d/freebsd-nq
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
51d025a596
freebsd-nq/sys/dev/ath/ath_hal/ar5416/ar5416_reset.c

2892 lines
94 KiB
C
Raw Normal View History

virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
* Copyright (c) 2002-2009 Sam Leffler, Errno Consulting
virgin import of ath hal
2008-11-28 00:03:41 +00:00
* Copyright (c) 2002-2008 Atheros Communications, Inc.
*
* Permission to use, copy, modify, and/or distribute this software for any
* purpose with or without fee is hereby granted, provided that the above
* copyright notice and this permission notice appear in all copies.
*
* THE SOFTWARE IS PROVIDED "AS IS" AND THE AUTHOR DISCLAIMS ALL WARRANTIES
* WITH REGARD TO THIS SOFTWARE INCLUDING ALL IMPLIED WARRANTIES OF
* MERCHANTABILITY AND FITNESS. IN NO EVENT SHALL THE AUTHOR BE LIABLE FOR
* ANY SPECIAL, DIRECT, INDIRECT, OR CONSEQUENTIAL DAMAGES OR ANY DAMAGES
* WHATSOEVER RESULTING FROM LOSS OF USE, DATA OR PROFITS, WHETHER IN AN
* ACTION OF CONTRACT, NEGLIGENCE OR OTHER TORTIOUS ACTION, ARISING OUT OF
* OR IN CONNECTION WITH THE USE OR PERFORMANCE OF THIS SOFTWARE.
*
fix return status handling by ar5XXXReset; this is the reason the driver sometimes reports reset failed w/ status 0
2009-01-23 05:33:48 +00:00
* $FreeBSD$
virgin import of ath hal
2008-11-28 00:03:41 +00:00
*/
#include "opt_ah.h"
#include "ah.h"
#include "ah_internal.h"
#include "ah_devid.h"
#include "ah_eeprom_v14.h"
#include "ar5416/ar5416.h"
#include "ar5416/ar5416reg.h"
#include "ar5416/ar5416phy.h"
/* Eeprom versioning macros. Returns true if the version is equal or newer than the ver specified */
#define EEP_MINOR(_ah) \
(AH_PRIVATE(_ah)->ah_eeversion & AR5416_EEP_VER_MINOR_MASK)
#define IS_EEP_MINOR_V2(_ah) (EEP_MINOR(_ah) >= AR5416_EEP_MINOR_VER_2)
#define IS_EEP_MINOR_V3(_ah) (EEP_MINOR(_ah) >= AR5416_EEP_MINOR_VER_3)
/* Additional Time delay to wait after activiting the Base band */
#define BASE_ACTIVATE_DELAY 100 /* 100 usec */
#define PLL_SETTLE_DELAY 300 /* 300 usec */
#define RTC_PLL_SETTLE_DELAY 1000 /* 1 ms */
static void ar5416InitDMA(struct ath_hal *ah);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
static void ar5416InitBB(struct ath_hal *ah, const struct ieee80211_channel *);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
static void ar5416InitIMR(struct ath_hal *ah, HAL_OPMODE opmode);
static void ar5416InitQoS(struct ath_hal *ah);
static void ar5416InitUserSettings(struct ath_hal *ah);
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
static void ar5416OverrideIni(struct ath_hal *ah, const struct ieee80211_channel *);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#if 0
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
static HAL_BOOL ar5416ChannelChange(struct ath_hal *, const struct ieee80211_channel *);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
static void ar5416SetDeltaSlope(struct ath_hal *, const struct ieee80211_channel *);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
static HAL_BOOL ar5416SetResetPowerOn(struct ath_hal *ah);
static HAL_BOOL ar5416SetReset(struct ath_hal *ah, int type);
static HAL_BOOL ar5416SetPowerPerRateTable(struct ath_hal *ah,
struct ar5416eeprom *pEepData,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
const struct ieee80211_channel *chan, int16_t *ratesArray,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint16_t cfgCtl, uint16_t AntennaReduction,
uint16_t twiceMaxRegulatoryPower,
uint16_t powerLimit);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
static void ar5416Set11nRegs(struct ath_hal *ah, const struct ieee80211_channel *chan);
Mark the PHY as inactive before the chip is reset. It's also marked inactive by the initvals, and enabled after the baseband/PLL has been configured, but before the RF registers have been programmed. The origin and reason for this particular change is currently unknown. Obtained from: Linux ath9k
2011-04-17 13:46:13 +00:00
static void ar5416MarkPhyInactive(struct ath_hal *ah);
Introduce an AR5416 flavour of the IFS and mac usec/timing configuration used when running the chips in half/quarter rate. This sets up some default parameters which are then overridden by the driver (which manually configures things like slot timing at interface start time.) Although this is a copy-and-modify from the AR5212 HAL, I did peek at the reference HAL and the ath9k driver to see what they did. Ath9k in particular doesn't hard-code this - instead, their version of ar5416InitUserSettings() does all of the relevant math. TODO: * do the math, not hard code things! * fix the mac clock calculation for the AR9287; since it runs the MAC clock at a higher rate, requiring all the duration calculations to change; * Do a whole lot more validation for half/quarter rates. Obtained from: Qualcomm Atheros, Linux ath9k
2012-09-13 07:22:40 +00:00
static void ar5416SetIFSTiming(struct ath_hal *ah,
const struct ieee80211_channel *chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Places the device in and out of reset and then places sane
* values in the registers based on EEPROM config, initialization
* vectors (as determined by the mode), and station configuration
*
* bChannelChange is used to preserve DMA/PCU registers across
* a HW Reset during channel change.
*/
HAL_BOOL
ar5416Reset(struct ath_hal *ah, HAL_OPMODE opmode,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
struct ieee80211_channel *chan,
ath(4): begin fleshing out a "reset type" extension to force cold/warn resets. Right now the only way to force a cold reset is: * The HAL itself detects it's needed, or * The sysctl, setting all resets to be cold. Trouble is, cold resets take quite a bit longer than warm resets. However, there are situations where a cold reset would be nice. Specifically, after a stuck beacon, BB/MAC hang, stuck calibration results, etc. The vendor HAL has a separate method to set the reset reason (which is how HAL_RESET_BBPANIC gets set) which informs the HAL during the reset path why it occured. This is almost but not quite the same; I may eventually unify both approaches in the future. This commit just extends HAL_RESET_TYPE to include both status (eg BBPANIC) and type (eg do COLD.) None of the HAL code uses it yet though; that'll come later. It also is a big no-op in each HAL - I need to go teach each of the HALs about cold/warm reset through this path.
2015-11-09 15:59:42 +00:00
HAL_BOOL bChannelChange,
HAL_RESET_TYPE resetType,
HAL_STATUS *status)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
#define N(a) (sizeof (a) / sizeof (a[0]))
#define FAIL(_code) do { ecode = _code; goto bad; } while (0)
struct ath_hal_5212 *ahp = AH5212(ah);
HAL_CHANNEL_INTERNAL *ichan;
uint32_t saveDefAntenna, saveLedState;
uint32_t macStaId1;
uint16_t rfXpdGain[2];
HAL_STATUS ecode;
uint32_t powerVal, rssiThrReg;
uint32_t ackTpcPow, ctsTpcPow, chirpTpcPow;
preliminary ar9280 support: o add 9280 attach that sets up ini, cal, etc. o new rf backend for 9280 and later parts o split ini setup and spur mitigation support out to methods and provide 9280-specific support o minor fixups to shared code to handle 9280-specific work Obtained from: Atheros (ini values and some code)
2009-03-12 18:18:28 +00:00
int i;
For chips that are full reset in ar5416ChipReset(), save and restore the TSF. Merlin (ar9280) and later were full-reset if they're doing open-loop TX power control but the TSF wasn't being saved/restored. Add ar5212SetTsf64() which sets the 64 bit TSF appropriately.
2011-03-09 04:39:35 +00:00
uint64_t tsf = 0;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_MARK(ah, AH_MARK_RESET, bChannelChange);
/* Bring out of sleep mode */
if (!ar5416SetPowerMode(ah, HAL_PM_AWAKE, AH_TRUE)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: chip did not wakeup\n",
__func__);
Revert a local change that should not have been in the last commit. Approved by: re (kib)
2009-06-28 11:32:52 +00:00
FAIL(HAL_EIO);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/*
* Map public channel to private.
*/
ichan = ath_hal_checkchannel(ah, chan);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (ichan == AH_NULL)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
FAIL(HAL_EINVAL);
switch (opmode) {
case HAL_M_STA:
case HAL_M_IBSS:
case HAL_M_HOSTAP:
case HAL_M_MONITOR:
break;
default:
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: invalid operating mode %u\n",
__func__, opmode);
FAIL(HAL_EINVAL);
break;
}
HALASSERT(AH_PRIVATE(ah)->ah_eeversion >= AR_EEPROM_VER14_1);
After thinking about this a bit more, let's not keep statistics per-channel in the HAL. That's very memory hungry (32k just for channel statistics) which would be better served by keeping a summary in the ANI state. Or, later, keep a survey history in net80211. So: * Migrate the ah_chansurvey array to be a single entry, for the current channel. * Change the ioctl interface and ANI code to just reference that. * Clear the ah_chansurvey array during channel reset, both in the AR5212 and AR5416 reset path.
2012-04-28 22:03:19 +00:00
/* Blank the channel survey statistics */
Move the HAL channel survey support out to be in the top-level HAL, rathe than private in each HAL module. Whilst here, modify ath_hal_private to always have the per-channel noisefloor stats, rather than conditionally. This just makes life easier in general (no strange ABI differences between different HAL compile options.) Add a couple of methods (clear/reset, add) rather than using hand-rolled versions of things.
2015-03-29 21:50:21 +00:00
ath_hal_survey_clear(ah);
After thinking about this a bit more, let's not keep statistics per-channel in the HAL. That's very memory hungry (32k just for channel statistics) which would be better served by keeping a summary in the ANI state. Or, later, keep a survey history in net80211. So: * Migrate the ah_chansurvey array to be a single entry, for the current channel. * Change the ioctl interface and ANI code to just reference that. * Clear the ah_chansurvey array during channel reset, both in the AR5212 and AR5416 reset path.
2012-04-28 22:03:19 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* XXX Turn on fast channel change for 5416 */
Move the HAL channel survey support out to be in the top-level HAL, rathe than private in each HAL module. Whilst here, modify ath_hal_private to always have the per-channel noisefloor stats, rather than conditionally. This just makes life easier in general (no strange ABI differences between different HAL compile options.) Add a couple of methods (clear/reset, add) rather than using hand-rolled versions of things.
2015-03-29 21:50:21 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Preserve the bmiss rssi threshold and count threshold
* across resets
*/
rssiThrReg = OS_REG_READ(ah, AR_RSSI_THR);
/* If reg is zero, first time thru set to default val */
if (rssiThrReg == 0)
rssiThrReg = INIT_RSSI_THR;
/*
* Preserve the antenna on a channel change
*/
saveDefAntenna = OS_REG_READ(ah, AR_DEF_ANTENNA);
Fix receive on the AR9285 (Kite) with only one antenna connected. The main problem here is that fast and driver RX diversity isn't actually configured; I need to figure out why that is. That said, this makes the single-antenna connected AR9285 and AR2427 (AR9285 w/ no 11n) work correctly. PR: kern/179269
2013-06-03 19:14:29 +00:00
/*
* Don't do this for the AR9285 - it breaks RX for single
* antenna designs when diversity is disabled.
*
* I'm not sure what this was working around; it may be
* something to do with the AR5416. Certainly this register
* isn't supposed to be used by the MIMO chips for anything
* except for defining the default antenna when an external
* phase array / smart antenna is connected.
*
* See PR: kern/179269 .
*/
if ((! AR_SREV_KITE(ah)) && saveDefAntenna == 0) /* XXX magic constants */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
saveDefAntenna = 1;
/* Save hardware flag before chip reset clears the register */
macStaId1 = OS_REG_READ(ah, AR_STA_ID1) &
(AR_STA_ID1_BASE_RATE_11B | AR_STA_ID1_USE_DEFANT);
/* Save led state from pci config register */
saveLedState = OS_REG_READ(ah, AR_MAC_LED) &
(AR_MAC_LED_ASSOC | AR_MAC_LED_MODE |
AR_MAC_LED_BLINK_THRESH_SEL | AR_MAC_LED_BLINK_SLOW);
For chips that are full reset in ar5416ChipReset(), save and restore the TSF. Merlin (ar9280) and later were full-reset if they're doing open-loop TX power control but the TSF wasn't being saved/restored. Add ar5212SetTsf64() which sets the 64 bit TSF appropriately.
2011-03-09 04:39:35 +00:00
/* For chips on which the RTC reset is done, save TSF before it gets cleared */
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
if (AR_SREV_HOWL(ah) ||
Add a new HAL parameter which forces a full reset rather than warm reset. This forces a full reset of the baseband/radio and seems needed to clear some issues (with Merlin at least) when the baseband gets confused in a very noisy environment. Sponsored by: Hobnob, Inc.
2011-11-09 05:30:24 +00:00
(AR_SREV_MERLIN(ah) &&
ath_hal_eepromGetFlag(ah, AR_EEP_OL_PWRCTRL)) ||
(ah->ah_config.ah_force_full_reset))
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
tsf = ar5416GetTsf64(ah);
For chips that are full reset in ar5416ChipReset(), save and restore the TSF. Merlin (ar9280) and later were full-reset if they're doing open-loop TX power control but the TSF wasn't being saved/restored. Add ar5212SetTsf64() which sets the 64 bit TSF appropriately.
2011-03-09 04:39:35 +00:00
Mark the PHY as inactive before the chip is reset. It's also marked inactive by the initvals, and enabled after the baseband/PLL has been configured, but before the RF registers have been programmed. The origin and reason for this particular change is currently unknown. Obtained from: Linux ath9k
2011-04-17 13:46:13 +00:00
/* Mark PHY as inactive; marked active in ar5416InitBB() */
ar5416MarkPhyInactive(ah);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (!ar5416ChipReset(ah, chan)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: chip reset failed\n", __func__);
FAIL(HAL_EIO);
}
For chips that are full reset in ar5416ChipReset(), save and restore the TSF. Merlin (ar9280) and later were full-reset if they're doing open-loop TX power control but the TSF wasn't being saved/restored. Add ar5212SetTsf64() which sets the 64 bit TSF appropriately.
2011-03-09 04:39:35 +00:00
/* Restore TSF */
if (tsf)
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
ar5416SetTsf64(ah, tsf);
For chips that are full reset in ar5416ChipReset(), save and restore the TSF. Merlin (ar9280) and later were full-reset if they're doing open-loop TX power control but the TSF wasn't being saved/restored. Add ar5212SetTsf64() which sets the 64 bit TSF appropriately.
2011-03-09 04:39:35 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_MARK(ah, AH_MARK_RESET_LINE, __LINE__);
Bring in a couple of fixes from the Linux ath9k related to chip hangs. While there, try to make the register write pattern look like what's done by ath9k. MFC after: 3 days
2010-06-01 15:33:10 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah))
OS_REG_SET_BIT(ah, AR_GPIO_INPUT_EN_VAL, AR_GPIO_JTAG_DISABLE);
preliminary ar9280 support: o add 9280 attach that sets up ini, cal, etc. o new rf backend for 9280 and later parts o split ini setup and spur mitigation support out to methods and provide 9280-specific support o minor fixups to shared code to handle 9280-specific work Obtained from: Atheros (ini values and some code)
2009-03-12 18:18:28 +00:00
AH5416(ah)->ah_writeIni(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Bring over my AR9287 work in progress. It isn't linked into the build because it's missing the TX power and PDADC programming code. This code is mostly based on the ath9k codebase, compared against the Atheros codebase as appropriate. What's implemented: * probe/attach * EEPROM board value programming * RX initial calibration * radio channel programming * general MAC / baseband setup * async fifo setup * open-loop tx power calibration What's missing before it can be enabled by default: * TX power / calibration setting code * closed-loop tx power calibration routines * TSF2 handling * generic timer support from ath9k Obtained from: Atheros, ath9k
2011-05-26 09:15:33 +00:00
if(AR_SREV_KIWI_13_OR_LATER(ah) ) {
/* Enable ASYNC FIFO */
OS_REG_SET_BIT(ah, AR_MAC_PCU_ASYNC_FIFO_REG3,
AR_MAC_PCU_ASYNC_FIFO_REG3_DATAPATH_SEL);
OS_REG_SET_BIT(ah, AR_PHY_MODE, AR_PHY_MODE_ASYNCFIFO);
OS_REG_CLR_BIT(ah, AR_MAC_PCU_ASYNC_FIFO_REG3,
AR_MAC_PCU_ASYNC_FIFO_REG3_SOFT_RESET);
OS_REG_SET_BIT(ah, AR_MAC_PCU_ASYNC_FIFO_REG3,
AR_MAC_PCU_ASYNC_FIFO_REG3_SOFT_RESET);
}
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
/* Override ini values (that can be overriden in this fashion) */
ar5416OverrideIni(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Setup 11n MAC/Phy mode registers */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416Set11nRegs(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_MARK(ah, AH_MARK_RESET_LINE, __LINE__);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
/*
* Some AR91xx SoC devices frequently fail to accept TSF writes
* right after the chip reset. When that happens, write a new
* value after the initvals have been applied, with an offset
* based on measured time difference
*/
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
if (AR_SREV_HOWL(ah) && (ar5416GetTsf64(ah) < tsf)) {
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
tsf += 1500;
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
ar5416SetTsf64(ah, tsf);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_RESET, ">>>2 %s: AR_PHY_DAG_CTRLCCK=0x%x\n",
__func__, OS_REG_READ(ah,AR_PHY_DAG_CTRLCCK));
HALDEBUG(ah, HAL_DEBUG_RESET, ">>>2 %s: AR_PHY_ADC_CTL=0x%x\n",
__func__, OS_REG_READ(ah,AR_PHY_ADC_CTL));
Break out the chainmask init code into a new function - ar5416InitChainMasks() . ath9k does a few different things here during config - if it's an early AR5416 with two chains, it enables all three chains for calibration and then restores the chainmask to the original values after initial calibration has completed. The reason behind this commit is to begin breaking out the chainmask configuration for this specific reason; follow-up commits will add the chainmask restore in the ar5416Reset() routine.
2011-01-26 10:08:37 +00:00
/*
* This routine swaps the analog chains - it should be done
* before any radio register twiddling is done.
*/
ar5416InitChainMasks(ah);
Remove some holdovers from the AR5212 origin of this code. These aren't relevant here.
2011-04-29 12:52:18 +00:00
Only do open loop power control and temperature compensation for the AR9280 based NICs if it's actually enabled. Some of the OLC code was erroneously called during setup and calibration. This may have caused some incorrect behaviour.
2011-05-13 14:33:45 +00:00
/* Setup the open-loop power calibration if required */
if (ath_hal_eepromGetFlag(ah, AR_EEP_OL_PWRCTRL)) {
AH5416(ah)->ah_olcInit(ah);
AH5416(ah)->ah_olcTempCompensation(ah);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Setup the transmit power values. */
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
if (!ah->ah_setTxPower(ah, chan, rfXpdGain)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: error init'ing transmit power\n", __func__);
FAIL(HAL_EIO);
}
/* Write the analog registers */
preliminary ar9280 support: o add 9280 attach that sets up ini, cal, etc. o new rf backend for 9280 and later parts o split ini setup and spur mitigation support out to methods and provide 9280-specific support o minor fixups to shared code to handle 9280-specific work Obtained from: Atheros (ini values and some code)
2009-03-12 18:18:28 +00:00
if (!ahp->ah_rfHal->setRfRegs(ah, chan,
IEEE80211_IS_CHAN_2GHZ(chan) ? 2: 1, rfXpdGain)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: ar5212SetRfRegs failed\n", __func__);
FAIL(HAL_EIO);
}
/* Write delta slope for OFDM enabled modes (A, G, Turbo) */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_OFDM(chan)|| IEEE80211_IS_CHAN_HT(chan))
ar5416SetDeltaSlope(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
preliminary ar9280 support: o add 9280 attach that sets up ini, cal, etc. o new rf backend for 9280 and later parts o split ini setup and spur mitigation support out to methods and provide 9280-specific support o minor fixups to shared code to handle 9280-specific work Obtained from: Atheros (ini values and some code)
2009-03-12 18:18:28 +00:00
AH5416(ah)->ah_spurMitigate(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Setup board specific options for EEPROM version 3 */
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
if (!ah->ah_setBoardValues(ah, chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: error setting board options\n", __func__);
FAIL(HAL_EIO);
}
OS_MARK(ah, AH_MARK_RESET_LINE, __LINE__);
OS_REG_WRITE(ah, AR_STA_ID0, LE_READ_4(ahp->ah_macaddr));
OS_REG_WRITE(ah, AR_STA_ID1, LE_READ_2(ahp->ah_macaddr + 4)
| macStaId1
| AR_STA_ID1_RTS_USE_DEF
| ahp->ah_staId1Defaults
);
ar5212SetOperatingMode(ah, opmode);
/* Set Venice BSSID mask according to current state */
OS_REG_WRITE(ah, AR_BSSMSKL, LE_READ_4(ahp->ah_bssidmask));
OS_REG_WRITE(ah, AR_BSSMSKU, LE_READ_2(ahp->ah_bssidmask + 4));
/* Restore previous led state */
Import some HOWL (AR9130) related fixes from Atheros. Obtained from: Atheros
2011-05-05 02:59:31 +00:00
if (AR_SREV_HOWL(ah))
OS_REG_WRITE(ah, AR_MAC_LED,
AR_MAC_LED_ASSOC_ACTIVE | AR_CFG_SCLK_32KHZ);
else
OS_REG_WRITE(ah, AR_MAC_LED, OS_REG_READ(ah, AR_MAC_LED) |
saveLedState);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Bring over my AR9287 work in progress. It isn't linked into the build because it's missing the TX power and PDADC programming code. This code is mostly based on the ath9k codebase, compared against the Atheros codebase as appropriate. What's implemented: * probe/attach * EEPROM board value programming * RX initial calibration * radio channel programming * general MAC / baseband setup * async fifo setup * open-loop tx power calibration What's missing before it can be enabled by default: * TX power / calibration setting code * closed-loop tx power calibration routines * TSF2 handling * generic timer support from ath9k Obtained from: Atheros, ath9k
2011-05-26 09:15:33 +00:00
/* Start TSF2 for generic timer 8-15 */
#ifdef NOTYET
if (AR_SREV_KIWI(ah))
ar5416StartTsf2(ah);
#endif
Add bluetooth fixes to the AR5416/AR92xx HAL: * Call the bluetooth setup function during the reset path, so the bluetooth settings are actually initialised. * Call the AR9285 diversity functions during bluetooth setup; so the AR9285 diversity and antenna configuration registers are correctly programmed * Misc debugging info. Tested: * AR9285+AR3011 bluetooth combo; this code itself doesn't enable bluetooth coexistence but it's part of what I'm currently using.
