[PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.

The newer single chip hardware family of chipsets have not been
experiencing issues with power saving set by default with recent
fixes merged (even into stable). The remaining issues are only
reported with AR5416 and since enabling PS by default can increase
power savings considerably best to take advantage of that feature
as this has been tested properly.

For more details on this issue see the bug report:

http://bugzilla.kernel.org/show_bug.cgi?id=14267

We leave AR5416 with PS disabled by default, that seems to require
some more work.

Cc: stable@kernel.org
Cc: Peter Stuge <peter@stuge.se>
Cc: Justin P. Mattock  <justinmattock@gmail.com>
Cc: Kristoffer Ericson <kristoffer.ericson@gmail.com>
Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
Signed-off-by: John W. Linville <linville@tuxdriver.com>
---

Greg, this is the long promised backport of the patch titled
"ath9k: re-enable ps by default for new single chip families" backported
down to 2.6.32.y. This just goes test compiled. Manual backport
was required from the upstream Linus patch since the flag
WIPHY_FLAG_PS_ON_BY_DEFAULT was not used back on 2.6.32 so instead
we use the equivalent hw->wiphy->ps_default bool.

Apologies for the delay, was just stuck with other stuff.

I'll remove this from the stable pending list for 802.11 [1] once
this gets sucked in.

[1] http://wireless.kernel.org/en/developers/stable-pending

 drivers/net/wireless/ath/ath9k/main.c |    7 ++++++-
 1 files changed, 6 insertions(+), 1 deletions(-)

diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
index 15eb245..dba27b7 100644
--- a/drivers/net/wireless/ath/ath9k/main.c
+++ b/drivers/net/wireless/ath/ath9k/main.c
@@ -1538,6 +1538,8 @@ bad_no_ah:
 
 void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
 {
+	struct ath_hw *ah = sc->sc_ah;
+
 	hw->flags = IEEE80211_HW_RX_INCLUDES_FCS |
 		IEEE80211_HW_HOST_BROADCAST_PS_BUFFERING |
 		IEEE80211_HW_SIGNAL_DBM |
@@ -1556,7 +1558,10 @@ void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
 		BIT(NL80211_IFTYPE_ADHOC) |
 		BIT(NL80211_IFTYPE_MESH_POINT);
 
-	hw->wiphy->ps_default = false;
+	if (AR_SREV_5416(ah))
+		hw->wiphy->ps_default = false;
+	else
+		hw->wiphy->ps_default = true;
 
