d/freebsd-dev
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity
23401900d3
freebsd-dev/sys/dev/ath/ath_hal/ar5416/ar5416_interrupts.c

307 lines
8.4 KiB
C
Raw Normal View History

virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Copyright (c) 2002-2008 Sam Leffler, Errno Consulting
* 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.
*
Fill in gpio support for 5416 and later parts: o add output mux support o gpio pin count is chip-dependent o 9280 and 9285 do input handling different o hookup gpio interrupts o no need to save/restore soft led state around reset
2009-02-24 00:33:08 +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 "ar5416/ar5416.h"
#include "ar5416/ar5416reg.h"
/*
* Checks to see if an interrupt is pending on our NIC
*
* Returns: TRUE if an interrupt is pending
* FALSE if not
*/
HAL_BOOL
ar5416IsInterruptPending(struct ath_hal *ah)
{
uint32_t isr;
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))
return AH_TRUE;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Some platforms trigger our ISR before applying power to
* the card, so make sure the INTPEND is really 1, not 0xffffffff.
*/
isr = OS_REG_READ(ah, AR_INTR_ASYNC_CAUSE);
if (isr != AR_INTR_SPURIOUS && (isr & AR_INTR_MAC_IRQ) != 0)
return AH_TRUE;
isr = OS_REG_READ(ah, AR_INTR_SYNC_CAUSE);
if (isr != AR_INTR_SPURIOUS && (isr & AR_INTR_SYNC_DEFAULT))
return AH_TRUE;
return AH_FALSE;
}
/*
* Reads the Interrupt Status Register value from the NIC, thus deasserting
* the interrupt line, and returns both the masked and unmasked mapped ISR
* values. The value returned is mapped to abstract the hw-specific bit
* locations in the Interrupt Status Register.
*
Fix a corner-case of interrupt handling which resulted in potentially spurious (and fatal) interrupt errors. One user reported seeing this: Apr 22 18:04:24 ceres kernel: ar5416GetPendingInterrupts: fatal error, ISR_RAC 0x0 SYNC_CAUSE 0x2000 SYNC_CAUSE of 0x2000 is AR_INTR_SYNC_LOCAL_TIMEOUT which is a bus timeout; this shouldn't cause HAL_INT_FATAL to be set. After checking out ath9k, ath9k_ar9002_hw_get_isr() clears (*masked) before continuing, regardless of whether any bits in the ISR registers are set. So if AR_INTR_SYNC_CAUSE is set to something that isn't treated as fatal, and AR_ISR isn't read or is read and is 0, then (*masked) wouldn't be cleared. Thus any of the existing bits set that were passed in would be preserved in the output. The caller in if_ath - ath_intr() - wasn't setting the masked value to 0 before calling ath_hal_getisr(), so anything that was present in that uninitialised variable would be preserved in the case above of AR_ISR=0, AR_INTR_SYNC_CAUSE != 0; and if the HAL_INT_FATAL bit was set, a fatal condition would be interpreted and the chip was reset. This patch does the following: * ath_intr() - set masked to 0 before calling ath_hal_getisr(); * ar5416GetPendingInterrupts() - clear (*masked) before processing continues; so if the interrupt source is AR_INTR_SYNC_CAUSE and it isn't fatal, the hardware isn't reset via returning HAL_INT_FATAL. This doesn't fix any underlying errors which trigger AR_INTR_SYNC_LOCAL_TIMEOUT - which is a bus timeout of some sort - so that likely should be further investigated.
2011-04-23 06:37:09 +00:00
* (*masked) is cleared on initial call.
*
virgin import of ath hal
2008-11-28 00:03:41 +00:00
* Returns: A hardware-abstracted bitmap of all non-masked-out
* interrupts pending, as well as an unmasked value
*/
HAL_BOOL
ar5416GetPendingInterrupts(struct ath_hal *ah, HAL_INT *masked)
{
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 isr, isr0, isr1, sync_cause = 0;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/*
* Verify there's a mac interrupt and the RTC is on.