2013-06-07 05:17:58 +00:00
/*
* Enable Bluetooth Coexistence if it's enabled.
*/
if (AH5416(ah)->ah_btCoexConfigType != HAL_BT_COEX_CFG_NONE)
ar5416InitBTCoex(ah);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Restore previous antenna */
OS_REG_WRITE(ah, AR_DEF_ANTENNA, saveDefAntenna);
Save and restore the association ID across interface resets. Obtained from: Atheros MFC after: 1 week
2011-10-25 23:13:36 +00:00
/* then our BSSID and associate id */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_BSS_ID0, LE_READ_4(ahp->ah_bssid));
Save and restore the association ID across interface resets. Obtained from: Atheros MFC after: 1 week
2011-10-25 23:13:36 +00:00
OS_REG_WRITE(ah, AR_BSS_ID1, LE_READ_2(ahp->ah_bssid + 4) |
(ahp->ah_assocId & 0x3fff) << AR_BSS_ID1_AID_S);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Restore bmiss rssi & count thresholds */
OS_REG_WRITE(ah, AR_RSSI_THR, ahp->ah_rssiThr);
OS_REG_WRITE(ah, AR_ISR, ~0); /* cleared on write */
Restore the RSSI threshold after writing the board values. This would be overwritten by the board initvals written in ah->writeIni().
2011-05-11 09:47:48 +00:00
/* Restore bmiss rssi & count thresholds */
OS_REG_WRITE(ah, AR_RSSI_THR, rssiThrReg);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (!ar5212SetChannel(ah, chan))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
FAIL(HAL_EIO);
OS_MARK(ah, AH_MARK_RESET_LINE, __LINE__);
/* Set 1:1 QCU to DCU mapping for all queues */
for (i = 0; i < AR_NUM_DCU; i++)
OS_REG_WRITE(ah, AR_DQCUMASK(i), 1 << i);
ahp->ah_intrTxqs = 0;
for (i = 0; i < AH_PRIVATE(ah)->ah_caps.halTotalQueues; i++)
Use the HAL method rather than directly calling ar5212ResetTxQueue(). Since ath9k does some slightly different bit fiddling when setting up the TX queues, it may that the TX queue setup/reset functions will need overriding later on.
2011-03-19 03:09:21 +00:00
ah->ah_resetTxQueue(ah, i);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416InitIMR(ah, opmode);
Introduce an AR5416 flavour of the IFS and mac usec/timing configuration used when running the chips in half/quarter rate. This sets up some default parameters which are then overridden by the driver (which manually configures things like slot timing at interface start time.) Although this is a copy-and-modify from the AR5212 HAL, I did peek at the reference HAL and the ath9k driver to see what they did. Ath9k in particular doesn't hard-code this - instead, their version of ar5416InitUserSettings() does all of the relevant math. TODO: * do the math, not hard code things! * fix the mac clock calculation for the AR9287; since it runs the MAC clock at a higher rate, requiring all the duration calculations to change; * Do a whole lot more validation for half/quarter rates. Obtained from: Qualcomm Atheros, Linux ath9k
2012-09-13 07:22:40 +00:00
ar5416SetCoverageClass(ah, AH_PRIVATE(ah)->ah_coverageClass, 1);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416InitQoS(ah);
* Add in a comment about ar5416InitUserSettings() potentially modifying AR_DIAG_SW. There's a hardware workaround which sets disabling some errors early at startup and clears said bits before the PCU begins receiving - it does this to avoid RX descriptor status errors. It's possible these bits aren't being completely properly twiddled in all instances; but in particular if the diag_reg HAL variable is set it won't be setting these bits correctly. I'll review this at some point. * Disable multicast search on mac address and key id - the driver doesn't use it at the moment and thus adhoc may be broken for merlin and later. * Change this to be for Merlin 1.0 (which from what I understand wasn't ever publicly released) to be more correct.
2011-05-08 05:25:42 +00:00
/* This may override the AR_DIAG_SW register */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416InitUserSettings(ah);
Introduce an AR5416 flavour of the IFS and mac usec/timing configuration used when running the chips in half/quarter rate. This sets up some default parameters which are then overridden by the driver (which manually configures things like slot timing at interface start time.) Although this is a copy-and-modify from the AR5212 HAL, I did peek at the reference HAL and the ath9k driver to see what they did. Ath9k in particular doesn't hard-code this - instead, their version of ar5416InitUserSettings() does all of the relevant math. TODO: * do the math, not hard code things! * fix the mac clock calculation for the AR9287; since it runs the MAC clock at a higher rate, requiring all the duration calculations to change; * Do a whole lot more validation for half/quarter rates. Obtained from: Qualcomm Atheros, Linux ath9k
2012-09-13 07:22:40 +00:00
/* XXX this won't work for AR9287! */
if (IEEE80211_IS_CHAN_HALF(chan) || IEEE80211_IS_CHAN_QUARTER(chan)) {
ar5416SetIFSTiming(ah, chan);
#if 0
/*
* AR5413?
* Force window_length for 1/2 and 1/4 rate channels,
* the ini file sets this to zero otherwise.
*/
OS_REG_RMW_FIELD(ah, AR_PHY_FRAME_CTL,
AR_PHY_FRAME_CTL_WINLEN, 3);
}
#endif
}
Bring over my AR9287 work in progress. It isn't linked into the build because it's missing the TX power and PDADC programming code. This code is mostly based on the ath9k codebase, compared against the Atheros codebase as appropriate. What's implemented: * probe/attach * EEPROM board value programming * RX initial calibration * radio channel programming * general MAC / baseband setup * async fifo setup * open-loop tx power calibration What's missing before it can be enabled by default: * TX power / calibration setting code * closed-loop tx power calibration routines * TSF2 handling * generic timer support from ath9k Obtained from: Atheros, ath9k
2011-05-26 09:15:33 +00:00
if (AR_SREV_KIWI_13_OR_LATER(ah)) {
/*
* Enable ASYNC FIFO
*
* If Async FIFO is enabled, the following counters change
* as MAC now runs at 117 Mhz instead of 88/44MHz when
* async FIFO is disabled.
*
* Overwrite the delay/timeouts initialized in ProcessIni()
* above.
*/
OS_REG_WRITE(ah, AR_D_GBL_IFS_SIFS,
AR_D_GBL_IFS_SIFS_ASYNC_FIFO_DUR);
OS_REG_WRITE(ah, AR_D_GBL_IFS_SLOT,
AR_D_GBL_IFS_SLOT_ASYNC_FIFO_DUR);
OS_REG_WRITE(ah, AR_D_GBL_IFS_EIFS,
AR_D_GBL_IFS_EIFS_ASYNC_FIFO_DUR);
OS_REG_WRITE(ah, AR_TIME_OUT,
AR_TIME_OUT_ACK_CTS_ASYNC_FIFO_DUR);
OS_REG_WRITE(ah, AR_USEC, AR_USEC_ASYNC_FIFO_DUR);
OS_REG_SET_BIT(ah, AR_MAC_PCU_LOGIC_ANALYZER,
AR_MAC_PCU_LOGIC_ANALYZER_DISBUG20768);
OS_REG_RMW_FIELD(ah, AR_AHB_MODE, AR_AHB_CUSTOM_BURST_EN,
AR_AHB_CUSTOM_BURST_ASYNC_FIFO_VAL);
}
if (AR_SREV_KIWI_13_OR_LATER(ah)) {
/* Enable AGGWEP to accelerate encryption engine */
OS_REG_SET_BIT(ah, AR_PCU_MISC_MODE2,
AR_PCU_MISC_MODE2_ENABLE_AGGWEP);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* disable seq number generation in hw
*/
OS_REG_WRITE(ah, AR_STA_ID1,
OS_REG_READ(ah, AR_STA_ID1) | AR_STA_ID1_PRESERVE_SEQNUM);
ar5416InitDMA(ah);
/*
* program OBS bus to see MAC interrupts
*/
OS_REG_WRITE(ah, AR_OBS, 8);
Document exactly what the RX interrupt mitigation timers do.
2011-10-02 13:51:26 +00:00
/*
* Disable the "general" TX/RX mitigation timers.
*/
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_MIRT, 0);
Bring over interrupt mitigation changes from ath9k. * The existing interrupt mitigation code didn't mitigate anything - the per-packet TX/RX interrupts are still occuring. It's possible this worked for the AR5416 but not any later chipsets; I'll investigate and update as needed. * Set both the RX and TX threshold registers whilst I'm at it. This is verified to work on the AR9220 and AR9160. I'm leaving it off by default in case it's truely broken, but I need to have it enabled when doing 11n testing or interrupt loads exceed 10,000 interrupts/sec.
2011-03-25 00:03:21 +00:00
Correct/complete a partially-disabled TX interrupt mitigation configuration. Although a previous commit disabled TX interrupt mitigation handling and configuration, the mask register bits weren't setup correctly.
2011-10-25 23:17:53 +00:00
#ifdef AH_AR5416_INTERRUPT_MITIGATION
Document exactly what the RX interrupt mitigation timers do.
2011-10-02 13:51:26 +00:00
/*
* This initialises the RX interrupt mitigation timers.
*
* The mitigation timers begin at idle and are triggered
* upon the RXOK of a single frame (or sub-frame, for A-MPDU.)
* Then, the RX mitigation interrupt will fire:
*
* + 250uS after the last RX'ed frame, or
* + 700uS after the first RX'ed frame
*
* Thus, the LAST field dictates the extra latency
* induced by the RX mitigation method and the FIRST
* field dictates how long to delay before firing an
* RX mitigation interrupt.
*
* Please note this only seems to be for RXOK frames;
* not CRC or PHY error frames.
*
*/
Disable TX interrupt mitigation just for the time being. There are some timing concerns which I've yet to fully map out. In any case, there's an existing software driven mitigation method for TX interrupts and when TX'ing 11n frames, the whole frame itself generates an interrupt rather then the subframes.
2011-10-02 13:43:06 +00:00
OS_REG_RMW_FIELD(ah, AR_RIMT, AR_RIMT_LAST, 250);
OS_REG_RMW_FIELD(ah, AR_RIMT, AR_RIMT_FIRST, 700);
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
#endif
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416InitBB(ah, chan);
/* Setup compression registers */
ar5212SetCompRegs(ah); /* XXX not needed? */
/*
* 5416 baseband will check the per rate power table
* and select the lower of the two
*/
ackTpcPow = 63;
ctsTpcPow = 63;
chirpTpcPow = 63;
powerVal = SM(ackTpcPow, AR_TPC_ACK) |
SM(ctsTpcPow, AR_TPC_CTS) |
SM(chirpTpcPow, AR_TPC_CHIRP);
OS_REG_WRITE(ah, AR_TPC, powerVal);
if (!ar5416InitCal(ah, chan))
FAIL(HAL_ESELFTEST);
Add ar5416RestoreChainMask() which will undo any AR5416 specific chainmask overriding after calibration. This will get set for other two chain radios, such as AR9280 and later on, AR9287. It should however be a nul operation.
2011-01-26 10:48:29 +00:00
ar5416RestoreChainMask(ah);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
AH_PRIVATE(ah)->ah_opmode = opmode; /* record operating mode */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (bChannelChange && !IEEE80211_IS_CHAN_DFS(chan))
chan->ic_state &= ~IEEE80211_CHANSTATE_CWINT;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Import some HOWL (AR9130) related fixes from Atheros. Obtained from: Atheros
2011-05-05 02:59:31 +00:00
if (AR_SREV_HOWL(ah)) {
/*
* Enable the MBSSID block-ack fix for HOWL.
* This feature is only supported on Howl 1.4, but it is safe to
* set bit 22 of STA_ID1 on other Howl revisions (1.1, 1.2, 1.3),
* since bit 22 is unused in those Howl revisions.
*/
unsigned int reg;
reg = (OS_REG_READ(ah, AR_STA_ID1) | (1<<22));
OS_REG_WRITE(ah,AR_STA_ID1, reg);
ath_hal_printf(ah, "MBSSID Set bit 22 of AR_STA_ID 0x%x\n", reg);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_RESET, "%s: done\n", __func__);
OS_MARK(ah, AH_MARK_RESET_DONE, 0);
return AH_TRUE;
bad:
OS_MARK(ah, AH_MARK_RESET_DONE, ecode);
fix return status handling by ar5XXXReset; this is the reason the driver sometimes reports reset failed w/ status 0
2009-01-23 05:33:48 +00:00
if (status != AH_NULL)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
*status = ecode;
return AH_FALSE;
#undef FAIL
#undef N
}
#if 0
/*
* This channel change evaluates whether the selected hardware can
* perform a synthesizer-only channel change (no reset). If the
* TX is not stopped, or the RFBus cannot be granted in the given
* time, the function returns false as a reset is necessary
*/
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416ChannelChange(struct ath_hal *ah, const structu ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
uint32_t ulCount;
uint32_t data, synthDelay, qnum;
uint16_t rfXpdGain[4];
struct ath_hal_5212 *ahp = AH5212(ah);
HAL_CHANNEL_INTERNAL *ichan;
/*
* Map public channel to private.
*/
ichan = ath_hal_checkchannel(ah, chan);
/* TX must be stopped or RF Bus grant will not work */
for (qnum = 0; qnum < AH_PRIVATE(ah)->ah_caps.halTotalQueues; qnum++) {
if (ar5212NumTxPending(ah, qnum)) {
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: frames pending on queue %d\n", __func__, qnum);
return AH_FALSE;
}
}
/*
* Kill last Baseband Rx Frame - Request analog bus grant
*/
OS_REG_WRITE(ah, AR_PHY_RFBUS_REQ, AR_PHY_RFBUS_REQ_REQUEST);
if (!ath_hal_wait(ah, AR_PHY_RFBUS_GNT, AR_PHY_RFBUS_GRANT_EN, AR_PHY_RFBUS_GRANT_EN)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: could not kill baseband rx\n",
__func__);
return AH_FALSE;
}
ar5416Set11nRegs(ah, chan); /* NB: setup 5416-specific regs */
/* Change the synth */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (!ar5212SetChannel(ah, chan))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_FALSE;
/* Setup the transmit power values. */
Don't call ar5416SetTransmitPower() directly from ar5416SetTxPowerLimit(); this is incorrect for Kite (AR9285) and any future chipsets that override the EEPROM related routines. It meant that a direct call to set the TX power would call the v14 EEPROM AR5416/AR9280 calibration routines, rather than the v4k EEPROM routines for the AR9285. It thus read the incorrect values from the EEPROM and programmed garbage PDADC and TX power values into the hardware.
2011-03-11 03:46:27 +00:00
if (!ah->ah_setTxPower(ah, chan, rfXpdGain)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: error init'ing transmit power\n", __func__);
return AH_FALSE;
}
/*
* Wait for the frequency synth to settle (synth goes on
* via PHY_ACTIVE_EN). Read the phy active delay register.
* Value is in 100ns increments.
*/
data = OS_REG_READ(ah, AR_PHY_RX_DELAY) & AR_PHY_RX_DELAY_DELAY;
if (IS_CHAN_CCK(ichan)) {
synthDelay = (4 * data) / 22;
} else {
synthDelay = data / 10;
}
OS_DELAY(synthDelay + BASE_ACTIVATE_DELAY);
/* Release the RFBus Grant */
OS_REG_WRITE(ah, AR_PHY_RFBUS_REQ, 0);
/* Write delta slope for OFDM enabled modes (A, G, Turbo) */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_OFDM(ichan)|| IEEE80211_IS_CHAN_HT(chan)) {
HALASSERT(AH_PRIVATE(ah)->ah_eeversion >= AR_EEPROM_VER5_3);
ar5212SetSpurMitigation(ah, chan);
ar5416SetDeltaSlope(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/* XXX spur mitigation for Melin */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (!IEEE80211_IS_CHAN_DFS(chan))
chan->ic_state &= ~IEEE80211_CHANSTATE_CWINT;
ichan->channel_time = 0;
Update the TSF and next-TBTT methods to work for the AR5416 and later NICs. This is another commit in a series of TDMA support fixes for the 11n NICs. * Move ath_hal_getnexttbtt() into the HAL; write methods for it. This returns a timer value in TSF, rather than TU. * Move ath_hal_getcca() and ath_hal_setcca() into the HAL too, where they likely now belong. * Create a new HAL capability: HAL_CAP_LONG_RXDESC_TSF. The pre-11n NICs write 15 bit TSF snapshots into the RX descriptor; the AR5416 and later write 32 bit TSF snapshots into the RX descriptor. * Use the new capability to choose between 15 and 31 bit TSF adjustment functions in ath_extend_tsf(). * Write ar5416GetTsf64() and ar5416SetTsf64() methods. ar5416GetTsf64() tries to compensate for TSF changes at the 32 bit boundary. According to yin, this fixes the TDMA beaconing on 11n chipsets and TDMA stations can now associate/talk, but there are still issues with traffic stability which need to be investigated. The ath_hal_extendtsf() function is also used in RX packet timestamping; this may improve adhoc mode on the 11n chipsets. It also will affect the timestamps seen in radiotap frames. Submitted by: Kang Yin Su <cantona@cantona.net> Approved by: re (kib)
2011-09-08 01:23:05 +00:00
ichan->tsf_last = ar5416GetTsf64(ah);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5212TxEnable(ah, AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
}
#endif
static void
ar5416InitDMA(struct ath_hal *ah)
{
Couple of suggestions from Sam regarding latest commit: o rename the new variables to comply with the naming scheme o move the new variables to an AR5212 specific struct o use ahp when available o revert to previous ts_flags check
2010-03-02 12:59:42 +00:00
struct ath_hal_5212 *ahp = AH5212(ah);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* set AHB_MODE not to do cacheline prefetches
*/
OS_REG_SET_BIT(ah, AR_AHB_MODE, AR_AHB_PREFETCH_RD_EN);
/*
* let mac dma reads be in 128 byte chunks
*/
OS_REG_WRITE(ah, AR_TXCFG,
(OS_REG_READ(ah, AR_TXCFG) & ~AR_TXCFG_DMASZ_MASK) | AR_TXCFG_DMASZ_128B);
/*
* let mac dma writes be in 128 byte chunks
*/
Be slightly more paranoid with the TX DMA buffer maximum threshold. Specifically - never jack the TX FIFO threshold up to the absolute maximum; always leave enough space for two DMA transactions to appear. This is a paranoia from the Linux ath9k driver. It can't hurt. Obtained from: Linux ath9k
2013-02-21 08:42:40 +00:00
/*
* XXX If you change this, you must change the headroom
* assigned in ah_maxTxTrigLev - see ar5416InitState().
*/
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_RXCFG,
(OS_REG_READ(ah, AR_RXCFG) & ~AR_RXCFG_DMASZ_MASK) | AR_RXCFG_DMASZ_128B);
Properly setup the TX FIFO threshold for AR5416 based chipsets, including the AR9285. This seems to fix some users's problems. Submitted by: Jorge Boncompte [DTI2] <jorge at dti2.net>
2010-03-01 17:04:19 +00:00
/* restore TX trigger level */
OS_REG_WRITE(ah, AR_TXCFG,
(OS_REG_READ(ah, AR_TXCFG) &~ AR_FTRIG) |
Couple of suggestions from Sam regarding latest commit: o rename the new variables to comply with the naming scheme o move the new variables to an AR5212 specific struct o use ahp when available o revert to previous ts_flags check
2010-03-02 12:59:42 +00:00
SM(ahp->ah_txTrigLev, AR_FTRIG));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Setup receive FIFO threshold to hold off TX activities
*/
OS_REG_WRITE(ah, AR_RXFIFO_CFG, 0x200);
/*
* reduce the number of usable entries in PCU TXBUF to avoid
* wrap around.
*/
Import a fix from the ath9k - reduce the TX FIFO size for Kite (AR9285.)
2011-04-03 12:02:49 +00:00
if (AR_SREV_KITE(ah))
/*
* For AR9285 the number of Fifos are reduced to half.
* So set the usable tx buf size also to half to
* avoid data/delimiter underruns
*/
OS_REG_WRITE(ah, AR_PCU_TXBUF_CTRL, AR_9285_PCU_TXBUF_CTRL_USABLE_SIZE);
else
OS_REG_WRITE(ah, AR_PCU_TXBUF_CTRL, AR_PCU_TXBUF_CTRL_USABLE_SIZE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
static void
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416InitBB(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
uint32_t synthDelay;
/*
* Wait for the frequency synth to settle (synth goes on
* via AR_PHY_ACTIVE_EN). Read the phy active delay register.
* Value is in 100ns increments.