 	hw->queues = 4;
 	hw->max_rates = 4;
-- 
1.6.3.3

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/15/2010 03:19 PM, Luis R. Rodriguez wrote:
> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>
> The newer single chip hardware family of chipsets have not been
> experiencing issues with power saving set by default with recent
> fixes merged (even into stable). The remaining issues are only
> reported with AR5416 and since enabling PS by default can increase
> power savings considerably best to take advantage of that feature
> as this has been tested properly.
>
> For more details on this issue see the bug report:
>
> http://bugzilla.kernel.org/show_bug.cgi?id=14267
>
> We leave AR5416 with PS disabled by default, that seems to require
> some more work.
>
> Cc: stable@kernel.org
> Cc: Peter Stuge<peter@stuge.se>
> Cc: Justin P. Mattock<justinmattock@gmail.com>
> Cc: Kristoffer Ericson<kristoffer.ericson@gmail.com>
> Signed-off-by: Luis R. Rodriguez<lrodriguez@atheros.com>
> Signed-off-by: John W. Linville<linville@tuxdriver.com>
> ---
>
> Greg, this is the long promised backport of the patch titled
> "ath9k: re-enable ps by default for new single chip families" backported
> down to 2.6.32.y. This just goes test compiled. Manual backport
> was required from the upstream Linus patch since the flag
> WIPHY_FLAG_PS_ON_BY_DEFAULT was not used back on 2.6.32 so instead
> we use the equivalent hw->wiphy->ps_default bool.
>
> Apologies for the delay, was just stuck with other stuff.
>
> I'll remove this from the stable pending list for 802.11 [1] once
> this gets sucked in.
>
> [1] http://wireless.kernel.org/en/developers/stable-pending
>
>   drivers/net/wireless/ath/ath9k/main.c |    7 ++++++-
>   1 files changed, 6 insertions(+), 1 deletions(-)
>
> diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
> index 15eb245..dba27b7 100644
> --- a/drivers/net/wireless/ath/ath9k/main.c
> +++ b/drivers/net/wireless/ath/ath9k/main.c
> @@ -1538,6 +1538,8 @@ bad_no_ah:
>
>   void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
>   {
> +	struct ath_hw *ah = sc->sc_ah;
> +
>   	hw->flags = IEEE80211_HW_RX_INCLUDES_FCS |
>   		IEEE80211_HW_HOST_BROADCAST_PS_BUFFERING |
>   		IEEE80211_HW_SIGNAL_DBM |
> @@ -1556,7 +1558,10 @@ void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
>   		BIT(NL80211_IFTYPE_ADHOC) |
>   		BIT(NL80211_IFTYPE_MESH_POINT);
>
> -	hw->wiphy->ps_default = false;
> +	if (AR_SREV_5416(ah))
> +		hw->wiphy->ps_default = false;
> +	else
> +		hw->wiphy->ps_default = true;
>
>   	hw->queues = 4;
>   	hw->max_rates = 4;

I'll have to check and see if my machine run's properly with the 
powersaving tuned on..

Justin P. Mattock

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Kristoffer Ericson]

On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
> 
> The newer single chip hardware family of chipsets have not been
> experiencing issues with power saving set by default with recent
> fixes merged (even into stable). The remaining issues are only
> reported with AR5416 and since enabling PS by default can increase
> power savings considerably best to take advantage of that feature
> as this has been tested properly.

I havent had any issues lately. But that said Ive moved on to
2.6.33, 2.6.34 which seems to work fine. Believe you set
to disabled by default on those?

/Kristoffer

> 
> For more details on this issue see the bug report:
> 
> http://bugzilla.kernel.org/show_bug.cgi?id=14267
> 
> We leave AR5416 with PS disabled by default, that seems to require
> some more work.
> 
> Cc: stable@kernel.org
> Cc: Peter Stuge <peter@stuge.se>
> Cc: Justin P. Mattock  <justinmattock@gmail.com>
> Cc: Kristoffer Ericson <kristoffer.ericson@gmail.com>
> Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
> Signed-off-by: John W. Linville <linville@tuxdriver.com>
> ---
> 
> Greg, this is the long promised backport of the patch titled
> "ath9k: re-enable ps by default for new single chip families" backported
> down to 2.6.32.y. This just goes test compiled. Manual backport
> was required from the upstream Linus patch since the flag
> WIPHY_FLAG_PS_ON_BY_DEFAULT was not used back on 2.6.32 so instead
> we use the equivalent hw->wiphy->ps_default bool.
> 
> Apologies for the delay, was just stuck with other stuff.
> 
> I'll remove this from the stable pending list for 802.11 [1] once
> this gets sucked in.
> 
> [1] http://wireless.kernel.org/en/developers/stable-pending
> 
>  drivers/net/wireless/ath/ath9k/main.c |    7 ++++++-
>  1 files changed, 6 insertions(+), 1 deletions(-)
> 
> diff --git a/drivers/net/wireless/ath/ath9k/main.c b/drivers/net/wireless/ath/ath9k/main.c
> index 15eb245..dba27b7 100644
> --- a/drivers/net/wireless/ath/ath9k/main.c
> +++ b/drivers/net/wireless/ath/ath9k/main.c
> @@ -1538,6 +1538,8 @@ bad_no_ah:
>  
>  void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
>  {
> +	struct ath_hw *ah = sc->sc_ah;
> +
>  	hw->flags = IEEE80211_HW_RX_INCLUDES_FCS |
>  		IEEE80211_HW_HOST_BROADCAST_PS_BUFFERING |
>  		IEEE80211_HW_SIGNAL_DBM |
> @@ -1556,7 +1558,10 @@ void ath_set_hw_capab(struct ath_softc *sc, struct ieee80211_hw *hw)
>  		BIT(NL80211_IFTYPE_ADHOC) |
>  		BIT(NL80211_IFTYPE_MESH_POINT);
>  
> -	hw->wiphy->ps_default = false;
> +	if (AR_SREV_5416(ah))
> +		hw->wiphy->ps_default = false;
> +	else
> +		hw->wiphy->ps_default = true;
>  
>  	hw->queues = 4;
>  	hw->max_rates = 4;
> -- 
> 1.6.3.3

