- 23 Mar, 2010 17 commits
-
-
Teemu Paasikivi authored
Changed driver to keep MMC host claimed while wlan interface is up. This keeps MMC stack from powering the SDIO device down. Signed-off-by:
Teemu Paasikivi <ext-teemu.3.paasikivi@nokia.com> Reviewed-by:
Juuso Oikarinen <juuso.oikarinen@nokia.com> Signed-off-by:
Luciano Coelho <luciano.coelho@nokia.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Teemu Paasikivi authored
Added method for wlan power control to io_ops struct and moved wl1271_power_on and wl1271_power_off functions to wl1271_io.h. Signed-off-by:
-
Juuso Oikarinen authored
The per-queue TX opportunity value is configured by the mac80211 in units of 32us. The firmware however wants it in us, so add the conversion. Signed-off-by:
Kalle Valo <kalle.valo@nokia.com> Signed-off-by:
-
Sujith authored
Features: * Station mode * IBSS mode * Monitor mode * Legacy support * HT support * TX/RX 11n Aggregation * HW encryption * LED * Suspend/Resume For more information: http://wireless.kernel.org/en/users/Drivers/ath9k_htcSigned-off-by:
Sujith <Sujith.Manoharan@atheros.com> Signed-off-by:
Vasanthakumar Thiagarajan <vasanth@atheros.com> Signed-off-by:
Senthil Balasubramanian <senthilkumar@atheros.com> Signed-off-by:
-
Sujith authored
Without this you will get a panic if the device initialization fails. Also, free ath_hw instance properly. ath9k_hw_deinit() shouldn't do it. Cc: stable@kernel.org Signed-off-by:
Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by:
-
Sujith authored
AR9271 needs a full reset only upon the first reset, add a call for the driver to enable these special resets. We can optimize this out later without an export. Signed-off-by:
-
Sujith authored
When initializing the PLL on AR9271 we always need to set the core clock to 117MHz. While at it remove the baud rate settings for the serial device on the AR9271, the default settings work well unless you want to customize it. Signed-off-by:
-
Sujith authored
Periodic power amplifier offset calibration is skipped on ath9k algorithmically, this is required on AR9271. Signed-off-by:
-
Sujith authored
Noisefloor values read on AR9271 are unreliable if they are less than -114, set those statically to -116. Signed-off-by:
-
Sujith authored
Noisefloor calibration involves querying hardware for samples and storing this information on a history buffer in hardware for actual noisefloor calibration processing in hardware. The history buffer supports collecting information for all Atheros hardware, one history buffer slot for each chain on each channel used for MIMO operation. For current hardware this means one history buffer slot for each chain on both the control (or primary) channel and the extension (or secondary) channel. We know which noisefloor registers to poke for collecting noisefloor data through the chainmask. For AR9285 and AR9271 devices, both 1x1, the chaimmask is defined as 0x9 = 0b0001001. The first four bits represent each chain out of a maximum of 4 chains [0-3] on the primary channel. The last four bits represent each chain on the extension channel. A chainmask of 0x9 therefore indicates chain 1 is active on both the primary and the extension channel. AR9271 only requires collecting and storing noisefloor history buffer data for the first chain on both the control and extension channel (nfarray[0] and nfarray[3]) so fix the code and avoid which reads and writes to the history buffer for the other chains. Since the noisefloor varies depending on the number of chains your device supports also initialize the noisefloor history buffer with reasonable values seen on 1x1 devices such as AR9285. Signed-off-by:
-
Sujith authored
After telling the AR9271 to go into full sleep we do not need to clear the RTC reset signal. Signed-off-by:
-
Sujith authored
The TX descriptors setup for AR971 requires the same setup as AR9285, so use that. Signed-off-by:
-
Sujith authored
The chip test is not required for AR9271 on the host driver code as the firmware will do the test internally on its own. Signed-off-by:
-
Sujith authored
Assign the proper number of GPIO pins for AR9271. Signed-off-by:
-
Sujith authored
Update the register initialization values for AR9271. This is based on our last review from our systems team. Signed-off-by:
-
Ming Lei authored
From e74b075cdb143d45be9b371ee8a8e2dcfc15ab34 Mon Sep 17 00:00:00 2001 From: Ming Lei <tom.leiming@gmail.com> Date: Sat, 27 Feb 2010 23:50:54 +0800 Subject: [PATCH] ath9k: decrease size of ath9k.ko The patch defines the fields of 'valid_single_stream' and 'valid' in struct ath_rate_table as char type, so decrease the size of ath9k.ko about 2KB. old ath9k.ko [tom@tom-lei ath9k]$ size ath9k.ko text data bss dec hex filename 69344 3080 168 72592 11b90 ath9k.ko new ath9k.ko [tom@tom-lei ath9k]$ size ath9k.ko text data bss dec hex filename 67304 3080 168 70552 11398 ath9k.ko Signed-off-by:
Larry Finger <Larry.Finger@lwfinger.net> Signed-off-by:
Ming Lei <tom.leiming@gmail.com> Signed-off-by:
-
John W. Linville authored
Merge branch 'wireless-next-2.6' of git://git.kernel.org/pub/scm/linux/kernel/git/iwlwifi/iwlwifi-2.6
-
- 19 Mar, 2010 10 commits
-
-
Wey-Yi Guy authored
Split the connectivity check and recovery routine into separated functions based on the types 1. iwl_good_ack_health() - check for ack count 2. iwl_good_plcp_health() - check for plcp error Based on the type of errors being detected, different recovery methods will be used to bring the system back to normal operational state. Because different NIC has different HW and uCode, the behavior is also different; these functions thus now form part of the ops infrastructure, so we can have more control on how to monitor and recover from error condition case per device. Signed-off-by:
Wey-Yi Guy <wey-yi.w.guy@intel.com> Signed-off-by:
Reinette Chatre <reinette.chatre@intel.com>
-
Wey-Yi Guy authored
Monitors the tx statistics to detect the drop in throughput. When the throughput drops, the ratio of the actual_ack_count and the expected_ack_count also drops. At the same time, the aggregated ba_timeout (the number of ba timeout retries) also rises. If the actual_ack_count/expected_ack_count ratio is 0 and the number of ba timeout retries rises to BA_TIMEOUT_MAX, no tx packets can be delivered. Reloading the uCode and bring the system back to normal operational state. Signed-off-by:
Trieu 'Andrew' Nguyen <trieux.t.nguyen@intel.com> Signed-off-by:
-
Wey-Yi Guy authored
Move the plcp error checking into stand alone function and pointed by ops to accommodate devices not needing this recovery. Signed-off-by:
-
Wey-Yi Guy authored
Monitors the internal TX queues periodically. When a queue is stuck for some unknown conditions causing the throughput to drop and the transfer is stop, the driver will force firmware reload and bring the system back to normal operational state. The iwlwifi devices behave differently in this regard so this feature is made part of the ops infrastructure so we can have more control on how to monitor and recover from tx queue stall case per device. Signed-off-by:
-
Shanyu Zhao authored
Align the driver names with official product names for 1000/5000/6000 series. This change mainly affects the debug messages show up in system log. No functional changes. Signed-off-by:
Shanyu Zhao <shanyu.zhao@intel.com> Signed-off-by:
-
Johannes Berg authored
The only caller of this is iwl_rs_rate_init which is only called with a valid sta_id. Signed-off-by:
Johannes Berg <johannes.berg@intel.com> Signed-off-by:
-
Reinette Chatre authored
mac80211 recently implemented two new callbacks that are used to request station add/remove from the driver. The benefot from these new callbacks are that they enable the driver to sleep while performing this work. This is a big patch since a few things need to be coordinated in this move. First we need to decouple station management from rate scaling, which caused a lot of code to be moved and/or deleted. Next we needed to tie in with mac80211's station management callback and let it direct our station management as well as trigger the rate scaling initialization. Signed-off-by: -
Reinette Chatre authored
Currently the broadcast station is added after every RXON command. Change this to only add the broadcast station when interface is added by mac80211. With this we need some extra work to ensure broadcast station is always present since station table is cleared when RXON without ASSOC bit set is sent. To deal with this we re-add all driver known stations to uCode after such an RXON command is sent. We also do some cleanup and remove the various calls to clear the station table. We now only clear the station table in two scenarios: - only clear uCode portion of station table when RXON command without ASSOC bit is sent - clear uCode and driver portion when interface goes down or is removed. We need to do this clearing when interface goes down to deal with the device restart/reconfigure routines which do not remove the interface, but do add the interface during reconfiguration. Previously the keys were also cleared when station table in driver is cleared, this is not done anymore since mac80211 will take care that keys are set and cleared correctly. There is a known issue with this change. Associating with different AP without bringing interface down fails with a firmware error. This is because of the lack of full station notification support and the later patches in this series that complete the station notification support will fix this. Signed-off-by: -