*/
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)) {
*masked = 0;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
isr = OS_REG_READ(ah, AR_ISR);
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
} else {
if ((OS_REG_READ(ah, AR_INTR_ASYNC_CAUSE) & AR_INTR_MAC_IRQ) &&
(OS_REG_READ(ah, AR_RTC_STATUS) & AR_RTC_STATUS_M) == AR_RTC_STATUS_ON)
isr = OS_REG_READ(ah, AR_ISR);
else
isr = 0;
sync_cause = OS_REG_READ(ah, AR_INTR_SYNC_CAUSE);
sync_cause &= AR_INTR_SYNC_DEFAULT;
*masked = 0;
Fix a corner-case of interrupt handling which resulted in potentially spurious (and fatal) interrupt errors. One user reported seeing this: Apr 22 18:04:24 ceres kernel: ar5416GetPendingInterrupts: fatal error, ISR_RAC 0x0 SYNC_CAUSE 0x2000 SYNC_CAUSE of 0x2000 is AR_INTR_SYNC_LOCAL_TIMEOUT which is a bus timeout; this shouldn't cause HAL_INT_FATAL to be set. After checking out ath9k, ath9k_ar9002_hw_get_isr() clears (*masked) before continuing, regardless of whether any bits in the ISR registers are set. So if AR_INTR_SYNC_CAUSE is set to something that isn't treated as fatal, and AR_ISR isn't read or is read and is 0, then (*masked) wouldn't be cleared. Thus any of the existing bits set that were passed in would be preserved in the output. The caller in if_ath - ath_intr() - wasn't setting the masked value to 0 before calling ath_hal_getisr(), so anything that was present in that uninitialised variable would be preserved in the case above of AR_ISR=0, AR_INTR_SYNC_CAUSE != 0; and if the HAL_INT_FATAL bit was set, a fatal condition would be interpreted and the chip was reset. This patch does the following: * ath_intr() - set masked to 0 before calling ath_hal_getisr(); * ar5416GetPendingInterrupts() - clear (*masked) before processing continues; so if the interrupt source is AR_INTR_SYNC_CAUSE and it isn't fatal, the hardware isn't reset via returning HAL_INT_FATAL. This doesn't fix any underlying errors which trigger AR_INTR_SYNC_LOCAL_TIMEOUT - which is a bus timeout of some sort - so that likely should be further investigated.
2011-04-23 06:37:09 +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 (isr == 0 && sync_cause == 0)
return AH_FALSE;
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (isr != 0) {
struct ath_hal_5212 *ahp = AH5212(ah);
uint32_t mask2;
mask2 = 0;
if (isr & AR_ISR_BCNMISC) {
uint32_t isr2 = OS_REG_READ(ah, AR_ISR_S2);
if (isr2 & AR_ISR_S2_TIM)
mask2 |= HAL_INT_TIM;
if (isr2 & AR_ISR_S2_DTIM)
mask2 |= HAL_INT_DTIM;
if (isr2 & AR_ISR_S2_DTIMSYNC)
mask2 |= HAL_INT_DTIMSYNC;
if (isr2 & (AR_ISR_S2_CABEND ))
mask2 |= HAL_INT_CABEND;
if (isr2 & AR_ISR_S2_GTT)
mask2 |= HAL_INT_GTT;
if (isr2 & AR_ISR_S2_CST)
mask2 |= HAL_INT_CST;
if (isr2 & AR_ISR_S2_TSFOOR)
mask2 |= HAL_INT_TSFOOR;
}
isr = OS_REG_READ(ah, AR_ISR_RAC);
if (isr == 0xffffffff) {
*masked = 0;
Remove extraneous semicolons, no functional changes. Submitted by: Marc Balmer <marc@msys.ch> MFC after: 1 week
2010-01-07 21:01:37 +00:00