*/
synthDelay = OS_REG_READ(ah, AR_PHY_RX_DELAY) & AR_PHY_RX_DELAY_DELAY;
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_CCK(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
synthDelay = (4 * synthDelay) / 22;
} else {
synthDelay /= 10;
}
/* Turn on PLL on 5416 */
HALDEBUG(ah, HAL_DEBUG_RESET, "%s %s channel\n",
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
__func__, IEEE80211_IS_CHAN_5GHZ(chan) ? "5GHz" : "2GHz");
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Activate the PHY (includes baseband activate and synthesizer on) */
OS_REG_WRITE(ah, AR_PHY_ACTIVE, AR_PHY_ACTIVE_EN);
/*
* If the AP starts the calibration before the base band timeout
Update to later code from my repository: o many bug fixes o add new periodic calibration api o break up 5416 periodic calibration code in preparation for 928x o move get noise floor to rf backends o 5416-specific ani (still disabled) o modularize 5210 eeprom format a la other eeprom formats o start cleaning up regdomain code o prepare for proper 1/2 and 1/4 width channel support o bring back 900MHz card support o clean up 5212 rf version handling o add 1/2 and 1/4 width channel support for 5212 parts o split 5212 rfgain handling out o improve ani debugging o add AH_USE_INIPDGAIN compile option o purge a bunch of dead 5212 state o add 1/2 and 1/4 rate modes o remove HAL_CAP_CHAN_HALFRATE and HAL_CAP_CHAN_QUARTERRATE; the same info can now be deduced from the set of supported modes
2008-11-28 00:48:05 +00:00
* completes we could get rx_clear false triggering. Add an
* extra BASE_ACTIVATE_DELAY usecs to ensure this condition
virgin import of ath hal
2008-11-28 00:03:41 +00:00
* does not happen.
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HALF(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_DELAY((synthDelay << 1) + BASE_ACTIVATE_DELAY);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
} else if (IEEE80211_IS_CHAN_QUARTER(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_DELAY((synthDelay << 2) + BASE_ACTIVATE_DELAY);
} else {
OS_DELAY(synthDelay + BASE_ACTIVATE_DELAY);
}
}
static void
ar5416InitIMR(struct ath_hal *ah, HAL_OPMODE opmode)
{
struct ath_hal_5212 *ahp = AH5212(ah);
/*
* Setup interrupt handling. Note that ar5212ResetTxQueue
* manipulates the secondary IMR's as queues are enabled
* and disabled. This is done with RMW ops to insure the
* settings we make here are preserved.
*/
ahp->ah_maskReg = AR_IMR_TXERR | AR_IMR_TXURN
| AR_IMR_RXERR | AR_IMR_RXORN
| AR_IMR_BCNMISC;
Introduce AH_AR5416_INTERRUPT_MITIGATION which enables interrupt mitigation for the AR5416 and later. Rename the older HAL option to use this.
2011-03-31 08:48:05 +00:00
#ifdef AH_AR5416_INTERRUPT_MITIGATION
Correct/complete a partially-disabled TX interrupt mitigation configuration. Although a previous commit disabled TX interrupt mitigation handling and configuration, the mask register bits weren't setup correctly.
2011-10-25 23:17:53 +00:00
ahp->ah_maskReg |= AR_IMR_RXINTM | AR_IMR_RXMINTR;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#else
Correct/complete a partially-disabled TX interrupt mitigation configuration. Although a previous commit disabled TX interrupt mitigation handling and configuration, the mask register bits weren't setup correctly.
2011-10-25 23:17:53 +00:00
ahp->ah_maskReg |= AR_IMR_RXOK;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
Correct/complete a partially-disabled TX interrupt mitigation configuration. Although a previous commit disabled TX interrupt mitigation handling and configuration, the mask register bits weren't setup correctly.
2011-10-25 23:17:53 +00:00
ahp->ah_maskReg |= AR_IMR_TXOK;
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (opmode == HAL_M_HOSTAP)
ahp->ah_maskReg |= AR_IMR_MIB;
OS_REG_WRITE(ah, AR_IMR, ahp->ah_maskReg);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
#ifdef ADRIAN_NOTYET
/* This is straight from ath9k */
if (! AR_SREV_HOWL(ah)) {
OS_REG_WRITE(ah, AR_INTR_SYNC_CAUSE, 0xFFFFFFFF);
OS_REG_WRITE(ah, AR_INTR_SYNC_ENABLE, AR_INTR_SYNC_DEFAULT);
OS_REG_WRITE(ah, AR_INTR_SYNC_MASK, 0);
}
#endif
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Enable bus errors that are OR'd to set the HIUERR bit */
#if 0
OS_REG_WRITE(ah, AR_IMR_S2,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
OS_REG_READ(ah, AR_IMR_S2) | AR_IMR_S2_GTT | AR_IMR_S2_CST);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
}
static void
ar5416InitQoS(struct ath_hal *ah)
{
/* QoS support */
OS_REG_WRITE(ah, AR_QOS_CONTROL, 0x100aa); /* XXX magic */
OS_REG_WRITE(ah, AR_QOS_SELECT, 0x3210); /* XXX magic */
/* Turn on NOACK Support for QoS packets */
OS_REG_WRITE(ah, AR_NOACK,
SM(2, AR_NOACK_2BIT_VALUE) |
SM(5, AR_NOACK_BIT_OFFSET) |
SM(0, AR_NOACK_BYTE_OFFSET));
/*
* initialize TXOP for all TIDs
*/
OS_REG_WRITE(ah, AR_TXOP_X, AR_TXOP_X_VAL);
OS_REG_WRITE(ah, AR_TXOP_0_3, 0xFFFFFFFF);
OS_REG_WRITE(ah, AR_TXOP_4_7, 0xFFFFFFFF);
OS_REG_WRITE(ah, AR_TXOP_8_11, 0xFFFFFFFF);
OS_REG_WRITE(ah, AR_TXOP_12_15, 0xFFFFFFFF);
}
static void
ar5416InitUserSettings(struct ath_hal *ah)
{
struct ath_hal_5212 *ahp = AH5212(ah);
/* Restore user-specified settings */
if (ahp->ah_miscMode != 0)
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
OS_REG_WRITE(ah, AR_MISC_MODE, OS_REG_READ(ah, AR_MISC_MODE)
| ahp->ah_miscMode);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (ahp->ah_sifstime != (u_int) -1)
ar5212SetSifsTime(ah, ahp->ah_sifstime);
if (ahp->ah_slottime != (u_int) -1)
ar5212SetSlotTime(ah, ahp->ah_slottime);
if (ahp->ah_acktimeout != (u_int) -1)
ar5212SetAckTimeout(ah, ahp->ah_acktimeout);
if (ahp->ah_ctstimeout != (u_int) -1)
ar5212SetCTSTimeout(ah, ahp->ah_ctstimeout);
if (AH_PRIVATE(ah)->ah_diagreg != 0)
OS_REG_WRITE(ah, AR_DIAG_SW, AH_PRIVATE(ah)->ah_diagreg);
Add global TX timeout handling. The global TX timeout counter increments whenever a frame is ready to be transmitted and the medium is busy.
2011-04-18 12:15:43 +00:00
if (AH5416(ah)->ah_globaltxtimeout != (u_int) -1)
ar5416SetGlobalTxTimeout(ah, AH5416(ah)->ah_globaltxtimeout);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Break out the RF mode setup into ar5416SetRfMode(), mirroring what ath9k does.
2011-03-22 00:52:44 +00:00
static void
ar5416SetRfMode(struct ath_hal *ah, const struct ieee80211_channel *chan)
{
uint32_t rfMode;
if (chan == AH_NULL)
return;
/* treat channel B as channel G , no B mode suport in owl */
rfMode = IEEE80211_IS_CHAN_CCK(chan) ?
AR_PHY_MODE_DYNAMIC : AR_PHY_MODE_OFDM;
if (AR_SREV_MERLIN_20(ah) && IS_5GHZ_FAST_CLOCK_EN(ah, chan)) {
/* phy mode bits for 5GHz channels require Fast Clock */
rfMode |= AR_PHY_MODE_DYNAMIC
| AR_PHY_MODE_DYN_CCK_DISABLE;
} else if (!AR_SREV_MERLIN_10_OR_LATER(ah)) {
rfMode |= IEEE80211_IS_CHAN_5GHZ(chan) ?
AR_PHY_MODE_RF5GHZ : AR_PHY_MODE_RF2GHZ;
}
Add some (totally untested!) code to correctly set the RF half/quarter mode configuration registers. This is apparently required for correct behaviour, but also requires the chip to actually officially support it. Sponsored by: Hobnob, Inc.
2011-11-19 21:12:35 +00:00
Break out the RF mode setup into ar5416SetRfMode(), mirroring what ath9k does.
2011-03-22 00:52:44 +00:00
OS_REG_WRITE(ah, AR_PHY_MODE, rfMode);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Places the hardware into reset and then pulls it out of reset
*/
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416ChipReset(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
OS_MARK(ah, AH_MARK_CHIPRESET, chan ? chan->ic_freq : 0);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
Add a new HAL parameter which forces a full reset rather than warm reset. This forces a full reset of the baseband/radio and seems needed to clear some issues (with Merlin at least) when the baseband gets confused in a very noisy environment. Sponsored by: Hobnob, Inc.
2011-11-09 05:30:24 +00:00
* Warm reset is optimistic for open-loop TX power control.
virgin import of ath hal
2008-11-28 00:03:41 +00:00
*/
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
if (AR_SREV_MERLIN(ah) &&
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ath_hal_eepromGetFlag(ah, AR_EEP_OL_PWRCTRL)) {
if (!ar5416SetResetReg(ah, HAL_RESET_POWER_ON))
return AH_FALSE;
Add a new HAL parameter which forces a full reset rather than warm reset. This forces a full reset of the baseband/radio and seems needed to clear some issues (with Merlin at least) when the baseband gets confused in a very noisy environment. Sponsored by: Hobnob, Inc.
2011-11-09 05:30:24 +00:00
} else if (ah->ah_config.ah_force_full_reset) {
if (!ar5416SetResetReg(ah, HAL_RESET_POWER_ON))
return AH_FALSE;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
if (!ar5416SetResetReg(ah, HAL_RESET_WARM))
return AH_FALSE;
}
/* Bring out of sleep mode (AGAIN) */
if (!ar5416SetPowerMode(ah, HAL_PM_AWAKE, AH_TRUE))
return AH_FALSE;
Fiddle with the PLL initialisation order to match ath9k/Atheros HAL. This seems to make the AR9160 behave better during heavy scanning, where before it'd hang and require a hard reset to recover. Obtained From: Linux ath9k, Atheros
2011-05-08 07:21:09 +00:00
#ifdef notyet
ahp->ah_chipFullSleep = AH_FALSE;
#endif
Break out the PLL setup into an overridable method. The only method right now is ar5416InitPLL() which handles multiple chipsets; this can now be overridden by newer chipset HAL code.
2011-04-24 15:53:57 +00:00
AH5416(ah)->ah_initPLL(ah, chan);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Perform warm reset before the mode/PLL/turbo registers
* are changed in order to deactivate the radio. Mode changes
* with an active radio can result in corrupted shifts to the
* radio device.
*/
Fiddle with the PLL initialisation order to match ath9k/Atheros HAL. This seems to make the AR9160 behave better during heavy scanning, where before it'd hang and require a hard reset to recover. Obtained From: Linux ath9k, Atheros
2011-05-08 07:21:09 +00:00
ar5416SetRfMode(ah, chan);
Break out the RF mode setup into ar5416SetRfMode(), mirroring what ath9k does.
2011-03-22 00:52:44 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
}
/*
* Delta slope coefficient computation.
* Required for OFDM operation.
*/
static void
ar5416GetDeltaSlopeValues(struct ath_hal *ah, uint32_t coef_scaled,
uint32_t *coef_mantissa, uint32_t *coef_exponent)
{
#define COEF_SCALE_S 24
uint32_t coef_exp, coef_man;
/*
* ALGO -> coef_exp = 14-floor(log2(coef));
* floor(log2(x)) is the highest set bit position
*/
for (coef_exp = 31; coef_exp > 0; coef_exp--)
if ((coef_scaled >> coef_exp) & 0x1)
break;
/* A coef_exp of 0 is a legal bit position but an unexpected coef_exp */
HALASSERT(coef_exp);
coef_exp = 14 - (coef_exp - COEF_SCALE_S);
/*
* ALGO -> coef_man = floor(coef* 2^coef_exp+0.5);
* The coefficient is already shifted up for scaling
*/
coef_man = coef_scaled + (1 << (COEF_SCALE_S - coef_exp - 1));
*coef_mantissa = coef_man >> (COEF_SCALE_S - coef_exp);
*coef_exponent = coef_exp - 16;
#undef COEF_SCALE_S
}
void
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416SetDeltaSlope(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
#define INIT_CLOCKMHZSCALED 0x64000000
uint32_t coef_scaled, ds_coef_exp, ds_coef_man;
misc fixups, mostly for code not compiled yet
2009-02-24 00:29:32 +00:00
uint32_t clockMhzScaled;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
CHAN_CENTERS centers;
/* half and quarter rate can divide the scaled clock by 2 or 4 respectively */
/* scale for selected channel bandwidth */
misc fixups, mostly for code not compiled yet
2009-02-24 00:29:32 +00:00
clockMhzScaled = INIT_CLOCKMHZSCALED;
if (IEEE80211_IS_CHAN_TURBO(chan))
clockMhzScaled <<= 1;
else if (IEEE80211_IS_CHAN_HALF(chan))
clockMhzScaled >>= 1;
else if (IEEE80211_IS_CHAN_QUARTER(chan))
clockMhzScaled >>= 2;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* ALGO -> coef = 1e8/fcarrier*fclock/40;
* scaled coef to provide precision for this floating calculation
*/
ar5416GetChannelCenters(ah, chan, &centers);
coef_scaled = clockMhzScaled / centers.synth_center;
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416GetDeltaSlopeValues(ah, coef_scaled, &ds_coef_man, &ds_coef_exp);
OS_REG_RMW_FIELD(ah, AR_PHY_TIMING3,
AR_PHY_TIMING3_DSC_MAN, ds_coef_man);
OS_REG_RMW_FIELD(ah, AR_PHY_TIMING3,
AR_PHY_TIMING3_DSC_EXP, ds_coef_exp);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* For Short GI,
* scaled coeff is 9/10 that of normal coeff
*/
coef_scaled = (9 * coef_scaled)/10;
ar5416GetDeltaSlopeValues(ah, coef_scaled, &ds_coef_man, &ds_coef_exp);
/* for short gi */
OS_REG_RMW_FIELD(ah, AR_PHY_HALFGI,
AR_PHY_HALFGI_DSC_MAN, ds_coef_man);
OS_REG_RMW_FIELD(ah, AR_PHY_HALFGI,
AR_PHY_HALFGI_DSC_EXP, ds_coef_exp);
#undef INIT_CLOCKMHZSCALED
}
/*
* Set a limit on the overall output power. Used for dynamic
* transmit power control and the like.
*
* NB: limit is in units of 0.5 dbM.
*/
HAL_BOOL
ar5416SetTxPowerLimit(struct ath_hal *ah, uint32_t limit)
{
uint16_t dummyXpdGains[2];
AH_PRIVATE(ah)->ah_powerLimit = AH_MIN(limit, MAX_RATE_POWER);
Don't call ar5416SetTransmitPower() directly from ar5416SetTxPowerLimit(); this is incorrect for Kite (AR9285) and any future chipsets that override the EEPROM related routines. It meant that a direct call to set the TX power would call the v14 EEPROM AR5416/AR9280 calibration routines, rather than the v4k EEPROM routines for the AR9285. It thus read the incorrect values from the EEPROM and programmed garbage PDADC and TX power values into the hardware.
2011-03-11 03:46:27 +00:00
return ah->ah_setTxPower(ah, AH_PRIVATE(ah)->ah_curchan,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
dummyXpdGains);
}
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416GetChipPowerLimits(struct ath_hal *ah,
struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
struct ath_hal_5212 *ahp = AH5212(ah);
int16_t minPower, maxPower;
/*
* Get Pier table max and min powers.
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (ahp->ah_rfHal->getChannelMaxMinPower(ah, chan, &maxPower, &minPower)) {
/* NB: rf code returns 1/4 dBm units, convert */
chan->ic_maxpower = maxPower / 2;
chan->ic_minpower = minPower / 2;
} else {
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: no min/max power for %u/0x%x\n",
__func__, chan->ic_freq, chan->ic_flags);
chan->ic_maxpower = AR5416_MAX_RATE_POWER;
chan->ic_minpower = 0;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
HALDEBUG(ah, HAL_DEBUG_RESET,
"Chan %d: MaxPow = %d MinPow = %d\n",
chan->ic_freq, chan->ic_maxpower, chan->ic_minpower);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
}
Eliminate code duplication between AR5416/AR9160/AR9280 and AR9285. Writing the TX power registers is the same between all of these chips and later NICs (AR9287, AR9271 USB, etc.) so this will reduce code duplication when those NICs are added to the HAL.
2011-04-24 14:50:29 +00:00
/**************************************************************
* ar5416WriteTxPowerRateRegisters
*
* Write the TX power rate registers from the raw values given
* in ratesArray[].
*
* The CCK and HT40 rate registers are only written if needed.
* HT20 and 11g/11a OFDM rate registers are always written.
*
* The values written are raw values which should be written
* to the registers - so it's up to the caller to pre-adjust
* them (eg CCK power offset value, or Merlin TX power offset,
* etc.)
*/
void
ar5416WriteTxPowerRateRegisters(struct ath_hal *ah,
const struct ieee80211_channel *chan, const int16_t ratesArray[])
{
#define POW_SM(_r, _s) (((_r) & 0x3f) << (_s))
/* Write the OFDM power per rate set */
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE1,
POW_SM(ratesArray[rate18mb], 24)
| POW_SM(ratesArray[rate12mb], 16)
| POW_SM(ratesArray[rate9mb], 8)
| POW_SM(ratesArray[rate6mb], 0)
);
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE2,
POW_SM(ratesArray[rate54mb], 24)
| POW_SM(ratesArray[rate48mb], 16)
| POW_SM(ratesArray[rate36mb], 8)
| POW_SM(ratesArray[rate24mb], 0)
);
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
/* Write the CCK power per rate set */
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE3,
POW_SM(ratesArray[rate2s], 24)
| POW_SM(ratesArray[rate2l], 16)
| POW_SM(ratesArray[rateXr], 8) /* XR target power */
| POW_SM(ratesArray[rate1l], 0)
);
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE4,
POW_SM(ratesArray[rate11s], 24)
| POW_SM(ratesArray[rate11l], 16)
| POW_SM(ratesArray[rate5_5s], 8)
| POW_SM(ratesArray[rate5_5l], 0)
);
HALDEBUG(ah, HAL_DEBUG_RESET,
"%s AR_PHY_POWER_TX_RATE3=0x%x AR_PHY_POWER_TX_RATE4=0x%x\n",
__func__, OS_REG_READ(ah,AR_PHY_POWER_TX_RATE3),
OS_REG_READ(ah,AR_PHY_POWER_TX_RATE4));
}
/* Write the HT20 power per rate set */
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE5,
POW_SM(ratesArray[rateHt20_3], 24)
| POW_SM(ratesArray[rateHt20_2], 16)
| POW_SM(ratesArray[rateHt20_1], 8)
| POW_SM(ratesArray[rateHt20_0], 0)
);
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE6,
POW_SM(ratesArray[rateHt20_7], 24)
| POW_SM(ratesArray[rateHt20_6], 16)
| POW_SM(ratesArray[rateHt20_5], 8)
| POW_SM(ratesArray[rateHt20_4], 0)
);
if (IEEE80211_IS_CHAN_HT40(chan)) {
/* Write the HT40 power per rate set */
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE7,
POW_SM(ratesArray[rateHt40_3], 24)
| POW_SM(ratesArray[rateHt40_2], 16)
| POW_SM(ratesArray[rateHt40_1], 8)
| POW_SM(ratesArray[rateHt40_0], 0)
);
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE8,
POW_SM(ratesArray[rateHt40_7], 24)
| POW_SM(ratesArray[rateHt40_6], 16)
| POW_SM(ratesArray[rateHt40_5], 8)
| POW_SM(ratesArray[rateHt40_4], 0)
);
/* Write the Dup/Ext 40 power per rate set */
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE9,
POW_SM(ratesArray[rateExtOfdm], 24)
| POW_SM(ratesArray[rateExtCck], 16)
| POW_SM(ratesArray[rateDupOfdm], 8)
| POW_SM(ratesArray[rateDupCck], 0)
);
}
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
/*
* Set max power to 30 dBm and, optionally,
* enable TPC in tx descriptors.
*/
OS_REG_WRITE(ah, AR_PHY_POWER_TX_RATE_MAX, MAX_RATE_POWER |
(AH5212(ah)->ah_tpcEnabled ? AR_PHY_POWER_TX_RATE_MAX_TPC_ENABLE : 0));
#undef POW_SM
Eliminate code duplication between AR5416/AR9160/AR9280 and AR9285. Writing the TX power registers is the same between all of these chips and later NICs (AR9287, AR9271 USB, etc.) so this will reduce code duplication when those NICs are added to the HAL.
2011-04-24 14:50:29 +00:00
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/**************************************************************
* ar5416SetTransmitPower
*
* Set the transmit power in the baseband for the given
* operating channel and mode.
*/
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416SetTransmitPower(struct ath_hal *ah,
const struct ieee80211_channel *chan, uint16_t *rfXpdGain)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
#define N(a) (sizeof (a) / sizeof (a[0]))
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
#define POW_SM(_r, _s) (((_r) & 0x3f) << (_s))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
MODAL_EEP_HEADER *pModal;
struct ath_hal_5212 *ahp = AH5212(ah);
int16_t txPowerIndexOffset = 0;
int i;
uint16_t cfgCtl;
uint16_t powerLimit;
uint16_t twiceAntennaReduction;
uint16_t twiceMaxRegulatoryPower;
int16_t maxPower;
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
HAL_EEPROM_v14 *ee = AH_PRIVATE(ah)->ah_eeprom;
struct ar5416eeprom *pEepData = &ee->ee_base;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALASSERT(AH_PRIVATE(ah)->ah_eeversion >= AR_EEPROM_VER14_1);
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
/*
* Default to 2, is overridden based on the EEPROM version / value.