Ben Cahill authored
FH register dump debug feature was recently added for 5000 and later; 4965 can use it also. Signed-off-by:
Ben Cahill <ben.m.cahill@intel.com> Signed-off-by:
-
florian@mickler.org authored
This patch renames the (never officially released) sysfs-knobs "blocked_hw" and "blocked_sw" to "hard" and "soft", as the hardware vs software conotation is misleading. It also gets rid of not needed locks around u32-read-access. Signed-off-by:
Florian Mickler <florian@mickler.org> Signed-off-by:
-
- 16 Mar, 2010 4 commits
-
-
Bruno Randolf authored
when an IBSS merge happened, the supported rates for the newly added station were left empty, causing the rate control module to be initialized with only the basic rates. the section of the ibss code which deals with updating supported rates for an already existing station failed to inform the rate control module about the new rates. as both minstrel and pid don't have an update function i just use the init function. also remove unnecessary (unsigned long long) casts and edit debug message. Signed-off-by:
Bruno Randolf <br1@einfach.org> Signed-off-by:
-
John W. Linville authored
I accidentally merged an incomplete version of the patch... This reverts commit b4d59a93. Signed-off-by:
-
Alban Browaeys authored
After testing, we found that TX_STA_FIFO_MCS is the last MCS value tried. If the transmission failed, 8 frames have been transmitted. If the transmission succeed, we can easily compute the number of retry. This patch fix the way status is reported to mac80211 rate control. It has 2 bugs : 1. mcs can contain the short preamble flag and it will lead to wrong computations. 2. minstrel nearly always say that 54 Mbits is the best rate, even if we are very far from the AP Signed-off-by:
Benoit Papillault <benoit.papillault@free.fr> Signed-off-by:
Alban Browaeys <prahal@yahoo.com> Acked-by:
Ivo van Doorn <ivdoorn@gmail.com> Signed-off-by:
Gertjan van Wingerde <gwingerde@gmail.com> Signed-off-by:
-
Alban Browaeys authored
This is an implementation that support WCID being the encryption key. Wireless Cli Id was set to be the encryption key in rt2800pci_write_tx_desc and read (TX_STA_FIFO_WCID) as the current queue entry index. Signed-off-by:
-
- 15 Mar, 2010 9 commits
-
-
Helmut Schaa authored
Fix compile warning "rt2800pci.c:1248: warning: 'rt2800pci_device_table' defined but not used" when building rt2800pci with only soc support (without pci). Signed-off-by:
Helmut Schaa <helmut.schaa@googlemail.com> Acked-by:
-
Pavel Roskin authored
"ieee80211" was the old name of the common library for ipw2100 and ipw2200. It was renamed to "libipw", but some occurrences of the old name remained. Rename alloc_ieee80211() to alloc_libipw() and free_ieee80211() to free_libipw(). Adjust comments and label names. Change prefixes in diagnostic messages. Keep /proc/net/ieee80211 under the original name to avoid breaking user interface. Move the affected EXPORT_SYMBOL macros to their proper places. Signed-off-by:
Pavel Roskin <proski@gnu.org> Signed-off-by:
-
Pavel Roskin authored
The hostap driver provides better support for Prism chipset. Signed-off-by:
-
Grazvydas Ignotas authored
4sec wait is way too pessimistic, TI driver uses 40ms here, and testing shows that is ebough, so let's also use that. While at it, add useful sounding comment from the TI driver. Signed-off-by:
Grazvydas Ignotas <notasas@gmail.com> Acked-by:
Kalle Valo <kalle.valo@iki.fi> Signed-off-by:
-
Grazvydas Ignotas authored
For some unknown reason ELP_CTRL can't be accesed using sdio_memcpy_* functions (any attemts to do so result in timeouts): wl1251: ERROR sdio write failed (-110) wl1251: ERROR sdio read failed (-110) wl1251: WARNING WLAN not ready To fix this, add special IO functions for ELP_CTRL access that are using sdio_readb/sdio_writeb. Similar handling is done in TI reference driver from Android code drop. Signed-off-by:
-
Grazvydas Ignotas authored
Make local functions and data static, also constify some structures. While at it, clean up unneeded includes. Signed-off-by:
-
Felix Fietkau authored
When a cooked monitor interface is active, ieee80211_tx_status() generates a radiotap header for every single frame, even if it wasn't injected and thus won't be sent to a monitor interface. This patch reduces cpu utilization by moving the cooked monitor check a bit earlier, before it generates the rtap header. Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Stewart Malik authored
Signed-off-by: -
John W. Linville authored
Signed-off-by:
-