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>
>> The newer single chip hardware family of chipsets have not been
>> experiencing issues with power saving set by default with recent
>> fixes merged (even into stable). The remaining issues are only
>> reported with AR5416 and since enabling PS by default can increase
>> power savings considerably best to take advantage of that feature
>> as this has been tested properly.
>
> I havent had any issues lately. But that said Ive moved on to
> 2.6.33, 2.6.34 which seems to work fine. Believe you set
> to disabled by default on those?
>
> /Kristoffer
>

just tried to add this patch to the latest HEAD, but realized 
ath_set_hw_capab is missing(or I cant seem to grep it)
if you have a patch for the latest tree let me know I can see if I'm 
hitting anything like before.

Justin P. Mattock

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

On Wed, Jun 16, 2010 at 1:12 AM, Kristoffer Ericson
<kristoffer.ericson@gmail.com> wrote:
> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>
>> The newer single chip hardware family of chipsets have not been
>> experiencing issues with power saving set by default with recent
>> fixes merged (even into stable). The remaining issues are only
>> reported with AR5416 and since enabling PS by default can increase
>> power savings considerably best to take advantage of that feature
>> as this has been tested properly.
>
> I havent had any issues lately. But that said Ive moved on to
> 2.6.33, 2.6.34 which seems to work fine. Believe you set
> to disabled by default on those?

2.6.33 and up all have PS enabled by default for all single chip
families. For AR5416 and related chipsets, its left disabled as it was
on 2.6.32 right now. The patch is intended only for 2.6.32.y as the
other kernels 2.6.33.y and up already have the same patch.

  Luis

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
<justinmattock@gmail.com> wrote:
> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
>>
>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>>>
>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>>
>>> The newer single chip hardware family of chipsets have not been
>>> experiencing issues with power saving set by default with recent
>>> fixes merged (even into stable). The remaining issues are only
>>> reported with AR5416 and since enabling PS by default can increase
>>> power savings considerably best to take advantage of that feature
>>> as this has been tested properly.
>>
>> I havent had any issues lately. But that said Ive moved on to
>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
>> to disabled by default on those?
>>
>> /Kristoffer
>>
>
> just tried to add this patch to the latest HEAD, but realized
> ath_set_hw_capab is missing(or I cant seem to grep it)
> if you have a patch for the latest tree let me know I can see if I'm hitting
> anything like before.

The patch would already be merged since 2.6.33.y, this patch is a
backport for 2.6.32.y. I just hadn't had time to backport it yet,
until today.

  Luis

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
> <justinmattock@gmail.com>  wrote:
>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
>>>
>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>>>>
>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>>>
>>>> The newer single chip hardware family of chipsets have not been
>>>> experiencing issues with power saving set by default with recent
>>>> fixes merged (even into stable). The remaining issues are only
>>>> reported with AR5416 and since enabling PS by default can increase
>>>> power savings considerably best to take advantage of that feature
>>>> as this has been tested properly.
>>>
>>> I havent had any issues lately. But that said Ive moved on to
>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
>>> to disabled by default on those?
>>>
>>> /Kristoffer
>>>
>>
>> just tried to add this patch to the latest HEAD, but realized
>> ath_set_hw_capab is missing(or I cant seem to grep it)
>> if you have a patch for the latest tree let me know I can see if I'm hitting
>> anything like before.
>
> The patch would already be merged since 2.6.33.y, this patch is a
> backport for 2.6.32.y. I just hadn't had time to backport it yet,
> until today.
>
>    Luis
>

so its already in there then.. if that's the case then the system runs 
good with powersave on then..