*/
AH5416(ah)->ah_ht40PowerIncForPdadc = 2;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Setup info for the actual eeprom */
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
OS_MEMZERO(AH5416(ah)->ah_ratesArray, sizeof(AH5416(ah)->ah_ratesArray));
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
cfgCtl = ath_hal_getctl(ah, chan);
powerLimit = chan->ic_maxregpower * 2;
twiceAntennaReduction = chan->ic_maxantgain;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
twiceMaxRegulatoryPower = AH_MIN(MAX_RATE_POWER, AH_PRIVATE(ah)->ah_powerLimit);
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal = &pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)];
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_RESET, "%s Channel=%u CfgCtl=%u\n",
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
__func__,chan->ic_freq, cfgCtl );
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (IS_EEP_MINOR_V2(ah)) {
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
AH5416(ah)->ah_ht40PowerIncForPdadc = pModal->ht40PowerIncForPdadc;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
if (!ar5416SetPowerPerRateTable(ah, pEepData, chan,
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
&AH5416(ah)->ah_ratesArray[0],
cfgCtl,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
twiceAntennaReduction,
twiceMaxRegulatoryPower, powerLimit)) {
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: unable to set tx power per rate table\n", __func__);
return AH_FALSE;
}
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
if (!AH5416(ah)->ah_setPowerCalTable(ah, pEepData, chan, &txPowerIndexOffset)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: unable to set power table\n",
__func__);
return AH_FALSE;
}
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
maxPower = AH_MAX(AH5416(ah)->ah_ratesArray[rate6mb],
AH5416(ah)->ah_ratesArray[rateHt20_0]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
maxPower = AH_MAX(maxPower, AH5416(ah)->ah_ratesArray[rate1l]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40(chan)) {
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
maxPower = AH_MAX(maxPower, AH5416(ah)->ah_ratesArray[rateHt40_0]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
ahp->ah_tx6PowerInHalfDbm = maxPower;
AH_PRIVATE(ah)->ah_maxPowerLevel = maxPower;
ahp->ah_txPowerIndexOffset = txPowerIndexOffset;
/*
* txPowerIndexOffset is set by the SetPowerTable() call -
* adjust the rate table (0 offset if rates EEPROM not loaded)
*/
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
for (i = 0; i < N(AH5416(ah)->ah_ratesArray); i++) {
AH5416(ah)->ah_ratesArray[i] =
(int16_t)(txPowerIndexOffset + AH5416(ah)->ah_ratesArray[i]);
if (AH5416(ah)->ah_ratesArray[i] > AR5416_MAX_RATE_POWER)
AH5416(ah)->ah_ratesArray[i] = AR5416_MAX_RATE_POWER;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
#ifdef AH_EEPROM_DUMP
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/*
* Dump the rate array whilst it represents the intended dBm*2
* values versus what's being adjusted before being programmed
* in. Keep this in mind if you code up this function and enable
* this debugging; the values won't necessarily be what's being
* programmed into the hardware.
*/
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
ar5416PrintPowerPerRate(ah, AH5416(ah)->ah_ratesArray);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/*
* Merlin and later have a power offset, so subtract
* pwr_table_offset * 2 from each value. The default
* power offset is -5 dBm - ie, a register value of 0
* equates to a TX power of -5 dBm.
*/
if (AR_SREV_MERLIN_20_OR_LATER(ah)) {
int8_t pwr_table_offset;
(void) ath_hal_eepromGet(ah, AR_EEP_PWR_TABLE_OFFSET,
&pwr_table_offset);
/* Underflow power gets clamped at raw value 0 */
/* Overflow power gets camped at AR5416_MAX_RATE_POWER */
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
for (i = 0; i < N(AH5416(ah)->ah_ratesArray); i++) {
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/*
* + pwr_table_offset is in dBm
* + ratesArray is in 1/2 dBm
*/
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
AH5416(ah)->ah_ratesArray[i] -= (pwr_table_offset * 2);
if (AH5416(ah)->ah_ratesArray[i] < 0)
AH5416(ah)->ah_ratesArray[i] = 0;
else if (AH5416(ah)->ah_ratesArray[i] > AR5416_MAX_RATE_POWER)
AH5416(ah)->ah_ratesArray[i] = AR5416_MAX_RATE_POWER;
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
}
}
/*
* Adjust rates for OLC where needed
*
* The following CCK rates need adjusting when doing 2.4ghz
* CCK transmission.
*
* + rate2s, rate2l, rate1l, rate11s, rate11l, rate5_5s, rate5_5l
* + rateExtCck, rateDupCck
*
* They're adjusted here regardless. The hardware then gets
* programmed as needed. 5GHz operation doesn't program in CCK
* rates for legacy mode but they seem to be initialised for
* HT40 regardless of channel type.
*/
if (AR_SREV_MERLIN_20_OR_LATER(ah) &&
ath_hal_eepromGetFlag(ah, AR_EEP_OL_PWRCTRL)) {
int adj[] = {
rate2s, rate2l, rate1l, rate11s, rate11l,
rate5_5s, rate5_5l, rateExtCck, rateDupCck
};
int cck_ofdm_delta = 2;
int i;
for (i = 0; i < N(adj); i++) {
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
AH5416(ah)->ah_ratesArray[adj[i]] -= cck_ofdm_delta;
if (AH5416(ah)->ah_ratesArray[adj[i]] < 0)
AH5416(ah)->ah_ratesArray[adj[i]] = 0;
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
}
}
Eliminate code duplication between AR5416/AR9160/AR9280 and AR9285. Writing the TX power registers is the same between all of these chips and later NICs (AR9287, AR9271 USB, etc.) so this will reduce code duplication when those NICs are added to the HAL.
2011-04-24 14:50:29 +00:00
/*
* Adjust the HT40 power to meet the correct target TX power
* for 40MHz mode, based on TX power curves that are established
* for 20MHz mode.
*
* XXX handle overflow/too high power level?
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40(chan)) {
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
AH5416(ah)->ah_ratesArray[rateHt40_0] +=
AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_1] +=
AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_2] += AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_3] += AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_4] += AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_5] += AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_6] += AH5416(ah)->ah_ht40PowerIncForPdadc;
AH5416(ah)->ah_ratesArray[rateHt40_7] += AH5416(ah)->ah_ht40PowerIncForPdadc;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Eliminate code duplication between AR5416/AR9160/AR9280 and AR9285. Writing the TX power registers is the same between all of these chips and later NICs (AR9287, AR9271 USB, etc.) so this will reduce code duplication when those NICs are added to the HAL.
2011-04-24 14:50:29 +00:00
/* Write the TX power rate registers */
Setup needed tables for TPC on AR5416->AR9287 chips. * Add ah_ratesArray[] to the ar5416 HAL state - this stores the maximum values permissable per rate. * Since different chip EEPROM formats store this value in a different place, store the HT40 power detector increment value in the ar5416 HAL state. * Modify the target power setup code to store the maximum values in the ar5416 HAL state rather than using a local variable. * Add ar5416RateToRateTable() - to convert a hardware rate code to the ratesArray enum / index. * Add ar5416GetTxRatePower() - which goes through the gymnastics required to correctly calculate the target TX power: + Add the power detector increment for ht40; + Take the power offset into account for AR9280 and later; + Offset the TX power correctly when doing open-loop TX power control; + Enforce the per-rate maximum value allowable. Note - setting a TPC value of 0x0 in the TX descriptor on (at least) the AR9160 resulted in the TX power being very high indeed. This didn't happen on the AR9220. I'm guessing it's a chip bug that was fixed at some point. So for now, just assume the AR5416/AR5418 and AR9130 are also suspect and clamp the minimum value here at 1. Tested: * AR5416, AR9160, AR9220 hostap, verified using (2GHz) spectrum analyser * Looked at target TX power in TX descriptor (using athalq) as well as TX power on the spectrum analyser. TODO: * The TX descriptor code sets the target TX power to 0 for AR9285 chips. I'm not yet sure why. Disable this for TPC and ensure that the TPC TX power is set. * AR9280, AR9285, AR9227, AR9287 testing! * 5GHz testing! Quirks: * The per-packet TPC code is only exercised when the tpc sysctl is set to 1. (dev.ath.X.tpc=1.) This needs to be done before you bring the interface up. * When TPC is enabled, setting the TX power doesn't end up with a call through to the HAL to update the maximum TX power. So ensure that you set the TPC sysctl before you bring the interface up and configure a lower TX power or the hardware will be clamped by the lower TX power (at least until the next channel change.) Thanks to Qualcomm Atheros for all the hardware, and Sam Leffler for use of his spectrum analyser to verify the TX channel power.
2013-04-17 07:31:53 +00:00
ar5416WriteTxPowerRateRegisters(ah, chan, AH5416(ah)->ah_ratesArray);
Eliminate code duplication between AR5416/AR9160/AR9280 and AR9285. Writing the TX power registers is the same between all of these chips and later NICs (AR9287, AR9271 USB, etc.) so this will reduce code duplication when those NICs are added to the HAL.
2011-04-24 14:50:29 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Write the Power subtraction for dynamic chain changing, for per-packet powertx */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
OS_REG_WRITE(ah, AR_PHY_POWER_TX_SUB,
POW_SM(pModal->pwrDecreaseFor3Chain, 6)
| POW_SM(pModal->pwrDecreaseFor2Chain, 0)
);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
#undef POW_SM
#undef N
}
/*
* Exported call to check for a recent gain reading and return
* the current state of the thermal calibration gain engine.
*/
HAL_RFGAIN
ar5416GetRfgain(struct ath_hal *ah)
{
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
return (HAL_RFGAIN_INACTIVE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/*
* Places all of hardware into reset
*/
HAL_BOOL
ar5416Disable(struct ath_hal *ah)
{
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
Call the correct chipset power routine when disabling the AR5416 and later NICs.
2011-11-04 13:32:13 +00:00
if (!ar5416SetPowerMode(ah, HAL_PM_AWAKE, AH_TRUE))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_FALSE;
Fiddle with the PLL initialisation order to match ath9k/Atheros HAL. This seems to make the AR9160 behave better during heavy scanning, where before it'd hang and require a hard reset to recover. Obtained From: Linux ath9k, Atheros
2011-05-08 07:21:09 +00:00
if (! ar5416SetResetReg(ah, HAL_RESET_COLD))
return AH_FALSE;
AH5416(ah)->ah_initPLL(ah, AH_NULL);
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
return (AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/*
* Places the PHY and Radio chips into reset. A full reset
* must be called to leave this state. The PCI/MAC/PCU are
* not placed into reset as we must receive interrupt to
* re-enable the hardware.
*/
HAL_BOOL
ar5416PhyDisable(struct ath_hal *ah)
{
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
Fiddle with the PLL initialisation order to match ath9k/Atheros HAL. This seems to make the AR9160 behave better during heavy scanning, where before it'd hang and require a hard reset to recover. Obtained From: Linux ath9k, Atheros
2011-05-08 07:21:09 +00:00
if (! ar5416SetResetReg(ah, HAL_RESET_WARM))
return AH_FALSE;
AH5416(ah)->ah_initPLL(ah, AH_NULL);
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
return (AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/*
* Write the given reset bit mask into the reset register
*/
HAL_BOOL
ar5416SetResetReg(struct ath_hal *ah, uint32_t type)
{
* Force the MAC to wakeup before we try resetting it, to ensure it actually _gets_ reset properly. * Add some more comments describing why things are done. Obtained from: Atheros
2011-11-09 14:34:25 +00:00
/*
* Set force wake
*/
OS_REG_WRITE(ah, AR_RTC_FORCE_WAKE,
AR_RTC_FORCE_WAKE_EN | AR_RTC_FORCE_WAKE_ON_INT);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
switch (type) {
case HAL_RESET_POWER_ON:
return ar5416SetResetPowerOn(ah);
case HAL_RESET_WARM:
case HAL_RESET_COLD:
return ar5416SetReset(ah, type);
default:
preliminary ar9280 support: o add 9280 attach that sets up ini, cal, etc. o new rf backend for 9280 and later parts o split ini setup and spur mitigation support out to methods and provide 9280-specific support o minor fixups to shared code to handle 9280-specific work Obtained from: Atheros (ini values and some code)
2009-03-12 18:18:28 +00:00
HALASSERT(AH_FALSE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_FALSE;
}
}
static HAL_BOOL
ar5416SetResetPowerOn(struct ath_hal *ah)
{
/* Power On Reset (Hard Reset) */
/*
* Set force wake
*
* If the MAC was running, previously calling
* reset will wake up the MAC but it may go back to sleep
* before we can start polling.
* Set force wake stops that
* This must be called before initiating a hard reset.
*/
OS_REG_WRITE(ah, AR_RTC_FORCE_WAKE,
AR_RTC_FORCE_WAKE_EN | AR_RTC_FORCE_WAKE_ON_INT);
/*
* Force the MAC to wakeup before we try resetting it, to ensure it actually _gets_ reset properly. * Add some more comments describing why things are done. Obtained from: Atheros
2011-11-09 14:34:25 +00:00
* PowerOn reset can be used in open loop power control or failure recovery.
* If we do RTC reset while DMA is still running, hardware may corrupt memory.
* Therefore, we need to reset AHB first to stop DMA.
virgin import of ath hal
2008-11-28 00:03:41 +00:00
*/
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
if (! AR_SREV_HOWL(ah))
OS_REG_WRITE(ah, AR_RC, AR_RC_AHB);
* Force the MAC to wakeup before we try resetting it, to ensure it actually _gets_ reset properly. * Add some more comments describing why things are done. Obtained from: Atheros
2011-11-09 14:34:25 +00:00
/*
* RTC reset and clear
*/
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_RTC_RESET, 0);
OS_DELAY(20);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
if (! AR_SREV_HOWL(ah))
OS_REG_WRITE(ah, AR_RC, 0);
Bring in a couple of fixes from the Linux ath9k related to chip hangs. While there, try to make the register write pattern look like what's done by ath9k. MFC after: 3 days
2010-06-01 15:33:10 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_RTC_RESET, 1);
/*
* Poll till RTC is ON
*/
if (!ath_hal_wait(ah, AR_RTC_STATUS, AR_RTC_PM_STATUS_M, AR_RTC_STATUS_ON)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: RTC not waking up\n", __func__);
return AH_FALSE;
}
return ar5416SetReset(ah, HAL_RESET_COLD);
}
static HAL_BOOL
ar5416SetReset(struct ath_hal *ah, int type)
{
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
uint32_t tmpReg, mask;
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
uint32_t rst_flags;
#ifdef AH_SUPPORT_AR9130 /* Because of the AR9130 specific registers */
if (AR_SREV_HOWL(ah)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "[ath] HOWL: Fiddling with derived clk!\n");
uint32_t val = OS_REG_READ(ah, AR_RTC_DERIVED_CLK);
val &= ~AR_RTC_DERIVED_CLK_PERIOD;
val |= SM(1, AR_RTC_DERIVED_CLK_PERIOD);
OS_REG_WRITE(ah, AR_RTC_DERIVED_CLK, val);
(void) OS_REG_READ(ah, AR_RTC_DERIVED_CLK);
}
#endif /* AH_SUPPORT_AR9130 */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Force wake
*/
OS_REG_WRITE(ah, AR_RTC_FORCE_WAKE,
AR_RTC_FORCE_WAKE_EN | AR_RTC_FORCE_WAKE_ON_INT);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
#ifdef AH_SUPPORT_AR9130
if (AR_SREV_HOWL(ah)) {
rst_flags = AR_RTC_RC_MAC_WARM | AR_RTC_RC_MAC_COLD |
AR_RTC_RC_COLD_RESET | AR_RTC_RC_WARM_RESET;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
#endif /* AH_SUPPORT_AR9130 */
/*
* Reset AHB
* Force the MAC to wakeup before we try resetting it, to ensure it actually _gets_ reset properly. * Add some more comments describing why things are done. Obtained from: Atheros
2011-11-09 14:34:25 +00:00
*
* (In case the last interrupt source was a bus timeout.)
* XXX TODO: this is not the way to do it! It should be recorded
* XXX by the interrupt handler and passed _into_ the
* XXX reset path routine so this occurs.
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
*/
tmpReg = OS_REG_READ(ah, AR_INTR_SYNC_CAUSE);
if (tmpReg & (AR_INTR_SYNC_LOCAL_TIMEOUT|AR_INTR_SYNC_RADM_CPL_TIMEOUT)) {
OS_REG_WRITE(ah, AR_INTR_SYNC_ENABLE, 0);
OS_REG_WRITE(ah, AR_RC, AR_RC_AHB|AR_RC_HOSTIF);
} else {
OS_REG_WRITE(ah, AR_RC, AR_RC_AHB);
}
rst_flags = AR_RTC_RC_MAC_WARM;
if (type == HAL_RESET_COLD)
rst_flags |= AR_RTC_RC_MAC_COLD;
#ifdef AH_SUPPORT_AR9130
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
#endif /* AH_SUPPORT_AR9130 */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
OS_REG_WRITE(ah, AR_RTC_RC, rst_flags);
Import some HOWL (AR9130) related fixes from Atheros. Obtained from: Atheros
2011-05-05 02:59:31 +00:00
if (AR_SREV_HOWL(ah))
OS_DELAY(10000);
else
OS_DELAY(100);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Clear resets and force wakeup
*/
OS_REG_WRITE(ah, AR_RTC_RC, 0);
if (!ath_hal_wait(ah, AR_RTC_RC, AR_RTC_RC_M, 0)) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: RTC stuck in MAC reset\n", __func__);
return AH_FALSE;
}
/* Clear AHB reset */
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
if (! AR_SREV_HOWL(ah))
OS_REG_WRITE(ah, AR_RC, 0);
if (AR_SREV_HOWL(ah))
OS_DELAY(50);
if (AR_SREV_HOWL(ah)) {
uint32_t mask;
mask = OS_REG_READ(ah, AR_CFG);
if (mask & (AR_CFG_SWRB | AR_CFG_SWTB | AR_CFG_SWRG)) {
HALDEBUG(ah, HAL_DEBUG_RESET,
"CFG Byte Swap Set 0x%x\n", mask);
} else {
mask =
INIT_CONFIG_STATUS | AR_CFG_SWRB | AR_CFG_SWTB;
OS_REG_WRITE(ah, AR_CFG, mask);
HALDEBUG(ah, HAL_DEBUG_RESET,
"Setting CFG 0x%x\n", OS_REG_READ(ah, AR_CFG));
}
} else {
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
if (type == HAL_RESET_COLD) {
if (isBigEndian()) {
/*
Stop using the hardware register value byte order swapping for now, at least until I can root cause what's going on. The only platform I've seen this on is the AR9220 when attached to the AR71xx CPUs. I get immediate PCIe bus errors and all subsequent accesses cause further MIPS bus exceptions. I don't have any other big-endian platforms to test this on. If I get a chance (or two), I'll try to whack this on a bus analyser and see exactly what happens. I'd rather leave this on, especially for slower, embedded platforms. But the #ifdef hell is something I'm trying to avoid.
2012-04-19 03:26:21 +00:00
* Set CFG, little-endian for descriptor accesses.
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
*/
Stop using the hardware register value byte order swapping for now, at least until I can root cause what's going on. The only platform I've seen this on is the AR9220 when attached to the AR71xx CPUs. I get immediate PCIe bus errors and all subsequent accesses cause further MIPS bus exceptions. I don't have any other big-endian platforms to test this on. If I get a chance (or two), I'll try to whack this on a bus analyser and see exactly what happens. I'd rather leave this on, especially for slower, embedded platforms. But the #ifdef hell is something I'm trying to avoid.
2012-04-19 03:26:21 +00:00
mask = INIT_CONFIG_STATUS | AR_CFG_SWRD;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#ifndef AH_NEED_DESC_SWAP
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
mask |= AR_CFG_SWTD;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
HALDEBUG(ah, HAL_DEBUG_RESET,
"%s Applying descriptor swap\n", __func__);
Stop using the hardware register value byte order swapping for now, at least until I can root cause what's going on. The only platform I've seen this on is the AR9220 when attached to the AR71xx CPUs. I get immediate PCIe bus errors and all subsequent accesses cause further MIPS bus exceptions. I don't have any other big-endian platforms to test this on. If I get a chance (or two), I'll try to whack this on a bus analyser and see exactly what happens. I'd rather leave this on, especially for slower, embedded platforms. But the #ifdef hell is something I'm trying to avoid.
2012-04-19 03:26:21 +00:00
OS_REG_WRITE(ah, AR_CFG, mask);
Fix ar5416 and later parts on big-endian platforms: setup the h/w byte swizzler using the same technique used everywhere else. Approved by: re (kib)
2009-07-07 18:11:05 +00:00
} else
OS_REG_WRITE(ah, AR_CFG, INIT_CONFIG_STATUS);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
}
Break out the chainmask init code into a new function - ar5416InitChainMasks() . ath9k does a few different things here during config - if it's an early AR5416 with two chains, it enables all three chains for calibration and then restores the chainmask to the original values after initial calibration has completed. The reason behind this commit is to begin breaking out the chainmask configuration for this specific reason; follow-up commits will add the chainmask restore in the ar5416Reset() routine.
2011-01-26 10:08:37 +00:00
void
ar5416InitChainMasks(struct ath_hal *ah)
{
Fiddle with the AR5416 1.0 chainmask setup. Apparently all three RX chains need to be enabled before initial calibration is done, even if only two are configured. Reorder the alt chain swap bit to match what the Atheros HAL is doing. Obtained From: ath9k, Atheros
2011-05-08 03:24:17 +00:00
int rx_chainmask = AH5416(ah)->ah_rx_chainmask;
Fix a regression I introduced - only swap analog chains if the RX chainmask is 0x5.
