[PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Bastien Nocera]

http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/focu
s=127896

Signed-off-by: Bastien Nocera <hadess@hadess.net>

---
 net/rfkill/rfkill-gpio.c | 1 +
 1 file changed, 1 insertion(+)

diff --git a/net/rfkill/rfkill-gpio.c b/net/rfkill/rfkill-gpio.c
index d5d58d9..9471024 100644
--- a/net/rfkill/rfkill-gpio.c
+++ b/net/rfkill/rfkill-gpio.c
@@ -168,6 +168,7 @@ static const struct acpi_device_id
rfkill_acpi_match[] = {
 	{ "BCM2E3D", RFKILL_TYPE_BLUETOOTH },
 	{ "BCM2E40", RFKILL_TYPE_BLUETOOTH },
 	{ "BCM2E64", RFKILL_TYPE_BLUETOOTH },
+	{ "OBDA8723", RFKILL_TYPE_BLUETOOTH },
 	{ "BCM4752", RFKILL_TYPE_GPS },
 	{ "LNV4752", RFKILL_TYPE_GPS },
 	{ },
-- 
2.4.3

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Larry Finger]

On 07/13/2015 05:11 PM, Bastien Nocera wrote:
> http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/focu
> s=127896
>
> Signed-off-by: Bastien Nocera <hadess@hadess.net>

Bastien,

The subject of this patch needs a bit more detail. I know that you mean the 
RTL8723B versions of the Bluetooth devices. The subject should say this.

The commit message needs a bit more detail. Referencing the E-mail thread is OK, 
but a bit more is needed. What fails without this change?

To help keep that patches straight, change the next submission to say "[PATCH 
V2]" in the subject. In addition, after the ---, add a section saying what 
changed in V2. That info is not retained in the final commit message in the 
repo, but it can be very useful to a maintainer.

Larry

>
> ---
>   net/rfkill/rfkill-gpio.c | 1 +
>   1 file changed, 1 insertion(+)
>
> diff --git a/net/rfkill/rfkill-gpio.c b/net/rfkill/rfkill-gpio.c
> index d5d58d9..9471024 100644
> --- a/net/rfkill/rfkill-gpio.c
> +++ b/net/rfkill/rfkill-gpio.c
> @@ -168,6 +168,7 @@ static const struct acpi_device_id
> rfkill_acpi_match[] = {
>   	{ "BCM2E3D", RFKILL_TYPE_BLUETOOTH },
>   	{ "BCM2E40", RFKILL_TYPE_BLUETOOTH },
>   	{ "BCM2E64", RFKILL_TYPE_BLUETOOTH },
> +	{ "OBDA8723", RFKILL_TYPE_BLUETOOTH },
>   	{ "BCM4752", RFKILL_TYPE_GPS },
>   	{ "LNV4752", RFKILL_TYPE_GPS },
>   	{ },
>

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Johannes Berg]

On Tue, 2015-07-14 at 08:20 -0500, Larry Finger wrote:
> On 07/13/2015 05:11 PM, Bastien Nocera wrote:
> > http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/
> > focu
> > s=127896
> > 
> > Signed-off-by: Bastien Nocera <hadess@hadess.net>
> 
> Bastien,
> 
> The subject of this patch needs a bit more detail. I know that you 
> mean the 
> RTL8723B versions of the Bluetooth devices. The subject should say 
> this.
> 
> The commit message needs a bit more detail. Referencing the E-mail 
> thread is OK, 
> but a bit more is needed. What fails without this change?
> 
> To help keep that patches straight, change the next submission to say 
> "[PATCH 
> V2]" in the subject. In addition, after the ---, add a section saying 
> what 
> changed in V2. That info is not retained in the final commit message 
> in the 
> repo, but it can be very useful to a maintainer.
> 

And while at it, please use links that include the message-id (e.g.
mid.gmane.org) so it's searchable even if the archive disappears.

johannes

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Bastien Nocera]

On Tue, 2015-07-14 at 15:21 +0200, Johannes Berg wrote:
> On Tue, 2015-07-14 at 08:20 -0500, Larry Finger wrote:
> > On 07/13/2015 05:11 PM, Bastien Nocera wrote:
> > > http://thread.gmane.org/gmane.linux.kernel.wireless.general/12770
> > > 6/
> > > focu
> > > s=127896
> > > 
> > > Signed-off-by: Bastien Nocera <hadess@hadess.net>
> > 
> > Bastien,
> > 
> > The subject of this patch needs a bit more detail. I know that you 
> > mean the 
> > RTL8723B versions of the Bluetooth devices. The subject should say 
> > this.
> > 
> > The commit message needs a bit more detail. Referencing the E-mail 
> > thread is OK, 
> > but a bit more is needed. What fails without this change?
> > 
> > To help keep that patches straight, change the next submission to 
> > say 
> > "[PATCH 
> > V2]" in the subject. In addition, after the ---, add a section 
> > saying 
> > what 
> > changed in V2. That info is not retained in the final commit 
> > message 
> > in the 
> > repo, but it can be very useful to a maintainer.
> > 
> 
> And while at it, please use links that include the message-id (e.g.
> mid.gmane.org) so it's searchable even if the archive disappears.

Didn't know that service, thanks.