return AH_FALSE;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
*masked = isr & HAL_INT_COMMON;
if (isr & (AR_ISR_RXOK | AR_ISR_RXERR))
*masked |= HAL_INT_RX;
if (isr & (AR_ISR_TXOK | AR_ISR_TXDESC | AR_ISR_TXERR | AR_ISR_TXEOL)) {
*masked |= HAL_INT_TX;
isr0 = OS_REG_READ(ah, AR_ISR_S0_S);
ahp->ah_intrTxqs |= MS(isr0, AR_ISR_S0_QCU_TXOK);
ahp->ah_intrTxqs |= MS(isr0, AR_ISR_S0_QCU_TXDESC);
isr1 = OS_REG_READ(ah, AR_ISR_S1_S);
ahp->ah_intrTxqs |= MS(isr1, AR_ISR_S1_QCU_TXERR);
ahp->ah_intrTxqs |= MS(isr1, AR_ISR_S1_QCU_TXEOL);
}
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(ah) || AR_SREV_KITE(ah)) {
uint32_t isr5;
isr5 = OS_REG_READ(ah, AR_ISR_S5_S);
if (isr5 & AR_ISR_S5_TIM_TIMER)
*masked |= HAL_INT_TIM_TIMER;
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Interrupt Mitigation on AR5416 */
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
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (isr & (AR_ISR_RXMINTR | AR_ISR_RXINTM))
*masked |= HAL_INT_RX;
if (isr & (AR_ISR_TXMINTR | AR_ISR_TXINTM))
*masked |= HAL_INT_TX;
#endif
*masked |= mask2;
}
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))
return AH_TRUE;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (sync_cause != 0) {
if (sync_cause & (AR_INTR_SYNC_HOST1_FATAL | AR_INTR_SYNC_HOST1_PERR)) {
*masked |= HAL_INT_FATAL;
}
if (sync_cause & AR_INTR_SYNC_RADM_CPL_TIMEOUT) {
HALDEBUG(ah, HAL_DEBUG_ANY, "%s: RADM CPL timeout\n",
__func__);
OS_REG_WRITE(ah, AR_RC, AR_RC_HOSTIF);
OS_REG_WRITE(ah, AR_RC, 0);
*masked |= HAL_INT_FATAL;
}
/*
* On fatal errors collect ISR state for debugging.
*/
if (*masked & HAL_INT_FATAL) {
AH_PRIVATE(ah)->ah_fatalState[0] = isr;
AH_PRIVATE(ah)->ah_fatalState[1] = sync_cause;
HALDEBUG(ah, HAL_DEBUG_ANY,
"%s: fatal error, ISR_RAC 0x%x SYNC_CAUSE 0x%x\n",
__func__, isr, sync_cause);
}
OS_REG_WRITE(ah, AR_INTR_SYNC_CAUSE_CLR, sync_cause);
/* NB: flush write */
(void) OS_REG_READ(ah, AR_INTR_SYNC_CAUSE_CLR);
}
return AH_TRUE;
}
/*
* Atomically enables NIC interrupts. Interrupts are passed in
* via the enumerated bitmask in ints.
*/
HAL_INT
ar5416SetInterrupts(struct ath_hal *ah, HAL_INT ints)
{
struct ath_hal_5212 *ahp = AH5212(ah);
uint32_t omask = ahp->ah_maskReg;
Fill in gpio support for 5416 and later parts: o add output mux support o gpio pin count is chip-dependent o 9280 and 9285 do input handling different o hookup gpio interrupts o no need to save/restore soft led state around reset
2009-02-24 00:33:08 +00:00
uint32_t mask, mask2;
virgin import of ath hal
2008-11-28 00:03:41 +00:00
HALDEBUG(ah, HAL_DEBUG_INTERRUPT, "%s: 0x%x => 0x%x\n",
__func__, omask, ints);
if (omask & HAL_INT_GLOBAL) {
HALDEBUG(ah, HAL_DEBUG_INTERRUPT, "%s: disable IER\n", __func__);
OS_REG_WRITE(ah, AR_IER, AR_IER_DISABLE);
(void) OS_REG_READ(ah, AR_IER);
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_INTR_ASYNC_ENABLE, 0);
(void) OS_REG_READ(ah, AR_INTR_ASYNC_ENABLE);
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_INTR_SYNC_ENABLE, 0);