2011-05-09 17:10:48 +00:00
/* Flip this for this chainmask regardless of chip */
if (rx_chainmask == 0x5)
Fiddle with the AR5416 1.0 chainmask setup. Apparently all three RX chains need to be enabled before initial calibration is done, even if only two are configured. Reorder the alt chain swap bit to match what the Atheros HAL is doing. Obtained From: ath9k, Atheros
2011-05-08 03:24:17 +00:00
OS_REG_SET_BIT(ah, AR_PHY_ANALOG_SWAP, AR_PHY_SWAP_ALT_CHAIN);
/*
* Workaround for OWL 1.0 calibration failure; enable multi-chain;
* then set true mask after calibration.
*/
if (IS_5416V1(ah) && (rx_chainmask == 0x5 || rx_chainmask == 0x3)) {
OS_REG_WRITE(ah, AR_PHY_RX_CHAINMASK, 0x7);
OS_REG_WRITE(ah, AR_PHY_CAL_CHAINMASK, 0x7);
} else {
OS_REG_WRITE(ah, AR_PHY_RX_CHAINMASK, AH5416(ah)->ah_rx_chainmask);
OS_REG_WRITE(ah, AR_PHY_CAL_CHAINMASK, AH5416(ah)->ah_rx_chainmask);
}
Break out the chainmask init code into a new function - ar5416InitChainMasks() . ath9k does a few different things here during config - if it's an early AR5416 with two chains, it enables all three chains for calibration and then restores the chainmask to the original values after initial calibration has completed. The reason behind this commit is to begin breaking out the chainmask configuration for this specific reason; follow-up commits will add the chainmask restore in the ar5416Reset() routine.
2011-01-26 10:08:37 +00:00
OS_REG_WRITE(ah, AR_SELFGEN_MASK, AH5416(ah)->ah_tx_chainmask);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
Fiddle with the AR5416 1.0 chainmask setup. Apparently all three RX chains need to be enabled before initial calibration is done, even if only two are configured. Reorder the alt chain swap bit to match what the Atheros HAL is doing. Obtained From: ath9k, Atheros
2011-05-08 03:24:17 +00:00
if (AH5416(ah)->ah_tx_chainmask == 0x5)
OS_REG_SET_BIT(ah, AR_PHY_ANALOG_SWAP, AR_PHY_SWAP_ALT_CHAIN);
Introduce AR9130 (HOWL) WMAC support to the FreeBSD HAL. The AR9130 is an AR9160/AR5416 family WMAC which is glued directly to the AR913x SoC peripheral bus (APB) rather than via a PCI/PCIe bridge. The specifics: * A new build option is required to use the AR9130 - AH_SUPPORT_AR9130. This is needed due to the different location the RTC registers live with this chip; hopefully this will be undone in the future. This does currently mean that enabling this option will break non-AR9130 builds, so don't enable it unless you're specifically building an image for the AR913x SoC. * Add the new probe, attach, EEPROM and PLL methods specific to Howl. * Add a work-around to ah_eeprom_v14.c which disables some of the checks for endian-ness and magic in the EEPROM image if an eepromdata block is provided. This'll be fixed at a later stage by porting the ath9k probe code and making sure it doesn't break in other setups (which my previous attempt at this did.) * Sprinkle Howl modifications throughput the interrupt path - it doesn't implement the SYNC interrupt registers, so ignore those. * Sprinkle Howl chip powerup/down throughout the reset path; the RTC methods were * Sprinkle some other Howl workarounds in the reset path. * Hard-code an alternative setup for the AR_CFG register for Howl, that sets up things suitable for Big-Endian MIPS (which is the only platform this chip is glued to.) This has been tested on the AR913x based TP-Link WR-1043nd mode, in legacy, HT/20 and HT/40 modes. Caveats: * 2ghz has only been tested. I've not seen any 5ghz radios glued to this chipset so I can't test it. * AR5416_INTERRUPT_MITIGATION is not supported on the AR9130. At least, it isn't implemented in ath9k. Please don't enable this. * This hasn't been tested in MBSS mode or in RX/TX block-aggregation mode.
2011-04-28 12:47:40 +00:00
if (AR_SREV_HOWL(ah)) {
OS_REG_WRITE(ah, AR_PHY_ANALOG_SWAP,
OS_REG_READ(ah, AR_PHY_ANALOG_SWAP) | 0x00000001);
}
Break out the chainmask init code into a new function - ar5416InitChainMasks() . ath9k does a few different things here during config - if it's an early AR5416 with two chains, it enables all three chains for calibration and then restores the chainmask to the original values after initial calibration has completed. The reason behind this commit is to begin breaking out the chainmask configuration for this specific reason; follow-up commits will add the chainmask restore in the ar5416Reset() routine.
2011-01-26 10:08:37 +00:00
}
Fiddle with the AR5416 1.0 chainmask setup. Apparently all three RX chains need to be enabled before initial calibration is done, even if only two are configured. Reorder the alt chain swap bit to match what the Atheros HAL is doing. Obtained From: ath9k, Atheros
2011-05-08 03:24:17 +00:00
/*
* Work-around for Owl 1.0 calibration failure.
*
* ar5416InitChainMasks sets the RX chainmask to 0x7 if it's Owl 1.0
* due to init calibration failures. ar5416RestoreChainMask restores
* these registers to the correct setting.
*/
Add ar5416RestoreChainMask() which will undo any AR5416 specific chainmask overriding after calibration. This will get set for other two chain radios, such as AR9280 and later on, AR9287. It should however be a nul operation.
2011-01-26 10:48:29 +00:00
void
ar5416RestoreChainMask(struct ath_hal *ah)
{
int rx_chainmask = AH5416(ah)->ah_rx_chainmask;
Fiddle with the AR5416 1.0 chainmask setup. Apparently all three RX chains need to be enabled before initial calibration is done, even if only two are configured. Reorder the alt chain swap bit to match what the Atheros HAL is doing. Obtained From: ath9k, Atheros
2011-05-08 03:24:17 +00:00
if (IS_5416V1(ah) && (rx_chainmask == 0x5 || rx_chainmask == 0x3)) {
Add ar5416RestoreChainMask() which will undo any AR5416 specific chainmask overriding after calibration. This will get set for other two chain radios, such as AR9280 and later on, AR9287. It should however be a nul operation.
2011-01-26 10:48:29 +00:00
OS_REG_WRITE(ah, AR_PHY_RX_CHAINMASK, rx_chainmask);
OS_REG_WRITE(ah, AR_PHY_CAL_CHAINMASK, rx_chainmask);
}
}
Break out the PLL setup into an overridable method. The only method right now is ar5416InitPLL() which handles multiple chipsets; this can now be overridden by newer chipset HAL code.
2011-04-24 15:53:57 +00:00
void
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416InitPLL(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
Break out the PLL setup into (mostly) per-chip methods, rather than polluting the AR5416 code with later chipset support. Note: ar9280InitPLL() supports Merlin (AR9280) and later (AR9285, AR9287.) Submitted by: ssgriffonuser@gmail.com Approved by: re (kib)
2011-07-21 08:35:10 +00:00
uint32_t pll = AR_RTC_PLL_REFDIV_5 | AR_RTC_PLL_DIV2;
if (chan != AH_NULL) {
if (IEEE80211_IS_CHAN_HALF(chan))
pll |= SM(0x1, AR_RTC_PLL_CLKSEL);
else if (IEEE80211_IS_CHAN_QUARTER(chan))
pll |= SM(0x2, AR_RTC_PLL_CLKSEL);
if (IEEE80211_IS_CHAN_5GHZ(chan))
pll |= SM(0xa, AR_RTC_PLL_DIV);
else
virgin import of ath hal
2008-11-28 00:03:41 +00:00
pll |= SM(0xb, AR_RTC_PLL_DIV);
Break out the PLL setup into (mostly) per-chip methods, rather than polluting the AR5416 code with later chipset support. Note: ar9280InitPLL() supports Merlin (AR9280) and later (AR9285, AR9287.) Submitted by: ssgriffonuser@gmail.com Approved by: re (kib)
2011-07-21 08:35:10 +00:00
} else
pll |= SM(0xb, AR_RTC_PLL_DIV);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_RTC_PLL_CONTROL, pll);
/* TODO:
* For multi-band owl, switch between bands by reiniting the PLL.
*/
OS_DELAY(RTC_PLL_SETTLE_DELAY);
OS_REG_WRITE(ah, AR_RTC_SLEEP_CLK, AR_RTC_SLEEP_DERIVED_CLK);
}
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
static void
ar5416SetDefGainValues(struct ath_hal *ah,
const MODAL_EEP_HEADER *pModal,
const struct ar5416eeprom *eep,
uint8_t txRxAttenLocal, int regChainOffset, int i)
{
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
if (IS_EEP_MINOR_V3(ah)) {
txRxAttenLocal = pModal->txRxAttenCh[i];
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah)) {
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_XATTEN1_MARGIN,
pModal->bswMargin[i]);
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_XATTEN1_DB,
pModal->bswAtten[i]);
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_XATTEN2_MARGIN,
pModal->xatten2Margin[i]);
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_XATTEN2_DB,
pModal->xatten2Db[i]);
} else {
Do a bit of spring cleaning in the board setup code, just to bring it in line with the rest of the register initialisation. I've verified that the 2/5ghz board values written to the chip match what was previously written.
2011-03-22 00:43:58 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_BSW_MARGIN,
pModal->bswMargin[i]);
OS_REG_RMW_FIELD(ah, AR_PHY_GAIN_2GHZ + regChainOffset,
AR_PHY_GAIN_2GHZ_BSW_ATTEN,
pModal->bswAtten[i]);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
}
}
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah)) {
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
OS_REG_RMW_FIELD(ah,
AR_PHY_RXGAIN + regChainOffset,
AR9280_PHY_RXGAIN_TXRX_ATTEN, txRxAttenLocal);
OS_REG_RMW_FIELD(ah,
AR_PHY_RXGAIN + regChainOffset,
AR9280_PHY_RXGAIN_TXRX_MARGIN, pModal->rxTxMarginCh[i]);
} else {
Do a bit of spring cleaning in the board setup code, just to bring it in line with the rest of the register initialisation. I've verified that the 2/5ghz board values written to the chip match what was previously written.
2011-03-22 00:43:58 +00:00
OS_REG_RMW_FIELD(ah,
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
AR_PHY_RXGAIN + regChainOffset,
Do a bit of spring cleaning in the board setup code, just to bring it in line with the rest of the register initialisation. I've verified that the 2/5ghz board values written to the chip match what was previously written.
2011-03-22 00:43:58 +00:00
AR_PHY_RXGAIN_TXRX_ATTEN, txRxAttenLocal);
OS_REG_RMW_FIELD(ah,
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
AR_PHY_GAIN_2GHZ + regChainOffset,
Do a bit of spring cleaning in the board setup code, just to bring it in line with the rest of the register initialisation. I've verified that the 2/5ghz board values written to the chip match what was previously written.
2011-03-22 00:43:58 +00:00
AR_PHY_GAIN_2GHZ_RXTX_MARGIN, pModal->rxTxMarginCh[i]);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
}
}
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/*
* Get the register chain offset for the given chain.
*
* Take into account the register chain swapping with AR5416 v2.0.
*
* XXX make sure that the reg chain swapping is only done for
* XXX AR5416 v2.0 or greater, and not later chips?
*/
int
ar5416GetRegChainOffset(struct ath_hal *ah, int i)
{
int regChainOffset;
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
Change AR_SREV_OWL_{X}_OR_LATER to AR_SREV_5416_{X}_OR_LATER. For now, these are equivalent macros. AR_SREV_OWL{X}_OR_LATER will later change to exclude Howl (AR9130) in line with what the Atheros HAL does. This should not functionally change anything. Obtained from: Atheros
2011-05-07 02:59:24 +00:00
if (AR_SREV_5416_V20_OR_LATER(ah) &&
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
(AH5416(ah)->ah_rx_chainmask == 0x5 ||
AH5416(ah)->ah_tx_chainmask == 0x5) && (i != 0)) {
/* Regs are swapped from chain 2 to 1 for 5416 2_0 with
* only chains 0 and 2 populated
*/
regChainOffset = (i == 1) ? 0x2000 : 0x1000;
} else {
regChainOffset = i * 0x1000;
}
return regChainOffset;
}
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Read EEPROM header info and program the device for correct operation
* given the channel value.
*/
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416SetBoardValues(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
const HAL_EEPROM_v14 *ee = AH_PRIVATE(ah)->ah_eeprom;
const struct ar5416eeprom *eep = &ee->ee_base;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
const MODAL_EEP_HEADER *pModal;
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
int i, regChainOffset;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint8_t txRxAttenLocal; /* workaround for eeprom versions <= 14.2 */
HALASSERT(AH_PRIVATE(ah)->ah_eeversion >= AR_EEPROM_VER14_1);
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal = &eep->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)];
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
/* NB: workaround for eeprom versions <= 14.2 */
txRxAttenLocal = IEEE80211_IS_CHAN_2GHZ(chan) ? 23 : 44;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
OS_REG_WRITE(ah, AR_PHY_SWITCH_COM, pModal->antCtrlCommon);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
for (i = 0; i < AR5416_MAX_CHAINS; i++) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
if (AR_SREV_MERLIN(ah)) {
if (i >= 2) break;
}
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
regChainOffset = ar5416GetRegChainOffset(ah, i);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
OS_REG_WRITE(ah, AR_PHY_SWITCH_CHAIN_0 + regChainOffset, pModal->antCtrlChain[i]);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_PHY_TIMING_CTRL4 + regChainOffset,
(OS_REG_READ(ah, AR_PHY_TIMING_CTRL4 + regChainOffset) &
~(AR_PHY_TIMING_CTRL4_IQCORR_Q_Q_COFF | AR_PHY_TIMING_CTRL4_IQCORR_Q_I_COFF)) |
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
SM(pModal->iqCalICh[i], AR_PHY_TIMING_CTRL4_IQCORR_Q_I_COFF) |
SM(pModal->iqCalQCh[i], AR_PHY_TIMING_CTRL4_IQCORR_Q_Q_COFF));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
* Large signal upgrade,
* If 14.3 or later EEPROM, use
* txRxAttenLocal = pModal->txRxAttenCh[i]
* else txRxAttenLocal is fixed value above.
virgin import of ath hal
2008-11-28 00:03:41 +00:00
*/
Change AR_SREV_OWL_{X}_OR_LATER to AR_SREV_5416_{X}_OR_LATER. For now, these are equivalent macros. AR_SREV_OWL{X}_OR_LATER will later change to exclude Howl (AR9130) in line with what the Atheros HAL does. This should not functionally change anything. Obtained from: Atheros
2011-05-07 02:59:24 +00:00
if ((i == 0) || AR_SREV_5416_V20_OR_LATER(ah))
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
ar5416SetDefGainValues(ah, pModal, eep, txRxAttenLocal, regChainOffset, i);
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah)) {
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
OS_A_REG_RMW_FIELD(ah, AR_AN_RF2G1_CH0, AR_AN_RF2G1_CH0_OB, pModal->ob);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF2G1_CH0, AR_AN_RF2G1_CH0_DB, pModal->db);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF2G1_CH1, AR_AN_RF2G1_CH1_OB, pModal->ob_ch1);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF2G1_CH1, AR_AN_RF2G1_CH1_DB, pModal->db_ch1);
} else {
OS_A_REG_RMW_FIELD(ah, AR_AN_RF5G1_CH0, AR_AN_RF5G1_CH0_OB5, pModal->ob);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF5G1_CH0, AR_AN_RF5G1_CH0_DB5, pModal->db);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF5G1_CH1, AR_AN_RF5G1_CH1_OB5, pModal->ob_ch1);
OS_A_REG_RMW_FIELD(ah, AR_AN_RF5G1_CH1, AR_AN_RF5G1_CH1_DB5, pModal->db_ch1);
}
OS_A_REG_RMW_FIELD(ah, AR_AN_TOP2, AR_AN_TOP2_XPABIAS_LVL, pModal->xpaBiasLvl);
Fix the merlin LNA configuration code - these are bit flags, not raw values to be written into the registers.
2011-04-22 17:57:13 +00:00
OS_A_REG_RMW_FIELD(ah, AR_AN_TOP2, AR_AN_TOP2_LOCALBIAS,
!!(pModal->flagBits & AR5416_EEP_FLAG_LOCALBIAS));
OS_A_REG_RMW_FIELD(ah, AR_PHY_XPA_CFG, AR_PHY_FORCE_XPA_CFG,
!!(pModal->flagBits & AR5416_EEP_FLAG_FORCEXPAON));
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
}
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_SETTLING, AR_PHY_SETTLING_SWITCH, pModal->switchSettling);
OS_REG_RMW_FIELD(ah, AR_PHY_DESIRED_SZ, AR_PHY_DESIRED_SZ_ADC, pModal->adcDesiredSize);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
if (! AR_SREV_MERLIN_10_OR_LATER(ah))
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_DESIRED_SZ, AR_PHY_DESIRED_SZ_PGA, pModal->pgaDesiredSize);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_WRITE(ah, AR_PHY_RF_CTL4,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
SM(pModal->txEndToXpaOff, AR_PHY_RF_CTL4_TX_END_XPAA_OFF)
| SM(pModal->txEndToXpaOff, AR_PHY_RF_CTL4_TX_END_XPAB_OFF)
| SM(pModal->txFrameToXpaOn, AR_PHY_RF_CTL4_FRAME_XPAA_ON)
| SM(pModal->txFrameToXpaOn, AR_PHY_RF_CTL4_FRAME_XPAB_ON));
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_RF_CTL3, AR_PHY_TX_END_TO_A2_RX_ON,
pModal->txEndToRxOn);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah)) {
OS_REG_RMW_FIELD(ah, AR_PHY_CCA, AR9280_PHY_CCA_THRESH62,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal->thresh62);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_EXT_CCA0, AR_PHY_EXT_CCA0_THRESH62,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal->thresh62);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
OS_REG_RMW_FIELD(ah, AR_PHY_CCA, AR_PHY_CCA_THRESH62,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal->thresh62);
Set the "right" CCA register. Obtained From: ath9k
2011-03-22 05:47:48 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_EXT_CCA, AR_PHY_EXT_CCA_THRESH62,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pModal->thresh62);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/* Minor Version Specific application */
if (IS_EEP_MINOR_V2(ah)) {
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_RF_CTL2, AR_PHY_TX_FRAME_TO_DATA_START,
pModal->txFrameToDataStart);
OS_REG_RMW_FIELD(ah, AR_PHY_RF_CTL2, AR_PHY_TX_FRAME_TO_PA_ON,
pModal->txFrameToPaOn);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
if (IS_EEP_MINOR_V3(ah) && IEEE80211_IS_CHAN_HT40(chan))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Overwrite switch settling with HT40 value */
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_REG_RMW_FIELD(ah, AR_PHY_SETTLING, AR_PHY_SETTLING_SWITCH,
pModal->swSettleHt40);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
if (AR_SREV_MERLIN_20_OR_LATER(ah) && EEP_MINOR(ah) >= AR5416_EEP_MINOR_VER_19)
OS_REG_RMW_FIELD(ah, AR_PHY_CCK_TX_CTRL, AR_PHY_CCK_TX_CTRL_TX_DAC_SCALE_CCK, pModal->miscBits);
if (AR_SREV_MERLIN_20(ah) && EEP_MINOR(ah) >= AR5416_EEP_MINOR_VER_20) {
if (IEEE80211_IS_CHAN_2GHZ(chan))
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_A_REG_RMW_FIELD(ah, AR_AN_TOP1, AR_AN_TOP1_DACIPMODE,
eep->baseEepHeader.dacLpMode);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
else if (eep->baseEepHeader.dacHiPwrMode_5G)
OS_A_REG_RMW_FIELD(ah, AR_AN_TOP1, AR_AN_TOP1_DACIPMODE, 0);
else
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_A_REG_RMW_FIELD(ah, AR_AN_TOP1, AR_AN_TOP1_DACIPMODE,
eep->baseEepHeader.dacLpMode);
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
Fixes from the Atheros HAL - formatting; update Merlin checks to be consistent. Nothing functional should change with this commit.
2011-05-14 04:05:23 +00:00
OS_DELAY(100);
OS_REG_RMW_FIELD(ah, AR_PHY_FRAME_CTL, AR_PHY_FRAME_CTL_TX_CLIP,
pModal->miscBits >> 2);
OS_REG_RMW_FIELD(ah, AR_PHY_TX_PWRCTRL9, AR_PHY_TX_DESIRED_SCALE_CCK,
eep->baseEepHeader.desiredScaleCCK);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
Add in some AR9280 specific board configuration options. * The existing radio config code was for the AR5416/AR9160 and missed out on some of the AR9280 specific stuff. Include said stuff from ath9k. * Refactor out the gain control settings into a new function, again pilfered from ath9k. * Use the analog register RMW macro when touching analog registers. Obtained from: Linux ath9k
2011-02-07 22:00:31 +00:00
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
return (AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/*
* Helper functions common for AP/CB/XB
*/
Remove duplicate code - add a function which calculates the ratesArray[] table which contains the per-rate target TX power. This code is shared between the v14 eeprom board setup (AR5416, AR9160, AR9280) and will also be used by the upcoming Kite (AR9287) support.
2011-05-13 10:36:38 +00:00
/*
* Set the target power array "ratesArray" from the
* given set of target powers.
*
* This is used by the various chipset/EEPROM TX power
* setup routines.
*/
void
ar5416SetRatesArrayFromTargetPower(struct ath_hal *ah,
const struct ieee80211_channel *chan,
int16_t *ratesArray,
const CAL_TARGET_POWER_LEG *targetPowerCck,
const CAL_TARGET_POWER_LEG *targetPowerCckExt,
const CAL_TARGET_POWER_LEG *targetPowerOfdm,
const CAL_TARGET_POWER_LEG *targetPowerOfdmExt,
const CAL_TARGET_POWER_HT *targetPowerHt20,
const CAL_TARGET_POWER_HT *targetPowerHt40)
{
#define N(a) (sizeof(a)/sizeof(a[0]))
int i;
/* Blank the rates array, to be consistent */
for (i = 0; i < Ar5416RateSize; i++)
ratesArray[i] = 0;
/* Set rates Array from collected data */
ratesArray[rate6mb] = ratesArray[rate9mb] = ratesArray[rate12mb] =
Sprinkle some style(9) things around.
