error -2 We know that the routines in this trace, iwlcore_eeprom_acquire_semaphore and iwlcore_eeprom_verify_signature, only access CSR registers and thus do not need the device to be awake if it is EEPROM. But for OTP it is required for the device to be awake to read these registers. Ensure device is awake before accessing these registers. Signed-off-by:
-
Wey-Yi Guy authored
Align the format for tx_statistics and rx_statistics debugfs output for better readability Signed-off-by:
Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by:
-
Wey-Yi Guy authored
In both 6x00 and 6x50 series, the enhanced/extended tx power table in EEPROM is used to set the max. tx power limit. This new tx power table is in 1/2 dBm format, which creates an issue of possibility of 1/2 dBm loss when driver set the tx power limit; because of driver keep track and report the tx power in dBm format. In order to prevent the 1/2 dBm loss, keep track of the true max tx power in 1/2 dBm format in driver; do the comparison and adjust the tx power if needed when send tx power command to uCode. Signed-off-by:
-
Ben Cahill authored
Power-saving logic will not re-issue a POWER_TABLE_CMD if a new command matches the prior one. This can be bad if we re-start the device due to e.g. uCode error; the new POWER_TABLE_CMD (required to invoke power-saving) may match the prior POWER_TABLE_CMD issued before the uCode error. Ensure the POWER_TABLE_CMD is sent to device when uCode is up. Signed-off-by:
-
Wey-Yi Guy authored
When issue REPLY_STATISTICS_CMD to uCode, two possible flag can be set in the configuration flags bit 0: Clear statistics 0: Do not clear Statistics counters 1: Clear to zero Statistics counters Allow "clear" parameter to be set from the caller. Add debugfs file to clear the statistics counters to help monitor and debug the uCode behavior. Signed-off-by:
-