Justin P. Mattock

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
> > On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
> > <justinmattock@gmail.com>  wrote:
> >> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
> >>>
> >>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
> >>>>
> >>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
> >>>>
> >>>> The newer single chip hardware family of chipsets have not been
> >>>> experiencing issues with power saving set by default with recent
> >>>> fixes merged (even into stable). The remaining issues are only
> >>>> reported with AR5416 and since enabling PS by default can increase
> >>>> power savings considerably best to take advantage of that feature
> >>>> as this has been tested properly.
> >>>
> >>> I havent had any issues lately. But that said Ive moved on to
> >>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
> >>> to disabled by default on those?
> >>>
> >>> /Kristoffer
> >>>
> >>
> >> just tried to add this patch to the latest HEAD, but realized
> >> ath_set_hw_capab is missing(or I cant seem to grep it)
> >> if you have a patch for the latest tree let me know I can see if I'm hitting
> >> anything like before.
> >
> > The patch would already be merged since 2.6.33.y, this patch is a
> > backport for 2.6.32.y. I just hadn't had time to backport it yet,
> > until today.
> >
> >    Luis
> >
> 
> so its already in there then..

Right

> if that's the case then the system runs 
> good with powersave on then..

Did you read the patch by chance? It only enables power save for
non-AR5416 hardware. What do you have?
 
  Luis

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/16/2010 10:53 AM, Luis R. Rodriguez wrote:
> On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
>> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
>>> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
>>> <justinmattock@gmail.com>   wrote:
>>>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
>>>>>
>>>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>>>>>>
>>>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>>>>>
>>>>>> The newer single chip hardware family of chipsets have not been
>>>>>> experiencing issues with power saving set by default with recent
>>>>>> fixes merged (even into stable). The remaining issues are only
>>>>>> reported with AR5416 and since enabling PS by default can increase
>>>>>> power savings considerably best to take advantage of that feature
>>>>>> as this has been tested properly.
>>>>>
>>>>> I havent had any issues lately. But that said Ive moved on to
>>>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
>>>>> to disabled by default on those?
>>>>>
>>>>> /Kristoffer
>>>>>
>>>>
>>>> just tried to add this patch to the latest HEAD, but realized
>>>> ath_set_hw_capab is missing(or I cant seem to grep it)
>>>> if you have a patch for the latest tree let me know I can see if I'm hitting
>>>> anything like before.
>>>
>>> The patch would already be merged since 2.6.33.y, this patch is a
>>> backport for 2.6.32.y. I just hadn't had time to backport it yet,
>>> until today.
>>>
>>>     Luis
>>>
>>
>> so its already in there then..
>
> Right
>
>> if that's the case then the system runs
>> good with powersave on then..
>
> Did you read the patch by chance? It only enables power save for
> non-AR5416 hardware. What do you have?
>
>    Luis
>