2012-03-22 21:47:14 +00:00
ratesArray[rate18mb] = ratesArray[rate24mb] =
targetPowerOfdm->tPow2x[0];
Remove duplicate code - add a function which calculates the ratesArray[] table which contains the per-rate target TX power. This code is shared between the v14 eeprom board setup (AR5416, AR9160, AR9280) and will also be used by the upcoming Kite (AR9287) support.
2011-05-13 10:36:38 +00:00
ratesArray[rate36mb] = targetPowerOfdm->tPow2x[1];
ratesArray[rate48mb] = targetPowerOfdm->tPow2x[2];
ratesArray[rate54mb] = targetPowerOfdm->tPow2x[3];
ratesArray[rateXr] = targetPowerOfdm->tPow2x[0];
for (i = 0; i < N(targetPowerHt20->tPow2x); i++) {
ratesArray[rateHt20_0 + i] = targetPowerHt20->tPow2x[i];
}
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
ratesArray[rate1l] = targetPowerCck->tPow2x[0];
ratesArray[rate2s] = ratesArray[rate2l] = targetPowerCck->tPow2x[1];
ratesArray[rate5_5s] = ratesArray[rate5_5l] = targetPowerCck->tPow2x[2];
ratesArray[rate11s] = ratesArray[rate11l] = targetPowerCck->tPow2x[3];
}
if (IEEE80211_IS_CHAN_HT40(chan)) {
for (i = 0; i < N(targetPowerHt40->tPow2x); i++) {
ratesArray[rateHt40_0 + i] = targetPowerHt40->tPow2x[i];
}
ratesArray[rateDupOfdm] = targetPowerHt40->tPow2x[0];
ratesArray[rateDupCck] = targetPowerHt40->tPow2x[0];
ratesArray[rateExtOfdm] = targetPowerOfdmExt->tPow2x[0];
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
ratesArray[rateExtCck] = targetPowerCckExt->tPow2x[0];
}
}
#undef N
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* ar5416SetPowerPerRateTable
*
* Sets the transmit power in the baseband for the given
* operating channel and mode.
*/
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
static HAL_BOOL
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416SetPowerPerRateTable(struct ath_hal *ah, struct ar5416eeprom *pEepData,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
const struct ieee80211_channel *chan,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
int16_t *ratesArray, uint16_t cfgCtl,
uint16_t AntennaReduction,
uint16_t twiceMaxRegulatoryPower,
uint16_t powerLimit)
{
#define N(a) (sizeof(a)/sizeof(a[0]))
/* Local defines to distinguish between extension and control CTL's */
#define EXT_ADDITIVE (0x8000)
#define CTL_11A_EXT (CTL_11A | EXT_ADDITIVE)
#define CTL_11G_EXT (CTL_11G | EXT_ADDITIVE)
#define CTL_11B_EXT (CTL_11B | EXT_ADDITIVE)
uint16_t twiceMaxEdgePower = AR5416_MAX_RATE_POWER;
int i;
int16_t twiceLargestAntenna;
CAL_CTL_DATA *rep;
CAL_TARGET_POWER_LEG targetPowerOfdm, targetPowerCck = {0, {0, 0, 0, 0}};
CAL_TARGET_POWER_LEG targetPowerOfdmExt = {0, {0, 0, 0, 0}}, targetPowerCckExt = {0, {0, 0, 0, 0}};
CAL_TARGET_POWER_HT targetPowerHt20, targetPowerHt40 = {0, {0, 0, 0, 0}};
int16_t scaledPower, minCtlPower;
#define SUB_NUM_CTL_MODES_AT_5G_40 2 /* excluding HT40, EXT-OFDM */
#define SUB_NUM_CTL_MODES_AT_2G_40 3 /* excluding HT40, EXT-OFDM, EXT-CCK */
static const uint16_t ctlModesFor11a[] = {
CTL_11A, CTL_5GHT20, CTL_11A_EXT, CTL_5GHT40
};
static const uint16_t ctlModesFor11g[] = {
CTL_11B, CTL_11G, CTL_2GHT20, CTL_11B_EXT, CTL_11G_EXT, CTL_2GHT40
};
const uint16_t *pCtlMode;
uint16_t numCtlModes, ctlMode, freq;
CHAN_CENTERS centers;
ar5416GetChannelCenters(ah, chan, &centers);
/* Compute TxPower reduction due to Antenna Gain */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
twiceLargestAntenna = AH_MAX(AH_MAX(
pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[0],
pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[1]),
pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[2]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#if 0
/* Turn it back on if we need to calculate per chain antenna gain reduction */
/* Use only if the expected gain > 6dbi */
/* Chain 0 is always used */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
twiceLargestAntenna = pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[0];
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Look at antenna gains of Chains 1 and 2 if the TX mask is set */
if (ahp->ah_tx_chainmask & 0x2)
twiceLargestAntenna = AH_MAX(twiceLargestAntenna,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[1]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (ahp->ah_tx_chainmask & 0x4)
twiceLargestAntenna = AH_MAX(twiceLargestAntenna,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].antennaGainCh[2]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
#endif
twiceLargestAntenna = (int16_t)AH_MIN((AntennaReduction) - twiceLargestAntenna, 0);
/* XXX setup for 5212 use (really used?) */
ath_hal_eepromSet(ah,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
IEEE80211_IS_CHAN_2GHZ(chan) ? AR_EEP_ANTGAINMAX_2 : AR_EEP_ANTGAINMAX_5,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
twiceLargestAntenna);
/*
* scaledPower is the minimum of the user input power level and
* the regulatory allowed power level
*/
scaledPower = AH_MIN(powerLimit, twiceMaxRegulatoryPower + twiceLargestAntenna);
/* Reduce scaled Power by number of chains active to get to per chain tx power level */
/* TODO: better value than these? */
switch (owl_get_ntxchains(AH5416(ah)->ah_tx_chainmask)) {
case 1:
break;
case 2:
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
scaledPower -= pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].pwrDecreaseFor2Chain;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
break;
case 3:
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
scaledPower -= pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].pwrDecreaseFor3Chain;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
break;
default:
return AH_FALSE; /* Unsupported number of chains */
}
scaledPower = AH_MAX(0, scaledPower);
/* Get target powers from EEPROM - our baseline for TX Power */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Setup for CTL modes */
numCtlModes = N(ctlModesFor11g) - SUB_NUM_CTL_MODES_AT_2G_40; /* CTL_11B, CTL_11G, CTL_2GHT20 */
pCtlMode = ctlModesFor11g;
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPowerCck,
AR5416_NUM_2G_CCK_TARGET_POWERS, &targetPowerCck, 4, AH_FALSE);
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPower2G,
AR5416_NUM_2G_20_TARGET_POWERS, &targetPowerOfdm, 4, AH_FALSE);
ar5416GetTargetPowers(ah, chan, pEepData->calTargetPower2GHT20,
AR5416_NUM_2G_20_TARGET_POWERS, &targetPowerHt20, 8, AH_FALSE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
numCtlModes = N(ctlModesFor11g); /* All 2G CTL's */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
ar5416GetTargetPowers(ah, chan, pEepData->calTargetPower2GHT40,
AR5416_NUM_2G_40_TARGET_POWERS, &targetPowerHt40, 8, AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Get target powers for extension channels */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPowerCck,
AR5416_NUM_2G_CCK_TARGET_POWERS, &targetPowerCckExt, 4, AH_TRUE);
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPower2G,
AR5416_NUM_2G_20_TARGET_POWERS, &targetPowerOfdmExt, 4, AH_TRUE);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
} else {
/* Setup for CTL modes */
numCtlModes = N(ctlModesFor11a) - SUB_NUM_CTL_MODES_AT_5G_40; /* CTL_11A, CTL_5GHT20 */
pCtlMode = ctlModesFor11a;
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPower5G,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
AR5416_NUM_5G_20_TARGET_POWERS, &targetPowerOfdm, 4, AH_FALSE);
ar5416GetTargetPowers(ah, chan, pEepData->calTargetPower5GHT20,
AR5416_NUM_5G_20_TARGET_POWERS, &targetPowerHt20, 8, AH_FALSE);
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
numCtlModes = N(ctlModesFor11a); /* All 5G CTL's */
ar5416GetTargetPowers(ah, chan, pEepData->calTargetPower5GHT40,
AR5416_NUM_5G_40_TARGET_POWERS, &targetPowerHt40, 8, AH_TRUE);
ar5416GetTargetPowersLeg(ah, chan, pEepData->calTargetPower5G,
AR5416_NUM_5G_20_TARGET_POWERS, &targetPowerOfdmExt, 4, AH_TRUE);
}
}
/*
* For MIMO, need to apply regulatory caps individually across dynamically
* running modes: CCK, OFDM, HT20, HT40
*
* The outer loop walks through each possible applicable runtime mode.
* The inner loop walks through each ctlIndex entry in EEPROM.
* The ctl value is encoded as [7:4] == test group, [3:0] == test mode.
*
*/
for (ctlMode = 0; ctlMode < numCtlModes; ctlMode++) {
HAL_BOOL isHt40CtlMode = (pCtlMode[ctlMode] == CTL_5GHT40) ||
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
(pCtlMode[ctlMode] == CTL_2GHT40);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (isHt40CtlMode) {
freq = centers.ctl_center;
} else if (pCtlMode[ctlMode] & EXT_ADDITIVE) {
freq = centers.ext_center;
} else {
freq = centers.ctl_center;
}
/* walk through each CTL index stored in EEPROM */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
for (i = 0; (i < AR5416_NUM_CTLS) && pEepData->ctlIndex[i]; i++) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint16_t twiceMinEdgePower;
/* compare test group from regulatory channel list with test mode from pCtlMode list */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
if ((((cfgCtl & ~CTL_MODE_M) | (pCtlMode[ctlMode] & CTL_MODE_M)) == pEepData->ctlIndex[i]) ||
virgin import of ath hal
2008-11-28 00:03:41 +00:00
(((cfgCtl & ~CTL_MODE_M) | (pCtlMode[ctlMode] & CTL_MODE_M)) ==
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
((pEepData->ctlIndex[i] & CTL_MODE_M) | SD_NO_CTL))) {
rep = &(pEepData->ctlData[i]);
twiceMinEdgePower = ar5416GetMaxEdgePower(freq,
rep->ctlEdges[owl_get_ntxchains(AH5416(ah)->ah_tx_chainmask) - 1],
IEEE80211_IS_CHAN_2GHZ(chan));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if ((cfgCtl & ~CTL_MODE_M) == SD_NO_CTL) {
/* Find the minimum of all CTL edge powers that apply to this channel */
twiceMaxEdgePower = AH_MIN(twiceMaxEdgePower, twiceMinEdgePower);
} else {
/* specific */
twiceMaxEdgePower = twiceMinEdgePower;
break;
}
}
}
minCtlPower = (uint8_t)AH_MIN(twiceMaxEdgePower, scaledPower);
/* Apply ctl mode to correct target power set */
switch(pCtlMode[ctlMode]) {
case CTL_11B:
for (i = 0; i < N(targetPowerCck.tPow2x); i++) {
targetPowerCck.tPow2x[i] = (uint8_t)AH_MIN(targetPowerCck.tPow2x[i], minCtlPower);
}
break;
case CTL_11A:
case CTL_11G:
for (i = 0; i < N(targetPowerOfdm.tPow2x); i++) {
targetPowerOfdm.tPow2x[i] = (uint8_t)AH_MIN(targetPowerOfdm.tPow2x[i], minCtlPower);
}
break;
case CTL_5GHT20:
case CTL_2GHT20:
for (i = 0; i < N(targetPowerHt20.tPow2x); i++) {
targetPowerHt20.tPow2x[i] = (uint8_t)AH_MIN(targetPowerHt20.tPow2x[i], minCtlPower);
}
break;
case CTL_11B_EXT:
targetPowerCckExt.tPow2x[0] = (uint8_t)AH_MIN(targetPowerCckExt.tPow2x[0], minCtlPower);
break;
case CTL_11A_EXT:
case CTL_11G_EXT:
targetPowerOfdmExt.tPow2x[0] = (uint8_t)AH_MIN(targetPowerOfdmExt.tPow2x[0], minCtlPower);
break;
case CTL_5GHT40:
case CTL_2GHT40:
for (i = 0; i < N(targetPowerHt40.tPow2x); i++) {
targetPowerHt40.tPow2x[i] = (uint8_t)AH_MIN(targetPowerHt40.tPow2x[i], minCtlPower);
}
break;
default:
return AH_FALSE;
break;
}
} /* end ctl mode checking */
/* Set rates Array from collected data */
Remove duplicate code - add a function which calculates the ratesArray[] table which contains the per-rate target TX power. This code is shared between the v14 eeprom board setup (AR5416, AR9160, AR9280) and will also be used by the upcoming Kite (AR9287) support.
2011-05-13 10:36:38 +00:00
ar5416SetRatesArrayFromTargetPower(ah, chan, ratesArray,
&targetPowerCck,
&targetPowerCckExt,
&targetPowerOfdm,
&targetPowerOfdmExt,
&targetPowerHt20,
&targetPowerHt40);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
return AH_TRUE;
#undef EXT_ADDITIVE
#undef CTL_11A_EXT
#undef CTL_11G_EXT
#undef CTL_11B_EXT
#undef SUB_NUM_CTL_MODES_AT_5G_40
#undef SUB_NUM_CTL_MODES_AT_2G_40
#undef N
}
/**************************************************************************
* fbin2freq
*
* Get channel value from binary representation held in eeprom
* RETURNS: the frequency in MHz
*/
static uint16_t
fbin2freq(uint8_t fbin, HAL_BOOL is2GHz)
{
/*
* Reserved value 0xFF provides an empty definition both as
* an fbin and as a frequency - do not convert
*/
if (fbin == AR5416_BCHAN_UNUSED) {
return fbin;
}
return (uint16_t)((is2GHz) ? (2300 + fbin) : (4800 + 5 * fbin));
}
/*
* ar5416GetMaxEdgePower
*
* Find the maximum conformance test limit for the given channel and CTL info
*/
Remove some duplicate code from the AR9285 TX power configuration path.
2011-04-16 11:59:37 +00:00
uint16_t
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
ar5416GetMaxEdgePower(uint16_t freq, CAL_CTL_EDGES *pRdEdgesPower, HAL_BOOL is2GHz)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
uint16_t twiceMaxEdgePower = AR5416_MAX_RATE_POWER;
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
int i;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Get the edge power */
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
for (i = 0; (i < AR5416_NUM_BAND_EDGES) && (pRdEdgesPower[i].bChannel != AR5416_BCHAN_UNUSED) ; i++) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* If there's an exact channel match or an inband flag set
* on the lower channel use the given rdEdgePower
*/
if (freq == fbin2freq(pRdEdgesPower[i].bChannel, is2GHz)) {
twiceMaxEdgePower = MS(pRdEdgesPower[i].tPowerFlag, CAL_CTL_EDGES_POWER);
break;
} else if ((i > 0) && (freq < fbin2freq(pRdEdgesPower[i].bChannel, is2GHz))) {
if (fbin2freq(pRdEdgesPower[i - 1].bChannel, is2GHz) < freq && (pRdEdgesPower[i - 1].tPowerFlag & CAL_CTL_EDGES_FLAG) != 0) {
twiceMaxEdgePower = MS(pRdEdgesPower[i - 1].tPowerFlag, CAL_CTL_EDGES_POWER);
}
/* Leave loop - no more affecting edges possible in this monotonic increasing list */
break;
}
}
HALASSERT(twiceMaxEdgePower > 0);
return twiceMaxEdgePower;
}
/**************************************************************
* ar5416GetTargetPowers
*
* Return the rates of target power for the given target power table
* channel, and number of channels
*/
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
void
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416GetTargetPowers(struct ath_hal *ah, const struct ieee80211_channel *chan,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
CAL_TARGET_POWER_HT *powInfo, uint16_t numChannels,
CAL_TARGET_POWER_HT *pNewPower, uint16_t numRates,
HAL_BOOL isHt40Target)
{
uint16_t clo, chi;
int i;
int matchIndex = -1, lowIndex = -1;
uint16_t freq;
CHAN_CENTERS centers;
ar5416GetChannelCenters(ah, chan, &centers);
freq = isHt40Target ? centers.synth_center : centers.ctl_center;
/* Copy the target powers into the temp channel list */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (freq <= fbin2freq(powInfo[0].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = 0;
} else {
for (i = 0; (i < numChannels) && (powInfo[i].bChannel != AR5416_BCHAN_UNUSED); i++) {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (freq == fbin2freq(powInfo[i].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = i;
break;
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
} else if ((freq < fbin2freq(powInfo[i].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) &&
(freq > fbin2freq(powInfo[i - 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
lowIndex = i - 1;
break;
}
}
if ((matchIndex == -1) && (lowIndex == -1)) {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
HALASSERT(freq > fbin2freq(powInfo[i - 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan)));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = i - 1;
}
}
if (matchIndex != -1) {
OS_MEMCPY(pNewPower, &powInfo[matchIndex], sizeof(*pNewPower));
} else {
HALASSERT(lowIndex != -1);
/*
* Get the lower and upper channels, target powers,
* and interpolate between them.
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
clo = fbin2freq(powInfo[lowIndex].bChannel, IEEE80211_IS_CHAN_2GHZ(chan));
chi = fbin2freq(powInfo[lowIndex + 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
for (i = 0; i < numRates; i++) {
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
pNewPower->tPow2x[i] = (uint8_t)ath_ee_interpolate(freq, clo, chi,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
powInfo[lowIndex].tPow2x[i], powInfo[lowIndex + 1].tPow2x[i]);
}
}
}
/**************************************************************
* ar5416GetTargetPowersLeg
*
* Return the four rates of target power for the given target power table
* channel, and number of channels
*/
Bring back AR9285 support. This fixes most of the issues and should be pretty usable. MFC after: 1 month
2010-02-15 17:49:49 +00:00
void
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416GetTargetPowersLeg(struct ath_hal *ah,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
const struct ieee80211_channel *chan,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
CAL_TARGET_POWER_LEG *powInfo, uint16_t numChannels,
CAL_TARGET_POWER_LEG *pNewPower, uint16_t numRates,
HAL_BOOL isExtTarget)
{
uint16_t clo, chi;
int i;
int matchIndex = -1, lowIndex = -1;
uint16_t freq;
CHAN_CENTERS centers;
ar5416GetChannelCenters(ah, chan, &centers);
freq = (isExtTarget) ? centers.ext_center :centers.ctl_center;
/* Copy the target powers into the temp channel list */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (freq <= fbin2freq(powInfo[0].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = 0;
} else {
for (i = 0; (i < numChannels) && (powInfo[i].bChannel != AR5416_BCHAN_UNUSED); i++) {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (freq == fbin2freq(powInfo[i].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = i;
break;
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
} else if ((freq < fbin2freq(powInfo[i].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))) &&
(freq > fbin2freq(powInfo[i - 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan))))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
lowIndex = i - 1;
break;
}
}
if ((matchIndex == -1) && (lowIndex == -1)) {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
HALASSERT(freq > fbin2freq(powInfo[i - 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan)));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
matchIndex = i - 1;
}
}
if (matchIndex != -1) {
OS_MEMCPY(pNewPower, &powInfo[matchIndex], sizeof(*pNewPower));
} else {
HALASSERT(lowIndex != -1);
/*
* Get the lower and upper channels, target powers,
* and interpolate between them.
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
clo = fbin2freq(powInfo[lowIndex].bChannel, IEEE80211_IS_CHAN_2GHZ(chan));
chi = fbin2freq(powInfo[lowIndex + 1].bChannel, IEEE80211_IS_CHAN_2GHZ(chan));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
for (i = 0; i < numRates; i++) {
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
pNewPower->tPow2x[i] = (uint8_t)ath_ee_interpolate(freq, clo, chi,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
powInfo[lowIndex].tPow2x[i], powInfo[lowIndex + 1].tPow2x[i]);
}
}
}
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/*
* Set the gain boundaries for the given radio chain.
*
* The gain boundaries tell the hardware at what point in the
* PDADC array to "switch over" from one PD gain setting
* to another. There's also a gain overlap between two
* PDADC array gain curves where there's valid PD values
* for 2 gain settings.
*
* The hardware uses the gain overlap and gain boundaries
* to determine which gain curve to use for the given
* target TX power.
*/
void
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
ar5416SetGainBoundariesClosedLoop(struct ath_hal *ah, int i,
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
uint16_t pdGainOverlap_t2, uint16_t gainBoundaries[])
{
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
int regChainOffset;
regChainOffset = ar5416GetRegChainOffset(ah, i);
HALDEBUG(ah, HAL_DEBUG_EEPROM, "%s: chain %d: gainOverlap_t2: %d,"
" gainBoundaries: %d, %d, %d, %d\n", __func__, i, pdGainOverlap_t2,
gainBoundaries[0], gainBoundaries[1], gainBoundaries[2],
gainBoundaries[3]);
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
OS_REG_WRITE(ah, AR_PHY_TPCRG5 + regChainOffset,
SM(pdGainOverlap_t2, AR_PHY_TPCRG5_PD_GAIN_OVERLAP) |
SM(gainBoundaries[0], AR_PHY_TPCRG5_PD_GAIN_BOUNDARY_1) |
SM(gainBoundaries[1], AR_PHY_TPCRG5_PD_GAIN_BOUNDARY_2) |
SM(gainBoundaries[2], AR_PHY_TPCRG5_PD_GAIN_BOUNDARY_3) |
SM(gainBoundaries[3], AR_PHY_TPCRG5_PD_GAIN_BOUNDARY_4));
}
/*
* Get the gain values and the number of gain levels given
* in xpdMask.
