- 16 Apr, 2010 40 commits
-
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
Vasanthakumar Thiagarajan <vasanth@atheros.com> Signed-off-by:
John W. Linville <linville@tuxdriver.com>
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
-
Luis R. Rodriguez authored
The AR9002 hardware code enables aggregation for WEP but mac80211 doesn't enable aggregation with WEP, and the AR9003 code family does not need this so skip it for now for AR9003 but leave the code and annotate we should eventually consider how to remove this in consideration for the HAL unification goals. Signed-off-by:
Luis R. Rodriguez <lrodriguez@atheros.com> Signed-off-by:
-
Luis R. Rodriguez authored
The asynch fifo code is specific to >= AR9287 so stuff it into the AR9002 hardware family code and skip it for AR9003 cards. Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
Felix Fietkau <nbd@openwrt.org> Signed-off-by:
-
Luis R. Rodriguez authored
This is done depending on what the EEPROM settings indicates. Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
-
Luis R. Rodriguez authored
We add this now as OFDM spur mitigation required accessing the EEPROM for the AR9003 devices. Signed-off-by:
-
Senthil Balasubramanian authored
Signed-off-by:
Senthil Balasubramanian <senthilkumar@atheros.com> Signed-off-by:
-
Luis R. Rodriguez authored
Once upon a time the AR_EEPROM_MAC macro was added to let us add a random attribute to the three 4-bytes of MAC addresses entries we read from the EEPROM. This was good while a random high-enough value was used which did not conflict with any of the already existing enum eeprom_param values. With AR9003 support the enums overlap and it means we either increment the random offset or just restore the reading logic to match what the HAL has. I choose to do the later to synchronize the logic on both code bases. This should fix reading the MAC address from the EEPROM on AR9003 hardware. Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
This goes with some new shiny TX IQ calibration that AR9003 hardware family supports. Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
This is the last call on calib.c which acceses PHY stuff, with this change we calib.c is now generic between both all supported hardware families. Signed-off-by:
-
Luis R. Rodriguez authored
This is so we can share routines which access this register on calib.c Signed-off-by:
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
-
Vasanthakumar Thiagarajan authored
Also move interrupt related code to mac.c Signed-off-by:
-
Luis R. Rodriguez authored
The AR9003 TX/RX gain is currently initialized with the other components, so for now AR9003 does not implment this callback, after hardware bring up we can test moving the TX/RX gain there as well and if it works well move them to its own callback as well. Since all INI stuff is now moved out hw.c no longer needs to include and touch any original INI headers/structs. Signed-off-by:
-
Luis R. Rodriguez authored
This is specific to the AR9002 family only. Signed-off-by:
-
Luis R. Rodriguez authored
Move out the generic hardware family code out into their own files, we have one for AR5008, AR9001, and AR9002 family (ar9002_hw.c) and another file for the new AR9003 hardware family (ar9003_hw.c). Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
Calibration code touches phy registers and since these change the calibration code needs to be abstracted. Noise floor calibration is the only thing remaining but since the remaining calls only touch the AR_PHY_AGC_CONTROL register we'll just define that register conditionally, that will be done separately. The goal is to remove the dependency of ar9002_phy.h on calib.c This also adds stubs to be filled for AR9003 calibration code. Signed-off-by:
-
Luis R. Rodriguez authored
The calibration settings should go into the respective hardware family AR9002 calibration settings callback, ar9002_hw_init_cal_settings(). Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
We can do the family revision check on the top level caller. Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
We're bailing out on the alternative code path so remove the else branch. Signed-off-by:
-
Luis R. Rodriguez authored
Signed-off-by:
-
Luis R. Rodriguez authored
Open Loop Control temperature compensation changes between our hardware so use a helper for it. Signed-off-by:
-
Luis R. Rodriguez authored
The code can be simplified and shared between two locations if we bring this into a helper. During reset we do not account for the skip count. Signed-off-by:
-
Senthil Balasubramanian authored
We can reorganize the code in such a way that eep_map can be removed, which makes the code more clearer. Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
-
Felix Fietkau authored
Signed-off-by:
-
Vasanthakumar Thiagarajan authored
Signed-off-by:
-