mine is the AR5008

Justin P. Mattock

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

On Wed, Jun 16, 2010 at 11:15:29AM -0700, Justin P. Mattock wrote:
> On 06/16/2010 10:53 AM, Luis R. Rodriguez wrote:
> > On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
> >> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
> >>> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
> >>> <justinmattock@gmail.com>   wrote:
> >>>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
> >>>>>
> >>>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
> >>>>>>
> >>>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
> >>>>>>
> >>>>>> The newer single chip hardware family of chipsets have not been
> >>>>>> experiencing issues with power saving set by default with recent
> >>>>>> fixes merged (even into stable). The remaining issues are only
> >>>>>> reported with AR5416 and since enabling PS by default can increase
> >>>>>> power savings considerably best to take advantage of that feature
> >>>>>> as this has been tested properly.
> >>>>>
> >>>>> I havent had any issues lately. But that said Ive moved on to
> >>>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
> >>>>> to disabled by default on those?
> >>>>>
> >>>>> /Kristoffer
> >>>>>
> >>>>
> >>>> just tried to add this patch to the latest HEAD, but realized
> >>>> ath_set_hw_capab is missing(or I cant seem to grep it)
> >>>> if you have a patch for the latest tree let me know I can see if I'm hitting
> >>>> anything like before.
> >>>
> >>> The patch would already be merged since 2.6.33.y, this patch is a
> >>> backport for 2.6.32.y. I just hadn't had time to backport it yet,
> >>> until today.
> >>>
> >>>     Luis
> >>>
> >>
> >> so its already in there then..
> >
> > Right
> >
> >> if that's the case then the system runs
> >> good with powersave on then..
> >
> > Did you read the patch by chance? It only enables power save for
> > non-AR5416 hardware. What do you have?
> >
> >    Luis
> >
> 
> mine is the AR5008

Ok I should say, all AR5008 and AR9001 family chipsets have
power save disabled still. Only AR9002 and up have it enabled
now. The patch was doing the same for 2.6.32.

  Luis

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/16/2010 11:16 AM, Luis R. Rodriguez wrote:
> On Wed, Jun 16, 2010 at 11:15:29AM -0700, Justin P. Mattock wrote:
>> On 06/16/2010 10:53 AM, Luis R. Rodriguez wrote:
>>> On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
>>>> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
>>>>> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
>>>>> <justinmattock@gmail.com>    wrote:
>>>>>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
>>>>>>>
>>>>>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>>>>>>>>
>>>>>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>>>>>>>
>>>>>>>> The newer single chip hardware family of chipsets have not been
>>>>>>>> experiencing issues with power saving set by default with recent
>>>>>>>> fixes merged (even into stable). The remaining issues are only
>>>>>>>> reported with AR5416 and since enabling PS by default can increase
>>>>>>>> power savings considerably best to take advantage of that feature
>>>>>>>> as this has been tested properly.
>>>>>>>
>>>>>>> I havent had any issues lately. But that said Ive moved on to
>>>>>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
>>>>>>> to disabled by default on those?
>>>>>>>
>>>>>>> /Kristoffer
>>>>>>>
>>>>>>
>>>>>> just tried to add this patch to the latest HEAD, but realized
>>>>>> ath_set_hw_capab is missing(or I cant seem to grep it)
>>>>>> if you have a patch for the latest tree let me know I can see if I'm hitting
>>>>>> anything like before.
>>>>>
>>>>> The patch would already be merged since 2.6.33.y, this patch is a
>>>>> backport for 2.6.32.y. I just hadn't had time to backport it yet,
>>>>> until today.
>>>>>
>>>>>      Luis
>>>>>
>>>>
>>>> so its already in there then..
>>>
>>> Right
>>>
>>>> if that's the case then the system runs
>>>> good with powersave on then..
>>>
>>> Did you read the patch by chance? It only enables power save for
>>> non-AR5416 hardware. What do you have?
>>>
>>>     Luis
>>>
>>
>> mine is the AR5008
>
> Ok I should say, all AR5008 and AR9001 family chipsets have
> power save disabled still. Only AR9002 and up have it enabled
> now. The patch was doing the same for 2.6.32.
>
>    Luis
>

o.k. I have not tried what powersave does since the last time I looked 
into this bug(if I remember 2.6.31/2 or so) I can manually
turn it on with iwconfig and see..