*
* The EEPROM xpdMask determines which power detector gain
* levels were used during calibration. Each of these mask
* bits maps to a fixed gain level in hardware.
*/
uint16_t
ar5416GetXpdGainValues(struct ath_hal *ah, uint16_t xpdMask,
uint16_t xpdGainValues[])
{
int i;
uint16_t numXpdGain = 0;
for (i = 1; i <= AR5416_PD_GAINS_IN_MASK; i++) {
if ((xpdMask >> (AR5416_PD_GAINS_IN_MASK - i)) & 1) {
if (numXpdGain >= AR5416_NUM_PD_GAINS) {
HALASSERT(0);
break;
}
xpdGainValues[numXpdGain] = (uint16_t)(AR5416_PD_GAINS_IN_MASK - i);
numXpdGain++;
}
}
return numXpdGain;
}
/*
* Write the detector gain and biases.
*
* There are four power detector gain levels. The xpdMask in the EEPROM
* determines which power detector gain levels have TX power calibration
* data associated with them. This function writes the number of
* PD gain levels and their values into the hardware.
*
* This is valid for all TX chains - the calibration data itself however
* will likely differ per-chain.
*/
void
ar5416WriteDetectorGainBiases(struct ath_hal *ah, uint16_t numXpdGain,
uint16_t xpdGainValues[])
{
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
HALDEBUG(ah, HAL_DEBUG_EEPROM, "%s: numXpdGain: %d,"
" xpdGainValues: %d, %d, %d\n", __func__, numXpdGain,
xpdGainValues[0], xpdGainValues[1], xpdGainValues[2]);
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
OS_REG_WRITE(ah, AR_PHY_TPCRG1, (OS_REG_READ(ah, AR_PHY_TPCRG1) &
~(AR_PHY_TPCRG1_NUM_PD_GAIN | AR_PHY_TPCRG1_PD_GAIN_1 |
AR_PHY_TPCRG1_PD_GAIN_2 | AR_PHY_TPCRG1_PD_GAIN_3)) |
SM(numXpdGain - 1, AR_PHY_TPCRG1_NUM_PD_GAIN) |
SM(xpdGainValues[0], AR_PHY_TPCRG1_PD_GAIN_1 ) |
SM(xpdGainValues[1], AR_PHY_TPCRG1_PD_GAIN_2) |
SM(xpdGainValues[2], AR_PHY_TPCRG1_PD_GAIN_3));
}
/*
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
* Write the PDADC array to the given radio chain i.
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
*
* The 32 PDADC registers are written without any care about
* their contents - so if various chips treat values as "special",
* this routine will not care.
*/
void
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
ar5416WritePdadcValues(struct ath_hal *ah, int i, uint8_t pdadcValues[])
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
{
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
int regOffset, regChainOffset;
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
int j;
int reg32;
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
regChainOffset = ar5416GetRegChainOffset(ah, i);
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
regOffset = AR_PHY_BASE + (672 << 2) + regChainOffset;
for (j = 0; j < 32; j++) {
reg32 = ((pdadcValues[4*j + 0] & 0xFF) << 0) |
((pdadcValues[4*j + 1] & 0xFF) << 8) |
((pdadcValues[4*j + 2] & 0xFF) << 16) |
((pdadcValues[4*j + 3] & 0xFF) << 24) ;
OS_REG_WRITE(ah, regOffset, reg32);
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
HALDEBUG(ah, HAL_DEBUG_EEPROM, "PDADC: Chain %d |"
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
" PDADC %3d Value %3d | PDADC %3d Value %3d | PDADC %3d"
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
" Value %3d | PDADC %3d Value %3d |\n",
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
i,
4*j, pdadcValues[4*j],
4*j+1, pdadcValues[4*j + 1],
4*j+2, pdadcValues[4*j + 2],
4*j+3, pdadcValues[4*j + 3]);
regOffset += 4;
}
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/**************************************************************
* ar5416SetPowerCalTable
*
* Pull the PDADC piers from cal data and interpolate them across the given
* points as well as from the nearest pier(s) to get a power detector
* linear voltage to power level table.
*/
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
HAL_BOOL
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416SetPowerCalTable(struct ath_hal *ah, struct ar5416eeprom *pEepData,
const struct ieee80211_channel *chan, int16_t *pTxPowerIndexOffset)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
CAL_DATA_PER_FREQ *pRawDataset;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint8_t *pCalBChans = AH_NULL;
uint16_t pdGainOverlap_t2;
static uint8_t pdadcValues[AR5416_NUM_PDADC_VALUES];
uint16_t gainBoundaries[AR5416_PD_GAINS_IN_MASK];
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
uint16_t numPiers, i;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
int16_t tMinCalPower;
uint16_t numXpdGain, xpdMask;
uint16_t xpdGainValues[AR5416_NUM_PD_GAINS];
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
uint32_t regChainOffset;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
kill more portability functions that are no longer useful
2009-05-08 00:23:00 +00:00
OS_MEMZERO(xpdGainValues, sizeof(xpdGainValues));
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
xpdMask = pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].xpdGain;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (IS_EEP_MINOR_V2(ah)) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pdGainOverlap_t2 = pEepData->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)].pdGainOverlap;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
pdGainOverlap_t2 = (uint16_t)(MS(OS_REG_READ(ah, AR_PHY_TPCRG5), AR_PHY_TPCRG5_PD_GAIN_OVERLAP));
}
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pCalBChans = pEepData->calFreqPier2G;
numPiers = AR5416_NUM_2G_CAL_PIERS;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
pCalBChans = pEepData->calFreqPier5G;
numPiers = AR5416_NUM_5G_CAL_PIERS;
}
/* Calculate the value of xpdgains from the xpdGain Mask */
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
numXpdGain = ar5416GetXpdGainValues(ah, xpdMask, xpdGainValues);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Write the detector gain biases and their number */
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
ar5416WriteDetectorGainBiases(ah, numXpdGain, xpdGainValues);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
for (i = 0; i < AR5416_MAX_CHAINS; i++) {
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
regChainOffset = ar5416GetRegChainOffset(ah, i);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
if (pEepData->baseEepHeader.txMask & (1 << i)) {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_2GHZ(chan)) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pRawDataset = pEepData->calPierData2G[i];
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
pRawDataset = pEepData->calPierData5G[i];
}
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
/* Fetch the gain boundaries and the PDADC values */
ar5416GetGainBoundariesAndPdadcs(ah, chan, pRawDataset,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
pCalBChans, numPiers,
pdGainOverlap_t2,
&tMinCalPower, gainBoundaries,
pdadcValues, numXpdGain);
Change AR_SREV_OWL_{X}_OR_LATER to AR_SREV_5416_{X}_OR_LATER. For now, these are equivalent macros. AR_SREV_OWL{X}_OR_LATER will later change to exclude Howl (AR9130) in line with what the Atheros HAL does. This should not functionally change anything. Obtained from: Atheros
2011-05-07 02:59:24 +00:00
if ((i == 0) || AR_SREV_5416_V20_OR_LATER(ah)) {
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
ar5416SetGainBoundariesClosedLoop(ah, i, pdGainOverlap_t2,
gainBoundaries);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
/* Write the power values into the baseband power table */
* Add in some board settings debugging to log what's being written to the TX closed-loop power control registers. * Modify a couple of functions to take the register chain number, rather than the regChainOffset value. This allows for the register chain to be logged.
2011-03-13 05:30:14 +00:00
ar5416WritePdadcValues(ah, i, pdadcValues);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
}
*pTxPowerIndexOffset = 0;
return AH_TRUE;
}
/**************************************************************
* ar5416GetGainBoundariesAndPdadcs
*
* Uses the data points read from EEPROM to reconstruct the pdadc power table
* Called by ar5416SetPowerCalTable only.
*/
Implement open-loop TX power control (OLC) for Merlin (AR9280) and generally tidy up the TX power programming code. Enforce that the TX power offset for Merlin is -5 dBm, rather than any other value programmable in the EEPROM. This requires some further code to be ported over from ath9k, so until that is done and tested, fail to attach NICs whose TX power offset isn't -5 dBm. This improves both legacy and HT transmission on my merlin board. It allows for stable MCS TX up to MCS15. Specifics: * Refactor out a bunch of the TX power calibration code - setting/obtaining the power detector / gain boundaries, programming the PDADC * Take the -5 dBm TX power offset into account on Merlin - "0" in the per-rate TX power register means -5 dBm, not 0 dBm * When doing OLC * Enforce min (0) and max (AR5416_MAX_RATE_POWER) when fiddling with the TX power, to avoid the TX power values from wrapping when low. * Implement the 1 dBm cck power offset when doing OLC * Implement temperature compensation for 2.4ghz mode when doing OLC * Implement an AR9280 specific TX power calibration routine which includes the OLC twiddles, leaving the earlier chipset path (AR5416, AR9160) alone Whilst here, use these refactored routines for the AR9285 TX power calibration/programming code and enforce correct overflow/underflow handling when fiddling with TX power values. Obtained from: linux ath9k
2011-03-08 06:59:59 +00:00
void
virgin import of ath hal
2008-11-28 00:03:41 +00:00
ar5416GetGainBoundariesAndPdadcs(struct ath_hal *ah,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
const struct ieee80211_channel *chan,
CAL_DATA_PER_FREQ *pRawDataSet,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint8_t * bChans, uint16_t availPiers,
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
uint16_t tPdGainOverlap, int16_t *pMinCalPower, uint16_t * pPdGainBoundaries,
virgin import of ath hal
2008-11-28 00:03:41 +00:00
uint8_t * pPDADCValues, uint16_t numXpdGains)
{
int i, j, k;
int16_t ss; /* potentially -ve index for taking care of pdGainOverlap */
uint16_t idxL, idxR, numPiers; /* Pier indexes */
/* filled out Vpd table for all pdGains (chanL) */
static uint8_t vpdTableL[AR5416_NUM_PD_GAINS][AR5416_MAX_PWR_RANGE_IN_HALF_DB];
/* filled out Vpd table for all pdGains (chanR) */
static uint8_t vpdTableR[AR5416_NUM_PD_GAINS][AR5416_MAX_PWR_RANGE_IN_HALF_DB];
/* filled out Vpd table for all pdGains (interpolated) */
static uint8_t vpdTableI[AR5416_NUM_PD_GAINS][AR5416_MAX_PWR_RANGE_IN_HALF_DB];
uint8_t *pVpdL, *pVpdR, *pPwrL, *pPwrR;
uint8_t minPwrT4[AR5416_NUM_PD_GAINS];
uint8_t maxPwrT4[AR5416_NUM_PD_GAINS];
int16_t vpdStep;
int16_t tmpVal;
uint16_t sizeCurrVpdTable, maxIndex, tgtIndex;
HAL_BOOL match;
int16_t minDelta = 0;
CHAN_CENTERS centers;
ar5416GetChannelCenters(ah, chan, &centers);
/* Trim numPiers for the number of populated channel Piers */
for (numPiers = 0; numPiers < availPiers; numPiers++) {
if (bChans[numPiers] == AR5416_BCHAN_UNUSED) {
break;
}
}
/* Find pier indexes around the current channel */
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
match = ath_ee_getLowerUpperIndex((uint8_t)FREQ2FBIN(centers.synth_center,
IEEE80211_IS_CHAN_2GHZ(chan)), bChans, numPiers, &idxL, &idxR);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (match) {
/* Directly fill both vpd tables from the matching index */
for (i = 0; i < numXpdGains; i++) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
minPwrT4[i] = pRawDataSet[idxL].pwrPdg[i][0];
maxPwrT4[i] = pRawDataSet[idxL].pwrPdg[i][4];
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
ath_ee_FillVpdTable(minPwrT4[i], maxPwrT4[i], pRawDataSet[idxL].pwrPdg[i],
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pRawDataSet[idxL].vpdPdg[i], AR5416_PD_GAIN_ICEPTS, vpdTableI[i]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
} else {
for (i = 0; i < numXpdGains; i++) {
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
pVpdL = pRawDataSet[idxL].vpdPdg[i];
pPwrL = pRawDataSet[idxL].pwrPdg[i];
pVpdR = pRawDataSet[idxR].vpdPdg[i];
pPwrR = pRawDataSet[idxR].pwrPdg[i];
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Start Vpd interpolation from the max of the minimum powers */
minPwrT4[i] = AH_MAX(pPwrL[0], pPwrR[0]);
/* End Vpd interpolation from the min of the max powers */
maxPwrT4[i] = AH_MIN(pPwrL[AR5416_PD_GAIN_ICEPTS - 1], pPwrR[AR5416_PD_GAIN_ICEPTS - 1]);
HALASSERT(maxPwrT4[i] > minPwrT4[i]);
/* Fill pier Vpds */
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
ath_ee_FillVpdTable(minPwrT4[i], maxPwrT4[i], pPwrL, pVpdL, AR5416_PD_GAIN_ICEPTS, vpdTableL[i]);
ath_ee_FillVpdTable(minPwrT4[i], maxPwrT4[i], pPwrR, pVpdR, AR5416_PD_GAIN_ICEPTS, vpdTableR[i]);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Interpolate the final vpd */
for (j = 0; j <= (maxPwrT4[i] - minPwrT4[i]) / 2; j++) {
Move out some of the shared eeprom board value calculation routines into ah.c rather than duplicating them for the v14 (ar5416+) and v4k (ar9285) codebases. Further chipsets (eg the AR9287) have yet another EEPROM format which will use these routines to calculate things.
2011-03-13 05:54:05 +00:00
vpdTableI[i][j] = (uint8_t)(ath_ee_interpolate((uint16_t)FREQ2FBIN(centers.synth_center,
IEEE80211_IS_CHAN_2GHZ(chan)),
virgin import of ath hal
2008-11-28 00:03:41 +00:00
bChans[idxL], bChans[idxR], vpdTableL[i][j], vpdTableR[i][j]));
}
}
}
*pMinCalPower = (int16_t)(minPwrT4[0] / 2);
k = 0; /* index for the final table */
for (i = 0; i < numXpdGains; i++) {
if (i == (numXpdGains - 1)) {
pPdGainBoundaries[i] = (uint16_t)(maxPwrT4[i] / 2);
} else {
pPdGainBoundaries[i] = (uint16_t)((maxPwrT4[i] + minPwrT4[i+1]) / 4);
}
pPdGainBoundaries[i] = (uint16_t)AH_MIN(AR5416_MAX_RATE_POWER, pPdGainBoundaries[i]);
Update to later code from my repository: o many bug fixes o add new periodic calibration api o break up 5416 periodic calibration code in preparation for 928x o move get noise floor to rf backends o 5416-specific ani (still disabled) o modularize 5210 eeprom format a la other eeprom formats o start cleaning up regdomain code o prepare for proper 1/2 and 1/4 width channel support o bring back 900MHz card support o clean up 5212 rf version handling o add 1/2 and 1/4 width channel support for 5212 parts o split 5212 rfgain handling out o improve ani debugging o add AH_USE_INIPDGAIN compile option o purge a bunch of dead 5212 state o add 1/2 and 1/4 rate modes o remove HAL_CAP_CHAN_HALFRATE and HAL_CAP_CHAN_QUARTERRATE; the same info can now be deduced from the set of supported modes
2008-11-28 00:48:05 +00:00
/* NB: only applies to owl 1.0 */
Change AR_SREV_OWL_{X}_OR_LATER to AR_SREV_5416_{X}_OR_LATER. For now, these are equivalent macros. AR_SREV_OWL{X}_OR_LATER will later change to exclude Howl (AR9130) in line with what the Atheros HAL does. This should not functionally change anything. Obtained from: Atheros
2011-05-07 02:59:24 +00:00
if ((i == 0) && !AR_SREV_5416_V20_OR_LATER(ah) ) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* fix the gain delta, but get a delta that can be applied to min to
* keep the upper power values accurate, don't think max needs to
* be adjusted because should not be at that area of the table?
*/
minDelta = pPdGainBoundaries[0] - 23;
pPdGainBoundaries[0] = 23;
}
else {
minDelta = 0;
}
/* Find starting index for this pdGain */
if (i == 0) {
Major fix: when doing open-loop TX power calibration, adjust the correct CCK rates rather than adjusting the first handful. This may have affected some AR9280 based NICs. Minor fix: merlin check update
2011-05-14 04:17:16 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah))
Merlin fix - first pdadc gain index is 0 - minpwr/2 . Obtained from: Linux ath9k
2011-03-10 06:06:26 +00:00
ss = (int16_t)(0 - (minPwrT4[i] / 2));
else
ss = 0; /* for the first pdGain, start from index 0 */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
} else {
/* need overlap entries extrapolated below. */
ss = (int16_t)((pPdGainBoundaries[i-1] - (minPwrT4[i] / 2)) - tPdGainOverlap + 1 + minDelta);
}
vpdStep = (int16_t)(vpdTableI[i][1] - vpdTableI[i][0]);
vpdStep = (int16_t)((vpdStep < 1) ? 1 : vpdStep);
/*
*-ve ss indicates need to extrapolate data below for this pdGain
*/
while ((ss < 0) && (k < (AR5416_NUM_PDADC_VALUES - 1))) {
tmpVal = (int16_t)(vpdTableI[i][0] + ss * vpdStep);
pPDADCValues[k++] = (uint8_t)((tmpVal < 0) ? 0 : tmpVal);
ss++;
}
Revert part of the 9285 support because it breaks the 9280 support. I'll try to do the 9285 support without interfering with any other chipset revisions support.
2010-02-14 16:26:32 +00:00
sizeCurrVpdTable = (uint8_t)((maxPwrT4[i] - minPwrT4[i]) / 2 +1);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
tgtIndex = (uint8_t)(pPdGainBoundaries[i] + tPdGainOverlap - (minPwrT4[i] / 2));
maxIndex = (tgtIndex < sizeCurrVpdTable) ? tgtIndex : sizeCurrVpdTable;
while ((ss < maxIndex) && (k < (AR5416_NUM_PDADC_VALUES - 1))) {
pPDADCValues[k++] = vpdTableI[i][ss++];
}
vpdStep = (int16_t)(vpdTableI[i][sizeCurrVpdTable - 1] - vpdTableI[i][sizeCurrVpdTable - 2]);
vpdStep = (int16_t)((vpdStep < 1) ? 1 : vpdStep);
/*
* for last gain, pdGainBoundary == Pmax_t2, so will
* have to extrapolate
*/
Merge in a fix for the power/(gain?) calculation. Apply it to both the 5416/9160 and 9285 code paths. Obtained from: OpenWRT r22123, 522-ath9k_pwrcal_fix.patch
2010-08-14 15:29:21 +00:00
if (tgtIndex >= maxIndex) { /* need to extrapolate above */
virgin import of ath hal
2008-11-28 00:03:41 +00:00
while ((ss <= tgtIndex) && (k < (AR5416_NUM_PDADC_VALUES - 1))) {
tmpVal = (int16_t)((vpdTableI[i][sizeCurrVpdTable - 1] +
(ss - maxIndex +1) * vpdStep));
pPDADCValues[k++] = (uint8_t)((tmpVal > 255) ? 255 : tmpVal);
ss++;
}
} /* extrapolated above */
} /* for all pdGainUsed */
/* Fill out pdGainBoundaries - only up to 2 allowed here, but hardware allows up to 4 */
while (i < AR5416_PD_GAINS_IN_MASK) {
pPdGainBoundaries[i] = pPdGainBoundaries[i-1];
i++;
}
while (k < AR5416_NUM_PDADC_VALUES) {
pPDADCValues[k] = pPDADCValues[k-1];
k++;
}
return;
}
Enable the 11n PHY by default whether or not 11n is configured. The linux ath9k driver and (from what I've been told) the atheros reference driver does this; it then leaves discarding 11n frames to the 802.11 layer. Whilst I'm here, merge in a fix from ath9k which maintains a turbo register setting when enabling the 11n register; and remove an un-needed (duplicate) flag setting.
2011-01-23 14:49:50 +00:00
/*
* The linux ath9k driver and (from what I've been told) the reference
* Atheros driver enables the 11n PHY by default whether or not it's
* configured.
*/
virgin import of ath hal
2008-11-28 00:03:41 +00:00
static void
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
ar5416Set11nRegs(struct ath_hal *ah, const struct ieee80211_channel *chan)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
uint32_t phymode;
Enable the 11n PHY by default whether or not 11n is configured. The linux ath9k driver and (from what I've been told) the atheros reference driver does this; it then leaves discarding 11n frames to the 802.11 layer. Whilst I'm here, merge in a fix from ath9k which maintains a turbo register setting when enabling the 11n register; and remove an un-needed (duplicate) flag setting.
2011-01-23 14:49:50 +00:00
uint32_t enableDacFifo = 0;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HAL_HT_MACMODE macmode; /* MAC - 20/40 mode */
Enable the 11n PHY by default whether or not 11n is configured. The linux ath9k driver and (from what I've been told) the atheros reference driver does this; it then leaves discarding 11n frames to the 802.11 layer. Whilst I'm here, merge in a fix from ath9k which maintains a turbo register setting when enabling the 11n register; and remove an un-needed (duplicate) flag setting.
2011-01-23 14:49:50 +00:00
if (AR_SREV_KITE_10_OR_LATER(ah))
enableDacFifo = (OS_REG_READ(ah, AR_PHY_TURBO) & AR_PHY_FC_ENABLE_DAC_FIFO);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Enable 11n HT, 20 MHz */
phymode = AR_PHY_FC_HT_EN | AR_PHY_FC_SHORT_GI_40
Enable the 11n PHY by default whether or not 11n is configured. The linux ath9k driver and (from what I've been told) the atheros reference driver does this; it then leaves discarding 11n frames to the 802.11 layer. Whilst I'm here, merge in a fix from ath9k which maintains a turbo register setting when enabling the 11n register; and remove an un-needed (duplicate) flag setting.