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Kalle Valo]

Johannes Berg <johannes@sipsolutions.net> writes:

> And while at it, please use links that include the message-id (e.g.
> mid.gmane.org) so it's searchable even if the archive disappears.

BTW, I recently found out that kernel.org has a this type of service as
well. From https://lkml.kernel.org/:

http://lkml.kernel.org/r/<message-id> - for marc.info links
http://lkml.kernel.org/g/<message-id> - for gmane.org links

-- 
Kalle Valo

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Bastien Nocera]

On Tue, 2015-07-14 at 08:20 -0500, Larry Finger wrote:
> On 07/13/2015 05:11 PM, Bastien Nocera wrote:
> > http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/
> > focu
> > s=127896
> > 
> > Signed-off-by: Bastien Nocera <hadess@hadess.net>
> 
> Bastien,
> 
> The subject of this patch needs a bit more detail. I know that you 
> mean the 
> RTL8723B versions of the Bluetooth devices. The subject should say 
> this.
> 
> The commit message needs a bit more detail. Referencing the E-mail 
> thread is OK, 
> but a bit more is needed. What fails without this change?

Nothing yet, as the current Bluetooth support for this device is
completely in user-space, and doesn't sync up with rfkill.

> To help keep that patches straight, change the next submission to say 
> "[PATCH 
> V2]" in the subject. In addition, after the ---, add a section saying 
> what 
> changed in V2. That info is not retained in the final commit message 
> in the 
> repo, but it can be very useful to a maintainer.


I already submitted this patch? I certainly forgot doing so :)

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Marcel Holtmann]

Hi Bastien,

> http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/focu
> s=127896
> 
> Signed-off-by: Bastien Nocera <hadess@hadess.net>
> 
> ---
> net/rfkill/rfkill-gpio.c | 1 +
> 1 file changed, 1 insertion(+)
> 
> diff --git a/net/rfkill/rfkill-gpio.c b/net/rfkill/rfkill-gpio.c
> index d5d58d9..9471024 100644
> --- a/net/rfkill/rfkill-gpio.c
> +++ b/net/rfkill/rfkill-gpio.c
> @@ -168,6 +168,7 @@ static const struct acpi_device_id
> rfkill_acpi_match[] = {
> 	{ "BCM2E3D", RFKILL_TYPE_BLUETOOTH },
> 	{ "BCM2E40", RFKILL_TYPE_BLUETOOTH },
> 	{ "BCM2E64", RFKILL_TYPE_BLUETOOTH },
> +	{ "OBDA8723", RFKILL_TYPE_BLUETOOTH },
> 	{ "BCM4752", RFKILL_TYPE_GPS },
> 	{ "LNV4752", RFKILL_TYPE_GPS },
> 	{ },

while this is most likely a good start to get people going and testing the hardware, I think long-term some these IDs have to move out of this driver and into a Realtek specific UART slave driver (once we get such a thing upstream).

I think one example would be what Fred is currently doing for the Broadcom UART devices. Essentially we are currently misusing the RFKILL switch for an on/off switch of the GPIO lines. This would be better handled by the Bluetooth subsystem itself when you power on the controller.

Regards

Marcel

Re: [PATCH] rfkill-gpio: Add support for the Realtek 8723 BT

[Bastien Nocera]

On Tue, 2015-07-14 at 17:29 +0200, Marcel Holtmann wrote:
> Hi Bastien,
> 
> > http://thread.gmane.org/gmane.linux.kernel.wireless.general/127706/
> > focu
> > s=127896
> > 
> > Signed-off-by: Bastien Nocera <hadess@hadess.net>
> > 
> > ---
> > net/rfkill/rfkill-gpio.c | 1 +
> > 1 file changed, 1 insertion(+)
> > 
> > diff --git a/net/rfkill/rfkill-gpio.c b/net/rfkill/rfkill-gpio.c
> > index d5d58d9..9471024 100644
> > --- a/net/rfkill/rfkill-gpio.c
> > +++ b/net/rfkill/rfkill-gpio.c
> > @@ -168,6 +168,7 @@ static const struct acpi_device_id
> > rfkill_acpi_match[] = {
> > 	{ "BCM2E3D", RFKILL_TYPE_BLUETOOTH },
> > 	{ "BCM2E40", RFKILL_TYPE_BLUETOOTH },
> > 	{ "BCM2E64", RFKILL_TYPE_BLUETOOTH },
> > +	{ "OBDA8723", RFKILL_TYPE_BLUETOOTH },
> > 	{ "BCM4752", RFKILL_TYPE_GPS },
> > 	{ "LNV4752", RFKILL_TYPE_GPS },
> > 	{ },
> 
> while this is most likely a good start to get people going and 
> testing the hardware, I think long-term some these IDs have to move 
> out of this driver and into a Realtek specific UART slave driver 
> (once we get such a thing upstream).
> 
> I think one example would be what Fred is currently doing for the 
> Broadcom UART devices. Essentially we are currently misusing the 
> RFKILL switch for an on/off switch of the GPIO lines. This would be 
> better handled by the Bluetooth subsystem itself when you power on 
> the controller.

Duly noted. I'll postpone that until I have time to do hardware
enablement on the Bluetooth side of that chipset.