(void) OS_REG_READ(ah, AR_INTR_SYNC_ENABLE);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
mask = ints & HAL_INT_COMMON;
mask2 = 0;
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
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
/*
* Overwrite default mask if Interrupt mitigation
* is specified for AR5416
*/
mask = ints & HAL_INT_COMMON;
if (ints & HAL_INT_TX)
mask |= AR_IMR_TXMINTR | AR_IMR_TXINTM;
if (ints & HAL_INT_RX)
mask |= AR_IMR_RXERR | AR_IMR_RXMINTR | AR_IMR_RXINTM;
if (ints & HAL_INT_TX) {
if (ahp->ah_txErrInterruptMask)
mask |= AR_IMR_TXERR;
if (ahp->ah_txEolInterruptMask)
mask |= AR_IMR_TXEOL;
}
#else
virgin import of ath hal
2008-11-28 00:03:41 +00:00
if (ints & HAL_INT_TX) {
if (ahp->ah_txOkInterruptMask)
mask |= AR_IMR_TXOK;
if (ahp->ah_txErrInterruptMask)
mask |= AR_IMR_TXERR;
if (ahp->ah_txDescInterruptMask)
mask |= AR_IMR_TXDESC;
if (ahp->ah_txEolInterruptMask)
mask |= AR_IMR_TXEOL;
}
if (ints & HAL_INT_RX)
mask |= AR_IMR_RXOK | AR_IMR_RXERR | AR_IMR_RXDESC;
#endif
if (ints & (HAL_INT_BMISC)) {
mask |= AR_IMR_BCNMISC;
if (ints & HAL_INT_TIM)
mask2 |= AR_IMR_S2_TIM;
if (ints & HAL_INT_DTIM)
mask2 |= AR_IMR_S2_DTIM;
if (ints & HAL_INT_DTIMSYNC)
mask2 |= AR_IMR_S2_DTIMSYNC;
if (ints & HAL_INT_CABEND)
mask2 |= (AR_IMR_S2_CABEND );
if (ints & HAL_INT_CST)
mask2 |= AR_IMR_S2_CST;
if (ints & HAL_INT_TSFOOR)
mask2 |= AR_IMR_S2_TSFOOR;
}
Rework the Global TX timeout handling to look more like ath9k. It correctly now sets the AR_IMR BCNMISC register, along with the GTT register in AR_IMR_S2.
2011-04-18 14:03:05 +00:00
if (ints & (HAL_INT_GTT | HAL_INT_CST)) {
mask |= AR_IMR_BCNMISC;
if (ints & HAL_INT_GTT)
mask2 |= AR_IMR_S2_GTT;
if (ints & HAL_INT_CST)
mask2 |= AR_IMR_S2_CST;
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
/* Write the new IMR and store off our SW copy. */
HALDEBUG(ah, HAL_DEBUG_INTERRUPT, "%s: new IMR 0x%x\n", __func__, mask);
OS_REG_WRITE(ah, AR_IMR, mask);
mask = OS_REG_READ(ah, AR_IMR_S2) & ~(AR_IMR_S2_TIM |
AR_IMR_S2_DTIM |
AR_IMR_S2_DTIMSYNC |
AR_IMR_S2_CABEND |
AR_IMR_S2_CABTO |
AR_IMR_S2_TSFOOR |
AR_IMR_S2_GTT |
AR_IMR_S2_CST);
OS_REG_WRITE(ah, AR_IMR_S2, mask | mask2);
ahp->ah_maskReg = ints;
/* Re-enable interrupts if they were enabled before. */
if (ints & HAL_INT_GLOBAL) {
HALDEBUG(ah, HAL_DEBUG_INTERRUPT, "%s: enable IER\n", __func__);
OS_REG_WRITE(ah, AR_IER, AR_IER_ENABLE);
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)) {
mask = AR_INTR_MAC_IRQ;
if (ints & HAL_INT_GPIO)
mask |= SM(AH5416(ah)->ah_gpioMask,
AR_INTR_ASYNC_MASK_GPIO);
OS_REG_WRITE(ah, AR_INTR_ASYNC_ENABLE, mask);
OS_REG_WRITE(ah, AR_INTR_ASYNC_MASK, mask);
mask = AR_INTR_SYNC_DEFAULT;
if (ints & HAL_INT_GPIO)
mask |= SM(AH5416(ah)->ah_gpioMask,
AR_INTR_SYNC_MASK_GPIO);
OS_REG_WRITE(ah, AR_INTR_SYNC_ENABLE, mask);
OS_REG_WRITE(ah, AR_INTR_SYNC_MASK, mask);
}
virgin import of ath hal
2008-11-28 00:03:41 +00:00
}
return omask;
}
Reference in New Issue Copy Permalink