2011-01-23 14:49:50 +00:00
| AR_PHY_FC_SINGLE_HT_LTF1 | AR_PHY_FC_WALSH | enableDacFifo;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Configure baseband for dynamic 20/40 operation */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40(chan)) {
Enable the 11n PHY by default whether or not 11n is configured. The linux ath9k driver and (from what I've been told) the atheros reference driver does this; it then leaves discarding 11n frames to the 802.11 layer. Whilst I'm here, merge in a fix from ath9k which maintains a turbo register setting when enabling the 11n register; and remove an un-needed (duplicate) flag setting.
2011-01-23 14:49:50 +00:00
phymode |= AR_PHY_FC_DYN2040_EN;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Configure control (primary) channel at +-10MHz */
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40U(chan))
virgin import of ath hal
2008-11-28 00:03:41 +00:00
phymode |= AR_PHY_FC_DYN2040_PRI_CH;
#if 0
/* Configure 20/25 spacing */
if (ht->ht_extprotspacing == HAL_HT_EXTPROTSPACING_25)
phymode |= AR_PHY_FC_DYN2040_EXT_CH;
#endif
macmode = HAL_HT_MACMODE_2040;
} else
macmode = HAL_HT_MACMODE_20;
OS_REG_WRITE(ah, AR_PHY_TURBO, phymode);
/* Configure MAC for 20/40 operation */
ar5416Set11nMac2040(ah, macmode);
/* global transmit timeout (25 TUs default)*/
/* XXX - put this elsewhere??? */
OS_REG_WRITE(ah, AR_GTXTO, 25 << AR_GTXTO_TIMEOUT_LIMIT_S) ;
/* carrier sense timeout */
OS_REG_SET_BIT(ah, AR_GTTM, AR_GTTM_CST_USEC);
bring this in line with what ath9k does.
2011-02-14 21:35:11 +00:00
OS_REG_WRITE(ah, AR_CST, 0xF << AR_CST_TIMEOUT_LIMIT_S);
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
void
ar5416GetChannelCenters(struct ath_hal *ah,
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
const struct ieee80211_channel *chan, CHAN_CENTERS *centers)
virgin import of ath hal
2008-11-28 00:03:41 +00:00
{
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
uint16_t freq = ath_hal_gethwchannel(ah, chan);
centers->ctl_center = freq;
centers->synth_center = freq;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* In 20/40 phy mode, the center frequency is
* "between" the control and extension channels.
*/
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
if (IEEE80211_IS_CHAN_HT40U(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
centers->synth_center += HT40_CHANNEL_CENTER_SHIFT;
centers->ext_center =
centers->synth_center + HT40_CHANNEL_CENTER_SHIFT;
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
} else if (IEEE80211_IS_CHAN_HT40D(chan)) {
virgin import of ath hal
2008-11-28 00:03:41 +00:00
centers->synth_center -= HT40_CHANNEL_CENTER_SHIFT;
centers->ext_center =
centers->synth_center - HT40_CHANNEL_CENTER_SHIFT;
} else {
Overhaul regulatory support: o remove HAL_CHANNEL; convert the hal to use net80211 channels; this mostly involves mechanical changes to variable names and channel attribute macros o gut HAL_CHANNEL_PRIVATE as most of the contents are now redundant with the net80211 channel available o change api for ath_hal_init_channels: no more reglass id's, no more outdoor indication (was a noop), anM contents o add ath_hal_getchannels to have the hal construct a channel list without altering runtime state; this is used to retrieve the calibration list for the device in ath_getradiocaps o add ath_hal_set_channels to take a channel list and regulatory data from above and construct internal state to match (maps frequencies for 900MHz cards, setup for CTL lookups, etc) o compact the private channel table: we keep one private channel per frequency instead of one per HAL_CHANNEL; this gives a big space savings and potentially improves ani and calibration by sharing state (to be seen; didn't see anything in testing); a new config option AH_MAXCHAN controls the table size (default to 96 which was chosen to be ~3x the largest expected size) o shrink ani state and change to mirror private channel table (one entry per frequency indexed by ic_devdata) o move ani state flags to private channel state o remove country codes; use net80211 definitions instead o remove GSM regulatory support; it's no longer needed now that we pass in channel lists from above o consolidate ADHOC_NO_11A attribute with DISALLOW_ADHOC_11A o simplify initial channel list construction based on the EEPROM contents; we preserve country code support for now but may want to just fallback to a WWR sku and dispatch the discovered country code up to user space so the channel list can be constructed using the master regdomain tables o defer to net80211 for max antenna gain o eliminate sorting of internal channel table; now that we use ic_devdata as an index, table lookups are O(1) o remove internal copy of the country code; the public one is sufficient o remove AH_SUPPORT_11D conditional compilation; we always support 11d o remove ath_hal_ispublicsafetysku; not needed any more o remove ath_hal_isgsmsku; no more GSM stuff o move Conformance Test Limit (CTL) state from private channel to a lookup using per-band pointers cached in the private state block o remove regulatory class id support; was unused and belongs in net80211 o fix channel list construction to set IEEE80211_CHAN_NOADHOC, IEEE80211_CHAN_NOHOSTAP, and IEEE80211_CHAN_4MSXMIT o remove private channel flags CHANNEL_DFS and CHANNEL_4MS_LIMIT; these are now set in the constructed net80211 channel o store CHANNEL_NFCREQUIRED (Noise Floor Required) channel attribute in one of the driver-private flag bits of the net80211 channel o move 900MHz frequency mapping into the hal; the mapped frequency is stored in the private channel and used throughout the hal (no more mapping in the driver and/or net80211) o remove ath_hal_mhz2ieee; it's no longer needed as net80211 does the calculation and available in the net80211 channel o change noise floor calibration logic to work with compacted private channel table setup; this may require revisiting as we no longer can distinguish channel attributes (e.g. 11b vs 11g vs turbo) but since the data is used only to calculate status data we can live with it for now o change ah_getChipPowerLimits internal method to operate on a single channel instead of all channels in the private channel table o add ath_hal_gethwchannel to map a net80211 channel to a h/w frequency (always the same except for 900MHz channels) o add HAL_EEBADREG and HAL_EEBADCC status codes to better identify regulatory problems o remove CTRY_DEBUG and CTRY_DEFAULT enum's; these come from net80211 now o change ath_hal_getwirelessmodes to really return wireless modes supported by the hardware (was previously applying regulatory constraints) o return channel interference status with IEEE80211_CHANSTATE_CWINT (should change to a callback so hal api's can take const pointers) o remove some #define's no longer needed with the inclusion of <net80211/_ieee80211.h> Sponsored by: Carlson Wireless
2009-01-28 18:00:22 +00:00
centers->ext_center = freq;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
}
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
/*
* Override the INI vals being programmed.
*/
static void
ar5416OverrideIni(struct ath_hal *ah, const struct ieee80211_channel *chan)
{
uint32_t val;
/*
* Set the RX_ABORT and RX_DIS and clear if off only after
* RXE is set for MAC. This prevents frames with corrupted
* descriptor status.
*/
OS_REG_SET_BIT(ah, AR_DIAG_SW, (AR_DIAG_RX_DIS | AR_DIAG_RX_ABORT));
Properly indent the WAR code i pasted in from ath9k a few months ago.
2011-05-08 05:45:06 +00:00
if (AR_SREV_MERLIN_10_OR_LATER(ah)) {
val = OS_REG_READ(ah, AR_PCU_MISC_MODE2);
* Add in a comment about ar5416InitUserSettings() potentially modifying AR_DIAG_SW. There's a hardware workaround which sets disabling some errors early at startup and clears said bits before the PCU begins receiving - it does this to avoid RX descriptor status errors. It's possible these bits aren't being completely properly twiddled in all instances; but in particular if the diag_reg HAL variable is set it won't be setting these bits correctly. I'll review this at some point. * Disable multicast search on mac address and key id - the driver doesn't use it at the moment and thus adhoc may be broken for merlin and later. * Change this to be for Merlin 1.0 (which from what I understand wasn't ever publicly released) to be more correct.
2011-05-08 05:25:42 +00:00
val &= (~AR_PCU_MISC_MODE2_ADHOC_MCAST_KEYID_ENABLE);
Properly indent the WAR code i pasted in from ath9k a few months ago.
2011-05-08 05:45:06 +00:00
if (!AR_SREV_9271(ah))
val &= ~AR_PCU_MISC_MODE2_HWWAR1;
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
Even though the HAL doesn't currently support Kiwi 1.0/1.1, be "more correct" about the Kiwi setup. Obtained from: Atheros
2011-11-09 19:09:03 +00:00
if (AR_SREV_KIWI_10_OR_LATER(ah))
Properly indent the WAR code i pasted in from ath9k a few months ago.
2011-05-08 05:45:06 +00:00
val = val & (~AR_PCU_MISC_MODE2_HWWAR2);
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
Properly indent the WAR code i pasted in from ath9k a few months ago.
2011-05-08 05:45:06 +00:00
OS_REG_WRITE(ah, AR_PCU_MISC_MODE2, val);
}
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
/*
* Disable RIFS search on some chips to avoid baseband
* hang issues.
*/
Add a function which enables or disables RX RIFS searching, and migrate the code which does this into it.
2011-05-06 15:33:56 +00:00
if (AR_SREV_HOWL(ah) || AR_SREV_SOWL(ah))
When disabling RIFS for Sowl (AR9160) and Howl (AR9130), make sure RIFS is totally disabled. The Atheros HAL code does this for Sowl/Howl but not for Owl (AR5416) where RIFS is disabled by default. This seems to quieten the occasional baseband hang I've been seeing with the AR9160 in STA mode under constant heavy traffic load. Obtained from: Atheros
2011-05-14 05:43:33 +00:00
(void) ar5416SetRifsDelay(ah, chan, AH_FALSE);
Change AR_SREV_OWL_{X}_OR_LATER to AR_SREV_5416_{X}_OR_LATER. For now, these are equivalent macros. AR_SREV_OWL{X}_OR_LATER will later change to exclude Howl (AR9130) in line with what the Atheros HAL does. This should not functionally change anything. Obtained from: Atheros
2011-05-07 02:59:24 +00:00
if (!AR_SREV_5416_V20_OR_LATER(ah) || AR_SREV_MERLIN(ah))
return;
/*
* Disable BB clock gating
* Necessary to avoid issues on AR5416 2.0
*/
OS_REG_WRITE(ah, 0x9800 + (651 << 2), 0x11);
Port over ar5416OverrideIni() from ath9k ar5008_hw_override_ini(). * change the BB gating logic to explicitly define which chips are covered; the ath9k method isn't as clear. * don't disable the BB gating for now, the ar5416 initvals have it, and the ar9160 initval sets it to 0x0. Figure out why before re-enabling this. * migrate the Merlin (ar9280) applicable WAR from the Kite (ar9285) code (which won't get called for Merlin!) and stuff it in here.
2011-03-03 08:38:31 +00:00
}
Bring over an XPA (external power amplifer) bias fix for the AR9160. This fix modifies the const addac initval array, rather than modifying a local copy. It means that running >1 AR9160 on a board may prove to be unpredictable. The AR5416 init path also does something similar, so supporting >1 AR5416 of different revisions could cause problems. The later fix will be to create a private copy of the Addac data for the AR5416, AR9160 (and AR9100 when it's merged in) and then modify that as needed. Obtained From: Linux ath9k
2011-03-22 10:29:36 +00:00
struct ini {
uint32_t *data; /* NB: !const */
int rows, cols;
};
/*
* Override XPA bias level based on operating frequency.
* This is a v14 EEPROM specific thing for the AR9160.
*/
void
ar5416EepromSetAddac(struct ath_hal *ah, const struct ieee80211_channel *chan)
{
#define XPA_LVL_FREQ(cnt) (pModal->xpaBiasLvlFreq[cnt])
MODAL_EEP_HEADER *pModal;
HAL_EEPROM_v14 *ee = AH_PRIVATE(ah)->ah_eeprom;
struct ar5416eeprom *eep = &ee->ee_base;
uint8_t biaslevel;
if (! AR_SREV_SOWL(ah))
return;
if (EEP_MINOR(ah) < AR5416_EEP_MINOR_VER_7)
return;
pModal = &(eep->modalHeader[IEEE80211_IS_CHAN_2GHZ(chan)]);
if (pModal->xpaBiasLvl != 0xff)
biaslevel = pModal->xpaBiasLvl;
else {
uint16_t resetFreqBin, freqBin, freqCount = 0;
CHAN_CENTERS centers;
ar5416GetChannelCenters(ah, chan, &centers);
resetFreqBin = FREQ2FBIN(centers.synth_center, IEEE80211_IS_CHAN_2GHZ(chan));
freqBin = XPA_LVL_FREQ(0) & 0xff;
biaslevel = (uint8_t) (XPA_LVL_FREQ(0) >> 14);
freqCount++;
while (freqCount < 3) {
if (XPA_LVL_FREQ(freqCount) == 0x0)
break;
freqBin = XPA_LVL_FREQ(freqCount) & 0xff;
if (resetFreqBin >= freqBin)
biaslevel = (uint8_t)(XPA_LVL_FREQ(freqCount) >> 14);
else
break;
freqCount++;
}
}
HALDEBUG(ah, HAL_DEBUG_EEPROM, "%s: overriding XPA bias level = %d\n",
__func__, biaslevel);
/*
* This is a dirty workaround for the const initval data,
* which will upset multiple AR9160's on the same board.
*
* The HAL should likely just have a private copy of the addac
* data per instance.
*/
if (IEEE80211_IS_CHAN_2GHZ(chan))
HAL_INI_VAL((struct ini *) &AH5416(ah)->ah_ini_addac, 7, 1) =
(HAL_INI_VAL(&AH5416(ah)->ah_ini_addac, 7, 1) & (~0x18)) | biaslevel << 3;
else
HAL_INI_VAL((struct ini *) &AH5416(ah)->ah_ini_addac, 6, 1) =
(HAL_INI_VAL(&AH5416(ah)->ah_ini_addac, 6, 1) & (~0xc0)) | biaslevel << 6;
#undef XPA_LVL_FREQ
}
Mark the PHY as inactive before the chip is reset. It's also marked inactive by the initvals, and enabled after the baseband/PLL has been configured, but before the RF registers have been programmed. The origin and reason for this particular change is currently unknown. Obtained from: Linux ath9k
2011-04-17 13:46:13 +00:00
static void
ar5416MarkPhyInactive(struct ath_hal *ah)
{
OS_REG_WRITE(ah, AR_PHY_ACTIVE, AR_PHY_ACTIVE_DIS);
}
Introduce an AR5416 flavour of the IFS and mac usec/timing configuration used when running the chips in half/quarter rate. This sets up some default parameters which are then overridden by the driver (which manually configures things like slot timing at interface start time.) Although this is a copy-and-modify from the AR5212 HAL, I did peek at the reference HAL and the ath9k driver to see what they did. Ath9k in particular doesn't hard-code this - instead, their version of ar5416InitUserSettings() does all of the relevant math. TODO: * do the math, not hard code things! * fix the mac clock calculation for the AR9287; since it runs the MAC clock at a higher rate, requiring all the duration calculations to change; * Do a whole lot more validation for half/quarter rates. Obtained from: Qualcomm Atheros, Linux ath9k
2012-09-13 07:22:40 +00:00
#define AR5416_IFS_SLOT_FULL_RATE_40 0x168 /* 9 us half, 40 MHz core clock (9*40) */
#define AR5416_IFS_SLOT_HALF_RATE_40 0x104 /* 13 us half, 20 MHz core clock (13*20) */
#define AR5416_IFS_SLOT_QUARTER_RATE_40 0xD2 /* 21 us quarter, 10 MHz core clock (21*10) */
#define AR5416_IFS_EIFS_FULL_RATE_40 0xE60 /* (74 + (2 * 9)) * 40MHz core clock */
#define AR5416_IFS_EIFS_HALF_RATE_40 0xDAC /* (149 + (2 * 13)) * 20MHz core clock */
#define AR5416_IFS_EIFS_QUARTER_RATE_40 0xD48 /* (298 + (2 * 21)) * 10MHz core clock */
#define AR5416_IFS_SLOT_FULL_RATE_44 0x18c /* 9 us half, 44 MHz core clock (9*44) */
#define AR5416_IFS_SLOT_HALF_RATE_44 0x11e /* 13 us half, 22 MHz core clock (13*22) */
#define AR5416_IFS_SLOT_QUARTER_RATE_44 0xe7 /* 21 us quarter, 11 MHz core clock (21*11) */
#define AR5416_IFS_EIFS_FULL_RATE_44 0xfd0 /* (74 + (2 * 9)) * 44MHz core clock */
#define AR5416_IFS_EIFS_HALF_RATE_44 0xf0a /* (149 + (2 * 13)) * 22MHz core clock */
#define AR5416_IFS_EIFS_QUARTER_RATE_44 0xe9c /* (298 + (2 * 21)) * 11MHz core clock */
#define AR5416_INIT_USEC_40 40
#define AR5416_HALF_RATE_USEC_40 19 /* ((40 / 2) - 1 ) */
#define AR5416_QUARTER_RATE_USEC_40 9 /* ((40 / 4) - 1 ) */
#define AR5416_INIT_USEC_44 44
#define AR5416_HALF_RATE_USEC_44 21 /* ((44 / 2) - 1 ) */
#define AR5416_QUARTER_RATE_USEC_44 10 /* ((44 / 4) - 1 ) */
/* XXX What should these be for 40/44MHz clocks (and half/quarter) ? */
#define AR5416_RX_NON_FULL_RATE_LATENCY 63
#define AR5416_TX_HALF_RATE_LATENCY 108
#define AR5416_TX_QUARTER_RATE_LATENCY 216
/*
* Adjust various register settings based on half/quarter rate clock setting.
* This includes:
*
* + USEC, TX/RX latency,
* + IFS params: slot, eifs, misc etc.
*
* TODO:
*
* + Verify which other registers need to be tweaked;
* + Verify the behaviour of this for 5GHz fast and non-fast clock mode;
* + This just plain won't work for long distance links - the coverage class
* code isn't aware of the slot/ifs/ACK/RTS timeout values that need to
* change;
* + Verify whether the 32KHz USEC value needs to be kept for the 802.11n
* series chips?
* + Calculate/derive values for 2GHz, 5GHz, 5GHz fast clock
*/
static void
ar5416SetIFSTiming(struct ath_hal *ah, const struct ieee80211_channel *chan)
{
uint32_t txLat, rxLat, usec, slot, refClock, eifs, init_usec;
int clk_44 = 0;
HALASSERT(IEEE80211_IS_CHAN_HALF(chan) ||
IEEE80211_IS_CHAN_QUARTER(chan));
/* 2GHz and 5GHz fast clock - 44MHz; else 40MHz */
if (IEEE80211_IS_CHAN_2GHZ(chan))
clk_44 = 1;
else if (IEEE80211_IS_CHAN_5GHZ(chan) &&
IS_5GHZ_FAST_CLOCK_EN(ah, chan))
clk_44 = 1;
/* XXX does this need save/restoring for the 11n chips? */
refClock = OS_REG_READ(ah, AR_USEC) & AR_USEC_USEC32;
/*
* XXX This really should calculate things, not use
* hard coded values! Ew.
*/
if (IEEE80211_IS_CHAN_HALF(chan)) {
if (clk_44) {
slot = AR5416_IFS_SLOT_HALF_RATE_44;
rxLat = AR5416_RX_NON_FULL_RATE_LATENCY <<
AR5416_USEC_RX_LAT_S;
txLat = AR5416_TX_HALF_RATE_LATENCY <<
AR5416_USEC_TX_LAT_S;
usec = AR5416_HALF_RATE_USEC_44;
eifs = AR5416_IFS_EIFS_HALF_RATE_44;
init_usec = AR5416_INIT_USEC_44 >> 1;
} else {
slot = AR5416_IFS_SLOT_HALF_RATE_40;
rxLat = AR5416_RX_NON_FULL_RATE_LATENCY <<
AR5416_USEC_RX_LAT_S;
txLat = AR5416_TX_HALF_RATE_LATENCY <<
AR5416_USEC_TX_LAT_S;
usec = AR5416_HALF_RATE_USEC_40;
eifs = AR5416_IFS_EIFS_HALF_RATE_40;
init_usec = AR5416_INIT_USEC_40 >> 1;
}
} else { /* quarter rate */
if (clk_44) {
slot = AR5416_IFS_SLOT_QUARTER_RATE_44;
rxLat = AR5416_RX_NON_FULL_RATE_LATENCY <<
AR5416_USEC_RX_LAT_S;
txLat = AR5416_TX_QUARTER_RATE_LATENCY <<
AR5416_USEC_TX_LAT_S;
usec = AR5416_QUARTER_RATE_USEC_44;
eifs = AR5416_IFS_EIFS_QUARTER_RATE_44;
init_usec = AR5416_INIT_USEC_44 >> 2;
} else {
slot = AR5416_IFS_SLOT_QUARTER_RATE_40;
rxLat = AR5416_RX_NON_FULL_RATE_LATENCY <<
AR5416_USEC_RX_LAT_S;
txLat = AR5416_TX_QUARTER_RATE_LATENCY <<
AR5416_USEC_TX_LAT_S;
usec = AR5416_QUARTER_RATE_USEC_40;
eifs = AR5416_IFS_EIFS_QUARTER_RATE_40;
init_usec = AR5416_INIT_USEC_40 >> 2;
}
}
/* XXX verify these! */
OS_REG_WRITE(ah, AR_USEC, (usec | refClock | txLat | rxLat));
OS_REG_WRITE(ah, AR_D_GBL_IFS_SLOT, slot);
OS_REG_WRITE(ah, AR_D_GBL_IFS_EIFS, eifs);
OS_REG_RMW_FIELD(ah, AR_D_GBL_IFS_MISC,
AR_D_GBL_IFS_MISC_USEC_DURATION, init_usec);
}
Reference in New Issue Copy Permalink