Justin P. Mattock

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Luis R. Rodriguez]

On Wed, Jun 16, 2010 at 12:14:51PM -0700, Justin P. Mattock wrote:
> On 06/16/2010 11:16 AM, Luis R. Rodriguez wrote:
> > On Wed, Jun 16, 2010 at 11:15:29AM -0700, Justin P. Mattock wrote:
> >> On 06/16/2010 10:53 AM, Luis R. Rodriguez wrote:
> >>> On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
> >>>> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
> >>>>> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
> >>>>> <justinmattock@gmail.com>    wrote:
> >>>>>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
> >>>>>>>
> >>>>>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
> >>>>>>>>
> >>>>>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
> >>>>>>>>
> >>>>>>>> The newer single chip hardware family of chipsets have not been
> >>>>>>>> experiencing issues with power saving set by default with recent
> >>>>>>>> fixes merged (even into stable). The remaining issues are only
> >>>>>>>> reported with AR5416 and since enabling PS by default can increase
> >>>>>>>> power savings considerably best to take advantage of that feature
> >>>>>>>> as this has been tested properly.
> >>>>>>>
> >>>>>>> I havent had any issues lately. But that said Ive moved on to
> >>>>>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
> >>>>>>> to disabled by default on those?
> >>>>>>>
> >>>>>>> /Kristoffer
> >>>>>>>
> >>>>>>
> >>>>>> just tried to add this patch to the latest HEAD, but realized
> >>>>>> ath_set_hw_capab is missing(or I cant seem to grep it)
> >>>>>> if you have a patch for the latest tree let me know I can see if I'm hitting
> >>>>>> anything like before.
> >>>>>
> >>>>> The patch would already be merged since 2.6.33.y, this patch is a
> >>>>> backport for 2.6.32.y. I just hadn't had time to backport it yet,
> >>>>> until today.
> >>>>>
> >>>>>      Luis
> >>>>>
> >>>>
> >>>> so its already in there then..
> >>>
> >>> Right
> >>>
> >>>> if that's the case then the system runs
> >>>> good with powersave on then..
> >>>
> >>> Did you read the patch by chance? It only enables power save for
> >>> non-AR5416 hardware. What do you have?
> >>>
> >>>     Luis
> >>>
> >>
> >> mine is the AR5008
> >
> > Ok I should say, all AR5008 and AR9001 family chipsets have
> > power save disabled still. Only AR9002 and up have it enabled
> > now. The patch was doing the same for 2.6.32.
> >
> >    Luis
> >
> 
> o.k. I have not tried what powersave does since the last time I looked 
> into this bug(if I remember 2.6.31/2 or so) I can manually
> turn it on with iwconfig and see..

Thanks but no need, its not the purpose of the patch.

  Luis

Re: [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Justin P. Mattock]

