2012-04-12 18:36:44

by Felix Fietkau

Subject: [PATCH 3.4] ath9k: wake up the hardware from full sleep when idle is turned off

The hardware needs a reset to recover from full sleep. Issue this reset
directly in the ath9k_config call that turns off idle, otherwise tx
remains dead until the first channel change after the idle state change

Signed-off-by: Felix Fietkau <[email protected]>
---
drivers/net/wireless/ath/ath9k/main.c | 9 ++++++++-
1 files changed, 8 insertions(+), 1 deletions(-)

diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
index ba5af8e..fcb13e3 100644
--- a/drivers/net/wireless/ath/ath9k/main.c
+++ b/drivers/net/wireless/ath/ath9k/main.c
@@ -1584,6 +1584,7 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
struct ath_hw *ah = sc->sc_ah;
struct ath_common *common = ath9k_hw_common(ah);
struct ieee80211_conf *conf = &hw->conf;
+ bool reset_channel = false;

ath9k_ps_wakeup(sc);
mutex_lock(&sc->mutex);
@@ -1592,6 +1593,12 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
sc->ps_idle = !!(conf->flags & IEEE80211_CONF_IDLE);
if (sc->ps_idle)
ath_cancel_work(sc);
+ else
+ /*
+ * The chip needs a reset to properly wake up from
+ * full sleep
+ */
+ reset_channel = ah->chip_fullsleep;
}

/*
@@ -1620,7 +1627,7 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
}
}

- if (changed & IEEE80211_CONF_CHANGE_CHANNEL) {
+ if ((changed & IEEE80211_CONF_CHANGE_CHANNEL) || reset_channel) {
struct ieee80211_channel *curchan = hw->conf.channel;
int pos = curchan->hw_value;
int old_pos = -1;
--
1.7.3.2

2012-04-13 08:04:58

by Felix Fietkau

Subject: Re: [PATCH 3.4] ath9k: wake up the hardware from full sleep when idle is turned off

On 2012-04-13 4:18 AM, Sujith Manoharan wrote:
> Felix Fietkau wrote:
>> The hardware needs a reset to recover from full sleep. Issue this reset
>> directly in the ath9k_config call that turns off idle, otherwise tx
>> remains dead until the first channel change after the idle state change
>>
>> Signed-off-by: Felix Fietkau <[email protected]>
>> ---
>> drivers/net/wireless/ath/ath9k/main.c | 9 ++++++++-
>> 1 files changed, 8 insertions(+), 1 deletions(-)
>>
>> diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
>> index ba5af8e..fcb13e3 100644
>> --- a/drivers/net/wireless/ath/ath9k/main.c
>> +++ b/drivers/net/wireless/ath/ath9k/main.c
>> @@ -1584,6 +1584,7 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
>> struct ath_hw *ah = sc->sc_ah;
>> struct ath_common *common = ath9k_hw_common(ah);
>> struct ieee80211_conf *conf = &hw->conf;
>> + bool reset_channel = false;
>>
>> ath9k_ps_wakeup(sc);
>> mutex_lock(&sc->mutex);
>> @@ -1592,6 +1593,12 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
>> sc->ps_idle = !!(conf->flags & IEEE80211_CONF_IDLE);
>> if (sc->ps_idle)
>> ath_cancel_work(sc);
>> + else
>> + /*
>> + * The chip needs a reset to properly wake up from
>> + * full sleep
>> + */
>> + reset_channel = ah->chip_fullsleep;
>
> This doesn't seem to work...
>
> Since IDLE is broken currently, the HW never enters FULL_SLEEP mode and
> reset_channel will always be zero, and a chip reset will be done only via
> IEEE80211_CONF_CHANGE_CHANNEL.
Consider this preparation for fixing idle mode.

- Felix

2012-04-13 02:24:19

by Sujith Manoharan

Subject: [PATCH 3.4] ath9k: wake up the hardware from full sleep when idle is turned off

Felix Fietkau wrote:
> The hardware needs a reset to recover from full sleep. Issue this reset
> directly in the ath9k_config call that turns off idle, otherwise tx
> remains dead until the first channel change after the idle state change
>
> Signed-off-by: Felix Fietkau <[email protected]>
> ---
> drivers/net/wireless/ath/ath9k/main.c | 9 ++++++++-
> 1 files changed, 8 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
> index ba5af8e..fcb13e3 100644
> --- a/drivers/net/wireless/ath/ath9k/main.c
> +++ b/drivers/net/wireless/ath/ath9k/main.c
> @@ -1584,6 +1584,7 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
> struct ath_hw *ah = sc->sc_ah;
> struct ath_common *common = ath9k_hw_common(ah);
> struct ieee80211_conf *conf = &hw->conf;
> + bool reset_channel = false;
>
> ath9k_ps_wakeup(sc);
> mutex_lock(&sc->mutex);
> @@ -1592,6 +1593,12 @@ static int ath9k_config(struct ieee80211_hw *hw, u32 changed)
> sc->ps_idle = !!(conf->flags & IEEE80211_CONF_IDLE);
> if (sc->ps_idle)
> ath_cancel_work(sc);
> + else
> + /*
> + * The chip needs a reset to properly wake up from
> + * full sleep
> + */
> + reset_channel = ah->chip_fullsleep;

This doesn't seem to work...

Since IDLE is broken currently, the HW never enters FULL_SLEEP mode and
reset_channel will always be zero, and a chip reset will be done only via
IEEE80211_CONF_CHANGE_CHANNEL.

Sujith