On 06/16/2010 01:06 PM, Luis R. Rodriguez wrote:
> On Wed, Jun 16, 2010 at 12:14:51PM -0700, Justin P. Mattock wrote:
>> On 06/16/2010 11:16 AM, Luis R. Rodriguez wrote:
>>> On Wed, Jun 16, 2010 at 11:15:29AM -0700, Justin P. Mattock wrote:
>>>> On 06/16/2010 10:53 AM, Luis R. Rodriguez wrote:
>>>>> On Wed, Jun 16, 2010 at 09:01:36AM -0700, Justin P. Mattock wrote:
>>>>>> On 06/16/2010 08:49 AM, Luis R. Rodriguez wrote:
>>>>>>> On Wed, Jun 16, 2010 at 6:33 AM, Justin P. Mattock
>>>>>>> <justinmattock@gmail.com>     wrote:
>>>>>>>> On 06/16/2010 01:12 AM, Kristoffer Ericson wrote:
>>>>>>>>>
>>>>>>>>> On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
>>>>>>>>>>
>>>>>>>>>> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
>>>>>>>>>>
>>>>>>>>>> The newer single chip hardware family of chipsets have not been
>>>>>>>>>> experiencing issues with power saving set by default with recent
>>>>>>>>>> fixes merged (even into stable). The remaining issues are only
>>>>>>>>>> reported with AR5416 and since enabling PS by default can increase
>>>>>>>>>> power savings considerably best to take advantage of that feature
>>>>>>>>>> as this has been tested properly.
>>>>>>>>>
>>>>>>>>> I havent had any issues lately. But that said Ive moved on to
>>>>>>>>> 2.6.33, 2.6.34 which seems to work fine. Believe you set
>>>>>>>>> to disabled by default on those?
>>>>>>>>>
>>>>>>>>> /Kristoffer
>>>>>>>>>
>>>>>>>>
>>>>>>>> just tried to add this patch to the latest HEAD, but realized
>>>>>>>> ath_set_hw_capab is missing(or I cant seem to grep it)
>>>>>>>> if you have a patch for the latest tree let me know I can see if I'm hitting
>>>>>>>> anything like before.
>>>>>>>
>>>>>>> The patch would already be merged since 2.6.33.y, this patch is a
>>>>>>> backport for 2.6.32.y. I just hadn't had time to backport it yet,
>>>>>>> until today.
>>>>>>>
>>>>>>>       Luis
>>>>>>>
>>>>>>
>>>>>> so its already in there then..
>>>>>
>>>>> Right
>>>>>
>>>>>> if that's the case then the system runs
>>>>>> good with powersave on then..
>>>>>
>>>>> Did you read the patch by chance? It only enables power save for
>>>>> non-AR5416 hardware. What do you have?
>>>>>
>>>>>      Luis
>>>>>
>>>>
>>>> mine is the AR5008
>>>
>>> Ok I should say, all AR5008 and AR9001 family chipsets have
>>> power save disabled still. Only AR9002 and up have it enabled
>>> now. The patch was doing the same for 2.6.32.
>>>
>>>     Luis
>>>
>>
>> o.k. I have not tried what powersave does since the last time I looked
>> into this bug(if I remember 2.6.31/2 or so) I can manually
>> turn it on with iwconfig and see..
>
> Thanks but no need, its not the purpose of the patch.
>
>    Luis
>

o.k.

Justin P. Mattock

Re: [stable] [PATCH 2.6.32.y] ath9k: re-enable ps by default for new single chip families

[Greg KH]

On Tue, Jun 15, 2010 at 06:19:19PM -0400, Luis R. Rodriguez wrote:
> commit 14acdde6e527950f66c084dbf19bad6fbfcaeedc upstream.
> 
> The newer single chip hardware family of chipsets have not been
> experiencing issues with power saving set by default with recent
> fixes merged (even into stable). The remaining issues are only
> reported with AR5416 and since enabling PS by default can increase
> power savings considerably best to take advantage of that feature
> as this has been tested properly.
> 
> For more details on this issue see the bug report:
> 
> http://bugzilla.kernel.org/show_bug.cgi?id=14267
> 
> We leave AR5416 with PS disabled by default, that seems to require
> some more work.
> 
> Cc: stable@kernel.org
> Cc: Peter Stuge <peter@stuge.se>
> Cc: Justin P. Mattock  <justinmattock@gmail.com>
> Cc: Kristoffer Ericson <kristoffer.ericson@gmail.com>
> Signed-off-by: Luis R. Rodriguez <lrodriguez@atheros.com>
> Signed-off-by: John W. Linville <linville@tuxdriver.com>
> ---
> 
> Greg, this is the long promised backport of the patch titled
> "ath9k: re-enable ps by default for new single chip families" backported
> down to 2.6.32.y. This just goes test compiled. Manual backport
> was required from the upstream Linus patch since the flag
> WIPHY_FLAG_PS_ON_BY_DEFAULT was not used back on 2.6.32 so instead
> we use the equivalent hw->wiphy->ps_default bool.
> 
> Apologies for the delay, was just stuck with other stuff.
> 
> I'll remove this from the stable pending list for 802.11 [1] once
> this gets sucked in.

Now queued up, thanks.

greg k-h